Charles Fain Lehman is a fellow at the Manhattan Institute and a senior editor at City Journal, where he covers crime, policing, and urban policy.

In this week’s conversation, Yascha Mounk and Charles Fain Lehman discuss why Baltimore failed to follow the crime declines that transformed other American cities, why strategic policing works, and how focused deterrence breaks cycles of retaliatory violence.

Watch the conversation below—the full video is behind the paywall on this page!

This transcript has been condensed and lightly edited for clarity.

Yascha Mounk: There is a broad question in the United States about why there was a lot of crime in the eighties and nineties, then a significant drop in crime in many cities across the United States in the following decades, and then possibly a slight uptick since the pandemic—that is somewhat contested and depends on the kinds of crime you look at. You have been trying to think about this by looking at a particular city that I know somewhat well: Baltimore, because I teach there some of the time. Baltimore to me was always a big intellectual puzzle because it had not gone through some of those same transformations. In the 2000s and 2010s, when Washington, D.C. became a lot safer, Philadelphia became a good bit safer, and New York became much, much safer than it used to be, Baltimore kind of did not. Perhaps we can start with that—why is it that even as all of these other cities on the eastern seaboard were really making inroads in reducing murder rates and becoming quite a bit safer, Baltimore at the time did not seem to be following the same positive trends?

Charles Fain Lehman: The first thing I want to emphasize is just the extent to which that is unusual—that Baltimore really struggled with its experience with homicide well after other cities had dealt with homicide as an issue or had substantially reduced the homicide level in their cities. If you look at young Black men in the city of Baltimore compared to other major cities, they experienced far less significant a decline in their risk of death by homicide over the period we are talking about, in the 1990s and 2000s. It is a dramatic reduction. As a result, when you get to even 2013 or 2014, their risk is substantially elevated over the rest of the nation. That is before 2015, when the death of Freddie Gray resulted in citywide protests, some rioting, and a durable crime spike. There have been a number of papers arguing persuasively that whatever little control the city had in the immediate prior period was totally lost yet again for the next several years.

There are different factors that are conceivably going on there. There has long been a struggle over the management, funding, and staffing levels of the BPD—the Baltimore Police Department’s trust in the community and the civilian leadership’s trust of the Baltimore Police Department. That relationship has long been problematic, and I think Freddie Gray was a microcosm of broader community concerns. I once met an MPD police officer in DC who would commute an hour and a half each way to work for MPD because he did not want to work for Baltimore PD. That is sort of the struggle that they had.

Mounk: Perhaps one way of thinking about this is through the comparative lens—this is a classic question of comparative politics as well as many other social science disciplines. Why does something happen in one set of places and not in another? Standard methodological advice holds that you can only explain a difference in outcome by a difference in independent variables, by a difference in causal factors. So perhaps let us make it comparative: what was happening at that time in New York City, in Washington, D.C., and so on that was reducing the crime rate?

Lehman: To a first approximation, the answer is deliberate, strategic policing. New York City is the prime example—the deployment of strategic policing under Bill Bratton, where they actually, in the 1990s, for the first time asked: where is crime really happening? What are the hotspots? How can we deploy NYPD resources, and how can we do that strategically to enhance safety in any given area? Think about Bryant Park or Times Square, where they really surged resources and cleaned up. The same thing happened in other jurisdictions—in LA later on, in Washington, D.C.—sort of across the board. There were efforts to do this in Baltimore in the nineties and two thousands as well.

Mounk: Before we get back to Baltimore, tell us a little bit more about what strategic policing looks like. There is obviously a huge debate about what led to this decline in crime in places like New York and LA. Explain the different kinds of explanations that have been given, and why you think strategic policing in particular is the thing really responsible—according to you, at least—for a significant chunk of that decline.

Lehman: It has to be a significant chunk, insofar as you see a large increase in homicide and other major crimes in the 1960s and 1970s, peaking in the eighties, starting to decline in the nineties, with a precipitous decline through the 2000s and early 2010s. For listeners who may not know: a large portion of that is demographic and structural, by which I mean the baby boomers aging into and out of crime. If you do age adjustment, about half of that effect disappears. People commit more crimes when they are younger; the baby boomers, who were in their peak crime-committing years during that hump, then aged out of it. That accounts for about half of it.

Mounk: To be clear, you are saying not that there is something particular about baby boomers that made them more prone to violence, but simply that there was a particularly large number of young men—a demographic that commits a lot of crimes—and so you would just expect there to be more murders when the number of people who are twenty years old is higher.

Lehman: Exactly. If we are talking systematically about the increase and decrease in violence, that is a big part of the story. Then there are other explanations worth addressing. I think there are a number of them—things like lead and crime, where I just do not think the evidence is there. Then you have to think about specific cities where we see very large decreases in violence. New York again is the paradigmatic example; its crime decline is twice as deep and twice as long as the rest of the nation. Boston is another good example, where there was a precipitous decline, particularly in youth violence in the 1990s.

When you look at what those departments did—and often those departments were talking to each other and communicating about their tactics, often run by the same people, very small circles—their theory was that through the seventies and eighties, policing had been highly reactive. You hear about a crime, you go, you arrest somebody, maybe, maybe you do not. They said: we have these resources and tools, let us use them proactively. Let us identify the people who are frequent flyers—the individuals we know are driving violence disproportionately—and the places where violence is happening disproportionately. Let us go there and focus on those people. If we do that, maybe we will see some substantial improvement. Indeed, many jurisdictions implemented this kind of strategy and saw the substantial improvement you would expect.

Mounk: Why is that so successful? What about the model causes this decrease in crime? It is not immediately obvious why police officers saying, this is where crimes happen, we will hang out there is going to reduce crime that much. If a lot of murders are connected to the drug trade and to gangs, you might think: if a lot of these shootings happen in place A, but we are still in competition with a rival gang and still want to avenge some other crime, we will just carry out a murder somewhere else. There would be all kinds of reasons, if you were telling me about this as an idea before the empirical evidence existed, to think: how can that possibly work? Why does it seem to be working?

Lehman: Your specific question is about displacement—the so-called balloon effect. Why is it that if you do something here, people do not just move over there? A lot of the answer comes down to the fact that murder is objectively very rare. Overwhelmingly, you are unlikely to be murdered, I am unlikely to be murdered. It is extremely unlikely, because murder—gun violence in particular—requires a brew of different things to go wrong in order for it to occur. You need, as I like to say, young men with guns, a culture of honor or culture of beef, and low levels of formal and informal social control. That mix does not occur naturally in most places or at most times, which is why violence is usually concentrated in pockets even within cities—and even within those places, often in a specific set of people.

Violence is not an efficiently produced phenomenon; it is not a market phenomenon. If you do not have the right environment for it, if one person gets pulled out of the social network, it can collapse really easily. In a modern, highly surveilled, enforced society, violence is not the norm—it is the exception. If you take one of the Jenga blocks out, all of a sudden violence becomes much less likely.

Mounk: What is it about this form of policing that is so successful at taking one of those Jenga blocks out? What does it actually look like? Cops know that a particular part of a particular neighborhood is where gang warfare is especially endemic or where there is a particularly high level of crime—they go and they hang out there. Is it to make people feel surveilled? Is it to build relationships? Is it to warn people? What does that look like concretely?

Lehman: “Strategic policing” is really just an umbrella term for police departments thinking proactively—rather than reactively—about how to allocate their resources, identifying places that are problems, and remediating them. One version of this is hotspot patrol policing, for which we have a substantial body of evidence. Basically, you make a heat map of crime, identify where it is happening, and put a cop there. We have a lot of evidence that when that happens, crime levels go down and do not displace. That hotspot is a hotspot because it is an area uniquely conducive to crime. Police are a visible deterrent—in many senses, deterrence is the primary way policing works. If you put a cop in a place, that officer is going to deter criminal behavior there and remove the opportunity for crime to be committed.

That is not the only approach. There is also place-based policing. We have a lot of evidence that the built environment matters a great deal for crime—that environments can be more or less criminogenic. One thing you can do is surge police resources to help clean up, beautify, and facilitate the surveillance of a neighborhood, a park, or an area that has been taken over by crime. That can durably reduce crime. Then there is a third approach—focused deterrence—which is about deterring the spread of violence as a kind of infectious pathogen within a social network.

Mounk: Tell us about some of these in a little bit more detail. You are not a kind of airy-fairy policy analyst, not one of the progressive voices in this debate who thinks the only cause of crime is deprivation and people being too hard on criminals. So it is perhaps somewhat surprising to hear you say that beautifying a public park or changing the built environment really has this big an impact on where crime takes place. In one register, that sounds like something the German artists I grew up around might say—if only our cities were more beautiful, there would be less crime. Yet here is a hard-nosed, right-of-center—if that is how you describe yourself—policy analyst saying that going out and beautifying a park is really going to bring down crime. Why should we believe that? What is it about these architectural interventions, these interventions in public space, that actually prove effective in influencing crime?

Lehman: The first answer is simply that we have a fair amount of high-quality evidence showing that cleaning up vacant lots, forcing landlords to clean up problem properties, and rehabilitating houses all causally reduce crime. The question then becomes: why? Some people argue it affects the psychology of criminals, which I do not think is true. What I do think is that it renders the space more governable. This goes back to Jane Jacobs and The Death and Life of Great American Cities—the idea that what deters antisocial behavior in public is first and foremost the informal surveillance that we all engage in, what Jacobs calls “eyes on the street.” The reason many spaces are not crime-ridden is that there is a high ratio of law-abiding to non-law-abiding people in that space.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

When you clean up a park, restore community space, clear lines of sight, or clear out Penn Station, you allow—or encourage—people who had otherwise been absent from that space to recolonize it. The natural governance that comes from informal social actors is restored. It is about facilitating that informal enforcement, that informal surveillance and coercion that is made possible by having a space that feels like people are welcome in it, can govern it, and feel that it is theirs.

Mounk: It is not about the psychology, then, but about the ability to exercise social control over the most dangerous parts of a city. How does that relate to this third idea you raised—focused policing?

Lehman: One of the things you get out of the criminological literature, as I said earlier, is that violence is really concentrated. A lot of things have to go wrong. A corollary of that is that violence is highly concentrated in social networks—most violence in America today is driven by people who know each other, who are often in conflict with one another, who have beef with one another, who are all dating the same people, who are all from the same blocks, and so on.

As a result, violence becomes contagious within these groups. You and I pick a fight, I shoot you, your friend comes back and shoots me, my friend comes back and shoots him—you get this cycle of escalation. I mentioned Boston earlier; part of how they drove down violence there is a strategy called focused deterrence or ceasefire. The idea is simple: you identify the people in the network who are connected to each other and commit crimes. You go to them and say, we are no longer tolerating violence. If you commit a shooting, we are going to come after you with everything we can. When a shooting happens, you follow up and say, remember what we told you—there are going to be consequences for escalating. You establish a deterrent threat. Implementation fidelity is challenging; it is hard to get this right. But when you do get it right, there are often dramatic reductions in violence, because you have established for that population: we know who you are, we are watching you, there will be consequences if you escalate, so do not escalate. If you stop the escalation—much like stopping a cancer from growing—you prevent many more murders than just the one interaction would otherwise generate.

Mounk: The idea is that one shooting can lead to a whole retaliatory cycle, but because you intervene fast enough, it does not. If you succeed at that, perhaps for six months there are no shootings, and there is no particular reason for another shooting to start.

In the background, I was wondering about the account of what actually drives these shootings in the first place, and how that matters. One way of thinking about it is as something very strategic: you have two different drug gangs trying to expand territory, engaging in a kind of urban warfare where the logic can be modeled with something like a rational choice framework—at a certain level, a very rational set of processes, and you might think you need one kind of intervention to contain it.

On the other hand, you might have a model that is far less rational: people at the age when men are most violent, carrying deadly weapons around for all kinds of reasons, who just happen to run into someone who slept with a girlfriend three months ago outside a bar when both are drunk, and that is where it kicks off. Or someone has the impression that someone else is disrespecting them, undermining their honor, and feels they must defend that honor in order not to become vulnerable to further attacks—and in that moment, they engage in conflict.

If you think a lot of the causes of violence are near-random in that way—where the structure of a situation determines the likelihood of conflict, but each particular conflict is not itself a rational choice—that might lead to a very different model of what it takes to rein in these crimes. Is it clear which of these two types of things really drives the most damaging violence? How do criminologists think about this?

Lehman: The short answer is it is the second one. The way I would push back on your framing is that there is an inherent rationality to honor culture. It is not just impulse—there is a logical form to it.

Mounk: I was starting to get at that, which is that if you allow yourself to be disrespected, you become a target for further disrespect down the line.

Lehman: Exactly. Being aggressive in a way that is antisocial can garner you benefits. The guy who is picking fights on the train, playing his music too loud, getting in somebody’s face, smoking on the train—he is doing it to establish his dominance in the public space. So it is not just reactive, not just: I am going to engage in this beef because otherwise I lose face and there is risk to me. It is also: if I provoke this beef, I will gain face, I will gain relative social standing, and I will be better off.

It is also the case—and there are a bunch of great books on this topic, Jill Leovy’s Ghettoside, and A Code of the Street by Elijah Anderson—that when you are in an environment of relatively low formal social control, where there is relatively low state legitimacy or projection of state power, honor culture is a very natural thing to emerge and to persist over time. When you are in that kind of environment, particularly if you are a young man, you are stuck in the game—you are going to fight back and forth, you are going to be obliged to engage in this sort of dispute of honor, in order to protect yourself but also in order to gain status. I think that is overwhelmingly more important as an explanation for why violence happens in America today than the story about rationally acting drug gangs. People tend to overstate the degree of violence associated with the drug trade. There is a particular political economy of the crack crisis in the 1980s, but that does not always generalize to what every illicit marketplace looks like. Today, a much smaller share of violence is explained by that versus by the honor dynamic.

Mounk: If you are saying it is about honor culture in that way, another way of reframing it is: how do you discourage honor culture, or differently put, how do you reduce the consequences within an honor culture of either having high status or having low status? The strange thing about honor culture is that the examples I can think of where it really governs social interactions include some schoolyards. A lot of what you were talking about applies to the school bully who you just know is crazy enough that they will always escalate—so if they ask for your lunchbox, you are just going to give it to them, because the stakes if you do not are too high. It applies to parts of inner-city America, and there is a lot of interesting sociological and anthropological research on this. It applies to parts of Brittany in France. It also applies to 17th and 18th century European aristocrats. When you read about all of these people dueling—including some of the founders of the United States—that is because they were embroiled in an honor culture where allowing someone to disrespect you had downstream consequences for your life and your reputation, and so you needed to engage in extremely risky acts to protect it.

This is not somehow characteristic of particularly poor neighborhoods mostly populated by a particular ethnic group in the United States today. Very different groups at very different ends of the socioeconomic scale, including some of the most privileged people in 17th and 18th century Europe, have been embroiled in these kinds of honor cultures. One question is: how do you change those cultures such that, as the 18th century gives way to the 19th, the response to an insult is no longer meet me tonight at midnight with your second but rather the bourgeois idea that sticks and stones may break my bones, but your insult will never hurt me—and so you simply smile and say, you are really just degrading yourself, and move on? How do these changes happen?

Lehman: Barry Latzer, the emeritus professor and CUNY criminologist, in his book The Rise and Fall of Violent Crime in America, makes a related argument about why you see honor culture disproportionately in inner-city Black communities in America. His answer is that it is simply southern honor culture—the American South has a long tradition of honor culture, and it is being transplanted northward as people move through the Great Migration. That gets to the point: it is not sui generis to race or specific to race or poverty status. It is about where the culture comes from.

Mounk: As a good Hobbesian, I would have thought it is about the absence of an external authority.

Lehman: I am getting there. What happens in the South is a little more complicated than simply saying “weak state formation”—they do ultimately have weaker states, but it is more nuanced than that. I think that is true also in the other environments you are talking about: honor culture is ultimately the result of the state having insufficient power to be the ultimate arbiter, to resolve all conflicts, or to claim a monopoly on the resolution of conflicts. The reason you do not have an honor culture is because the state undermines any rival claim to settle disputes. The way you get away from honor culture is through legitimate dispute resolution through the courts. How do you get to that bourgeois attitude? Because you have recourse to the legal infrastructure, which is itself undergirded by the state’s threat of violence.

Mounk: Imagining being someone who is growing up embroiled in honor culture and who is aware enough to recognize that it is dangerous—that it can lead to an early death through a duel or a shooting—and who is trying to get out of it: the ability to secure your basic rights, belongings, and safety through authorities is going to govern a lot of what your incentives are. Perhaps here we are in a kind of strategic logic. If the schoolyard bully comes to take your lunchbox and you feel you can tell a teacher who will reliably punish the bully—and doing so will not lead to your classmates turning on you as a snitch—then why fight? But if you think the teachers will not care, or that even if they do, your classmates will ostracize you for snitching and your social life will be miserable, then you face a different calculation. It is not just about the lunchbox today. Once you acquire the reputation of being easily bullied, perhaps you should fight—because that is the only way to protect yourself going forward. That ability to seek recourse against injustice, and to avoid becoming vulnerable by stepping out of honor culture, governs the strategic incentives. That is true for a trivial example like the lunchbox, but it also seems to be true for far more life-and-death situations.

Lehman: The bully analogies are useful because the bully gets something out of it too—it redounds to his benefit to engage in bullying behavior. In some sense, the reason we want states and law enforcement and everything that goes with them is that the predatory conduct of the bully is not something society should have to tolerate. To aggressively abuse the metaphor: we want one very big bully whose job it is to suppress all the others. We want the teacher. Having to live around predators is extraordinarily socially toxic. If you spend your whole life worrying about being victimized, you are much worse off. We know that kids who are exposed to shootings do worse in school, that kids regularly exposed to violence have worse outcomes as adults. There is disinvestment, harm to property values—all of these objective measures that correspond to the intuition that if I have to walk through my community every day worrying that one to five percent of the young men there are going to victimize me at random, I will be much worse off.

Mounk: Let us bring that back to the areas in the United States where there is the most crime. Is this partially about trust between residents of those neighborhoods and the police? If your best friend or your brother is shot and you more or less know who did it, one response is to go to the police and make sure that person goes to jail. Whether the norms of that neighborhood allow you to do that without yourself becoming a target for further violence and retaliation is going to strongly influence whether, faced with tragedy, you go to the police—or decide that is not a route available to you, and instead get in a car with a couple of friends and go looking for the person who did it.

Lehman: When we are talking about the modern era of high violence—the great crime wave of the sixties, seventies, and eighties—there are a couple of different factors going on. One is that, to return to this idea of governance, many parts of America’s big cities become ungoverned or ungovernable. There is a systematic decay of both the small things and the large things. Think about New York in the 1980s—or cities that never recovered, like Detroit, or Gary, Indiana, or large parts of Washington, D.C.—that become just totally lost. There is a fixed stock of police resources, and if crime becomes sufficiently bad, the police may be inadequate to addressing the issue in the stable equilibrium. You have to do something to shift the equilibrium, because in the stable equilibrium the police are simply outmatched.

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

In that period, that dynamic is exacerbated by—to put it oversimplistically—our collective belief that policing does not do anything. This is a real position held by many people of influence in American society in the sixties, seventies, and eighties: that policing has no effect on crime, that the criminal justice system cannot work until you address the root causes of crime. This goes back to the Kerner Commission and the Katzenbach Commission in the 1960s under the Johnson administration: until you address the root causes—poverty, deprivation, racism—you will not have any impact. So not only has crime gotten bad enough that the corresponding police response is relatively ineffective in equilibrium, but we have also decided that police cannot do anything, and so police gravitate toward the least useful things to do with their time—and are often quite corrupt on top of that. It is not a good equilibrium on either side of the equation.

Mounk: There is a huge debate on this in Europe and in Britain as well. One of the things that marked out Tony Blair’s third way was a famous soundbite from the run-up to his 1997 victory: tough on crime, tough on the causes of crime. He was essentially saying we do not need to choose between those—but you can only understand the impact of that sentence because the conventional wisdom in the Labour Party for the previous twenty years had been that it is just about the causes of crime, not about reining crime in directly.

To people who might be tempted by that position: why is it unhelpful to think that we simply need to eliminate the deeper causes of crime? Certainly, when I look at relatively affluent countries—Switzerland, Singapore, Japan—there are much lower levels of crime, perhaps because they have wealthier states and less social deprivation. What role do the root causes of crime play in bringing about the relatively high levels of crime in the United States—even in places where things have improved a great deal? New York is much safer than it was, many cities in America are much safer than they were, and yet they remain a lot less safe than Zurich, Singapore, or Tokyo.

Lehman: Two things. One is that there are structural factors that explain that variance and that matter—though I tend to point to more direct ones. We have a lot more guns in the United States than peer countries. I am actually a squish on guns and the Second Amendment relative to others on the right, but it is a fact of American life, and it explains a lot of the variance in violence specifically—it is just much easier to harm other people in the United States than in peer countries. That does not explain all of it, but it explains some. There are structural factors that, plugged into a model, will explain some of the variance.

That said, it is important to observe that rising affluence is not necessarily correlated with receding crime. There is a great line from James Q. Wilson, the political scientist, who observed that in the 1960s we dramatically overhauled the criminal justice system, made it much less punitive and much more rehabilitative, launched the Great Society, and made huge investments in the inner city—in welfare, in transfers—through the first half of the decade. Then crime exploded. Similarly, in the 1990s, inequality was rising, there were real problems with poverty particularly in major cities, the poverty rate went up under Giuliani in New York City—and yet crime dramatically declined.

You can talk about the big structural variables, but they are relatively hard to move. Part of my argument is that crime is policy-accessible. We actually know today, in a way we did not in the seventies, what tools work to meaningfully reduce crime. We know we can use policy to bring crime down. Last point: crime is itself a cause, not just an effect. Crime is a cause of deprivation. If you live in a crime-ridden neighborhood, you are systematically worse off, systematically disadvantaged. Policy interested in remediating disadvantage should say: we need to free children from the risk—and the fear—of being shot and murdered. That is a disadvantage they should not have to suffer. Policy can change crime in a way that it is much harder to change those other core variables. We should understand crime as an independent cause of the so-called root causes, not just an effect of them.

Mounk: To spell this out a little: it sounds like you are making two different kinds of responses. The first is that looking at the root causes does not give you much explanatory power. You look at times when the supposed root causes improve—the United States in 1970 had much more generous welfare programs and was much less overtly racist than twenty years previously—and yet crime is higher in 1970 than in 1950. Then you look at the period when crime really declines, and it is not obvious that those factors had gotten much better. So the root causes are just not that explanatory.

The second point, which is slightly different but important, is that sometimes the root causes are not the things that are actionable. I have drawn on this James Q. Wilson example in talking about a very different topic—populism—in some of my work. Perhaps the root causes of populism are globalization, or whatever; I am not sure I fully buy that argument, but even if it were true, we are not going to undo globalization in order to save ourselves from populism. It is just not clear how that might work. But you might learn how to run a different kind of political campaign, or adopt certain kinds of policies—moderate political parties might learn how to speak to people in ways they find less off-putting. Those are interventions that are doable in a way that changing the entire global structure is not.

Whatever exact role the root causes play, what we do have is a lot of empirical evidence that specific interventions—focused policing, changing the architectural environment of cities, warning people who are at particular risk of committing a serious crime that there will be consequences—can actually reduce crime.

To bring it back to Baltimore: this is going on in New York, in LA, in Boston, in Washington, D.C., and these cities transform quite radically. There are still some corners of New York that are quite violent, but whenever friends come to visit and ask if they should be worried, I say you would really have to set out to find a dangerous place to end up there. That is very different from what New York felt like in 1990. Washington, D.C. is an even more extreme case—it used to be that the area around Dupont Circle was relatively safe, but only on particular blocks, and straying a couple of blocks in the wrong direction was genuinely dangerous.

Lehman: There were crack markets behind the White House.

Mounk: Right. Washington, D.C. today has also radically transformed. Yet throughout all of this, Baltimore did not follow. Is it that people in Baltimore just did not learn these lessons? What happened there?

Lehman: It is a really great question, one to which I only have part of the answer. There were several parts. One is that, as I alluded to earlier, Baltimore had real problems making the Baltimore Police Department an effective tool for doing some of these things, just at an institutional and cultural level. BPD has had real scandals, real struggles, and a lot of adversarialism between BPD and the civilian leadership of the city. As a result, they never quite built the institutional infrastructure needed. Conversely, when they tried to do these things, they could not build the executive support they needed. We talked about focused deterrence; I have written about their most recent focused deterrence implementation, which I argue works. Everyone I talked to in the city said: we have done this before, we did it in 1999, we did it in 2014, and both times the implementation fell apart. The reason was they did not have executive support, they did not have people at the mayoral level who said, we believe violence is a problem and we are going to put our shoulders into it throughout city government. So they did not get the resources they needed, and there was a great deal of skepticism. That is the challenge of strategic policing and more strategic approaches generally: it takes deliberate action, institutional know-how, and institutional commitment. Without those, you can get stuck in a quagmire. All of that is downstream of the more deeply entrenched problems, which is simply that many neighborhoods in Baltimore have deeply entrenched gang cultures. When I say gang cultures, I mean these tight social networks of young men with guns who all know each other and are all primed to shoot at each other, because the city government has not been committed to solving the problem.

Mounk: This is a story of the city not quite getting its act together because of structural features that do not allow it to effectively deploy the tactics that have worked elsewhere, even when it tries. If we had recorded this conversation two or three years ago, that is roughly where it would have stopped. Until quite recently, Baltimore was still the most dangerous major city on the eastern seaboard. As Washington D.C., New York, and Boston had improved very significantly, Baltimore had improved a little from the heights of crime in the nineties, but was just notably less safe than those other major cities. It turns out, though, that over the last few years, there have been the beginnings of quite a remarkable success story. What is the evidence that Baltimore is finally turning it around?

Lehman: The most straightforward answer is that you can draw a straight line down in terms of the homicide rate. In 2022, when things started to turn, there were about 333 murders in Baltimore. Last year, three years later, there were 133 murders, a more than 60 percent decline. In count terms, that is the fewest murders Baltimore has seen since 1965, a stunning and dramatic decline. If you talk to people on the street, and there is some polling, less than I would like, and if you go and talk to people who live in Baltimore’s most violent communities, they will tell you things feel safer, that they can go out and walk around. The change is felt particularly in those places. It is not every category of crime, not that all measures of antisocial behavior have gotten better across the board. It is that the city has said we are no longer going to tolerate murder, and murder has precipitously declined, and shootings have declined in kind.

Mounk: We talked about the structural features that made it hard for Baltimore to implement some of these things: the tension between the Baltimore Police Department and its leadership, the high levels of corruption relative to other police departments on the eastern seaboard, and so on. One thing I have heard anecdotally, though I do not know whether there is reliable evidence for it, is that a lot of Baltimore Police Department officers do not live in the city of Baltimore but in Baltimore County, and so do not feel the same sense of ownership that perhaps the average NYPD officer might, being more likely to live in the city. Presumably none of that changed overnight. So what did change? What allowed Baltimore to finally implement some of the successful tactics from other places?

Lehman: There are two answers. One is that they got a new prosecutor who made a real difference. A prosecutor is a kind of gating function: a good prosecutor on its own will not make a difference, but a bad prosecutor can cause real problems when you are trying to fix this situation. So they addressed that, and then they built out an implementation of focused deterrence that actually has executive buy-in, that is a genuine all-of-government effort, and that, for the first time, makes a real effort to get the implementation details right rather than just attempting it and not really doing it. Everyone tends to pick one or the other of those two factors, and I say no, it is probably both, they are probably reinforcing each other, and the bigger picture is that they are simply trying to run the criminal justice machine properly. When you do that, things get better.

Mounk: Tell me about the role of prosecutors here. How was the previous prosecutor in Baltimore failing, and how has the new prosecutor made a difference? More broadly, you have argued that the movement to install progressive prosecutors in major American cities has backfired and led to significant rises in crime. What is the evidence for that?

Lehman: Baltimore’s state’s attorney is a man named Ivan Bates, elected in 2022 and taking office in 2023. He replaced Marilyn Mosby, who left under a cloud of scandal, having been under multiple federal investigations and ultimately convicted federally of various fraud charges.

Subscribe now

Mosby came in in 2015 as a self-described progressive prosecutor. The idea behind the progressive prosecutor movement is that prosecutors in the United States have prosecutorial discretion: they can decide whether or not to prosecute offenses. Progressive prosecutors argue you can use that power to unilaterally limit the reach of the criminal justice system, or simply stop prosecuting whole classes of crimes.

What do we know about progressive prosecution? It is a little thorny, because it is very easy to call yourself a progressive prosecutor without implementing any particular policy mix. There are a couple of studies that look at the effect of electing a self-identified progressive prosecutor on crime in a given jurisdiction. Depending on which study you prefer, the findings show an increase in petty or disorder offending, or an increase in property offending. Neither study finds an increase in violence. My response to this is that many progressive prosecutors run on things that do not make a big difference. If you promise to stop prosecuting jaywalking, for example, the response is that American prosecutors do not prosecute jaywalking anyway, so you are not making a policy change of any meaningful significance.

But then you have to look under the hood at what they are actually doing, and there are decisions that can be really impactful. Refusing to prosecute felons in possession, for instance: if you are a felon, you cannot carry a firearm, and some big-city prosecutors will refuse to prosecute that charge even though it is a way to get a criminal with a gun off the street. Or categorical non-prosecution of drug-related offenses, or not using gang enhancements. In Mosby’s case, I think she was fairly aggressive, particularly after 2020, stopping prosecution of a whole variety of so-called low-level offenses. That, concurrent with de-policing in 2020 and everything else that drove the increase in violence at the time, led to a similar increase in violence in Baltimore as many other jurisdictions saw.

When you talk about the role of a prosecutor, you need the prosecutor as a backstop, the person who establishes the credibility of the threat of arrest. When you do not have a prosecutor likely to send people to prison, the credibility of that threat declines significantly and it becomes harder for police to do their job.

Mounk: In terms of the impact of these things, let me steel-man the case on the other side. There is a widespread conviction, often based on somewhat shaky empirical evidence, that a huge share of the U.S. prison population is there because of nonviolent drug offenses. This is partially rooted in Michelle Alexander’s book The New Jim Crow. If you believe that, it seems very reasonable for prosecutors to say we should not be locking all of these people up because they are drug addicts who sadly use drugs, and some cop encounters them with a few grams of marijuana or a more serious substance in their pocket, and we are locking them away for a really long time. That seems genuinely bad.

More broadly, the argument would be that things like gang enhancements are racially discriminatory: that they are simply a way of discriminating against Black men who grew up in poor neighborhoods with a lot of gang presence, where it is inevitable that some of their associates or acquaintances are gang members, meaning the same crime gets punished far more harshly than the same crime committed by a 17-year-old white teenager in an affluent suburb making a dumb mistake and getting off with a warning. Why should one person go to prison for far longer than the other? That is the steel-man. Why does that story miss the mark?

Lehman: The answer is that it simply does not line up with the empirical facts. I once did the work of trying to run down some of the claims in Michelle Alexander’s book. I may get this wrong, but I believe she claims that the majority of the increase in the prison population since the 1980s is attributable to drug offenses. If you go into the footnotes and figure out where she got that number, she clearly misread the relevant table, and the actual figure is much smaller. There is a lot of that kind of thing in the literature, where you poke around and the empirics are fairly shoddy.

The bigger picture is: when you look at who is really getting prosecuted for the kinds of offenses we are talking about, who was the actual object of prosecution in the war on drugs, my argument has been that the war on drugs was strategically misguided. It did not accomplish its primary goal of aggressively suppressing the drug trade. But it gets a bad rap for a bunch of other reasons that are not well supported. One of those is the assumption that we were locking up low-level nonviolent drug offenders for very long periods. There is a study I find useful that looks at the composition of the prison population in 1997, examines drug offenders, and finds that something like 2 percent of state offenders and 6 percent of federal offenders can be characterized as nonviolent, low-level drug offenders. Everyone else has a violent prior, is in a gang, was involved at the trafficking level, or was some sort of major offender.

That gets to the deeper issue: the reason for prosecutorial discretion is that you need to be able to make decisions about who is and is not worth your time, and that is a person-level decision as much as an offense-level decision. Those critiques often look at what people get locked up for, what they get arrested for, but they cannot observe who is actually getting arrested and locked up. There are lots of arrests for drug possession every year, and almost nobody goes to prison for drug possession. Why? Because drug possession arrests are mostly a pretext to bring in someone you need to bring in, since it is easier to establish probable cause for drug possession than for the thing you actually need to lock them up for. That is a common dynamic.

Mounk: Presumably, for some portion of that 2 percent of state prisoners and 6 percent of federal prisoners, the nonviolent conviction may also be the result of a plea deal: you likely have evidence of a violent crime, you are not entirely sure you will win the prosecution, or there is a lot of pressure on the system to avoid jury trials because they are so much more lengthy and costly, and so what you agree to in the plea deal is a nonviolent drug offense, even though the underlying reason for the prosecution is something quite different.

Lehman: That is also often the dynamic, and it gets to the deeper point that in many cases what looks strange on paper is happening because of circumstances outside of what the paper is measuring. The deeper conversation is about how much discretion the criminal justice system should have. Often what progressive prosecutors are doing is saying this whole category of offense, this whole tactic or strategy, is going to be off limits, when a previous prosecutor would have said, I am going to use this as leverage toward my overall goal of incapacitating people who really need to be incapacitated. The progressive prosecutor worries a great deal about false positives, but as a result incorporates a lot of false negatives: this person should not be locked up because I do not like the charge we would use, I do not feel we can get them on the bigger thing, when a prior prosecutor would say, I can get them on this, and they need to go away for five to ten, so I am going to get them on this.

In the rest of this conversation, Yascha and Charles discuss how to balance public safety whilst minimizing the risk of false convictions, and whether crime in major cities will continue to decline. This part of the conversation is reserved for paying subscribers…

Read more

In this week’s episode of The Good Fight Club, Matthew Yglesias, Claire Ainsley, and Yascha Mounk examine why center-left parties are losing groun…

Read more

Mark Leonard is co-founder and director of the European Council on Foreign Relations, the first pan-European think–tank. His latest book is Surviving Chaos: Geopolitics When the Rules Fail.

In this week’s conversation, Yascha Mounk and Mark Leonard discuss why Europe is behind, the global impact of China’s rise, and whether Europe can learn to defend its…

Read more

In this special episode of The Good Fight, recorded at the How The Light Gets In Festival, Roger Hearing moderates a debate between Curtis Yarvin, Minna Salami, and Yascha Mounk on whether liberalism can ever be neutral, what a truly free society would look like, and whether liberalism’s heyday is over. Find out more about the Institute of Arts and Idea…

Read more

Samuel Moyn is the Kent Professor of Law and History at Yale University. His books include Humane: How the United States Abandoned Peace and Reinvented War, The Last Utopia, and Not Enough: Human Rights in an Unequal World. Cohost of the Digging a Hole podcast, he is a frequent contributor to The New York Times and many other publications.

In this week’s conversation, Yascha Mounk and Sam Moyn discuss whether a truly fair democracy might weigh different citizens’ votes differently, whether the emphasis on human rights have got us into the mess we’re in today, and to what extent our democracy is in danger from populism.

This transcript has been condensed and lightly edited for clarity.

Yascha Mounk: So why are the old people the problem, Sam?

Samuel Moyn: I don’t think old people are a problem. I think gerontocracy is—in the same way that whites may not be a problem, but white supremacy is real.

My goal is to show how older people actually sometimes suffer too under a regime of gerontocracy, and to look at how that regime came about and what its current form is—not just for the sake of younger folks, but also for the mass of older people themselves.

Mounk: All right, I’ll refrain from being flippant for the rest of the conversation. Walk us through some of the ways in which America is ruled by gerontocracy today.

Moyn: In the book I look at the first coinage of the term. It’s not Greek, but it’s based on Greek words. It’s supposed to add a dimension to our classification of forms of government like monarchy or oligarchy. The person who coined it made it very clear that it was taking a different form in modern times, having to do with the organization of the electoral system—how people have elections in a democratic society—and then a lot to do with how wealth is allocated and transmitted. I definitely talk about the politicians of our day, who in America are old men and women, but I’m mainly interested in showing that gerontocracy is systemic and that it affects what it means to go to the polls, the outcomes, how elections are financed, and how our economy is organized. I have this sense that we should think about power in the broadest sense—rather than just looking at who’s in office, as the Greeks did—and look more at who has power, meaning effective control of the lives of most people.

Mounk: Let’s go through some of these areas. As you’re saying, politics is the most straightforward, and then there are ones you think are perhaps more important or deeper. But just at a purely descriptive level, what’s the evidence that old people are ruling us in Congress, in the White House, and so on, in a way that hasn’t been the case in past eras?

Moyn: The most graphic evidence, if we’re talking about the United States, is the two most recent presidents. There had really only been one very old man in the presidency before Ronald Reagan, and that was William Henry Harrison, who died in short order of a cold in the nineteenth century. With Reagan we began to see older and older men in office, capped by the last two presidents, Joe Biden and Donald Trump. Now, in fairness, more than half of humanity is ruled by an old man over 70 at this point, largely because of the aging of Vladimir Putin, Xi Jinping, and so forth. The American syndrome is specific.

What I try to do—and this I think is tolerably well known—is go beyond the focus on Biden and cognitive decline to ask just how unrepresentative the political class generally is in the United States, looking in particular at Congress and the federal judiciary. What I’ve tried to do in the book across all these topics is provide the receipts, mainly generated by other people, and gather them in one place. Congress is aging; the Supreme Court is aging in the sense that even when appointed young, the justices stay for a long time.

Aside from cognitive decline, the risk of death is a very serious aspect of the problem. Donald Trump’s one big beautiful bill was facilitated in its passage by the death in droves of Democrats in the House in the six months before that bill. Ruth Bader Ginsburg, notorious for other reasons, died setting up the end of federal abortion rights. All of that pales in the end, if we’re talking about the political class, beside representational fairness, because cognitive decline or death are just the flagrant symptoms of a broader political class that’s old. Very few people in their thirties and forties hold any position of significance in the federal government. That’s the story when it comes to politicians.

Mounk: What do you think is the explanation here? Obviously, when it comes to the presidency, you said that’s been the trend since Ronald Reagan. Reagan was quite old when he assumed the presidency, but the three presidents afterwards were relatively young. He was followed by Bill Clinton, George W. Bush, and Barack Obama, all of whom were reasonably young men at the time they took office. Some of the dynamics in the Senate, and perhaps in the House of Representatives, have to do with very specific American institutions—though I think that’s harder as an explanation for the presidency. What I mean is that if you’re an incumbent senator, particularly in a safe state, of which there are now very many on both sides, it’s very hard to displace you. If you’re 70 or 80 years old and running for reelection, you probably wouldn’t win a primary contest when you’re first being elected, but with 25 years of service and all the favors owed to you and all the fundraising muscle at your disposal, you may be able to hold off a primary challenge—and then you’re going to sail to reelection in Texas or Massachusetts, if you’re in the right political party.

The presidency is harder to explain. In 2020, Joe Biden faced a lot of younger rivals for the Democratic nomination, and yet voters in the primaries freely chose him. In 2024, Donald Trump faced a number of opponents who were quite a bit younger than him, but he was clearly able to retain the loyalty of the Republican primary electorate to such an extent as to gain reelection. I haven’t seen analyses showing a very obvious age gradient in which primary voters voted for those candidates. In 2024, for example, it’s not clear to me that the people who voted for Trump in Republican primaries were systematically older than the people who voted for his rivals. Even in the general election itself, there was less of an age difference than we might have expected, and most of that is produced by partisanship rather than a preference for a particular age. So what do you think produces this? If this is the free choice of American voters, I’m free to disagree with that choice—and I think there’s good reason to disagree in both of those cases—but it’s somewhat unclear that we shouldn’t just leave it to American voters to make their own decisions.

Moyn: One reason I’m interested in pushing beyond politicians in general, and not just the president, is that I agree there’s a lot of contingency and idiosyncrasy in the political story—and yet it’s the one that, to the extent Americans care about this politics, they care about. So I’m using it as a door.

I agree with you, and I’d even add that if Donald Trump were to fulfill the fondest hopes of those who’ve resisted him and depart or die within a year, we’d immediately have the youngest president in U.S. history, because J.D. Vance would enter the office at an age younger than Theodore Roosevelt, who currently holds the record for youngest president.

Mounk: When you look at some of the leading Democratic contenders for the nomination in ’28, they go from middle-aged men—like Gavin Newsom—to very young contenders like AOC.

Moyn: Absolutely. The foundations of what I’m thinking of as the structure of gerontocracy—as a structural phenomenon that goes beyond these contingencies and therefore leads us beyond the level of politicians—the main driver is going to be the aging of humanity, and certainly of Americans. I would push back a bit on the moral you seem to hypothesize when it turns out that there’s a lot of contingency in accounting for the age of politicians.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

Voters only select and express their preferences within a system, and we have to critically inquire into that system. Let’s continue to take my analogy of white supremacy as something many people would acknowledge exists and may be systemic in the way I’m saying gerontocracy is. If you said that, after the Supreme Court’s decision in Callais, no Black politicians are being elected in the Deep South anymore—is that the will of the people, or is it because of the way the will of the people has been organized in an electoral system?

As far as I see it, it really matters that there are advantages of incumbency—which you mentioned—which means that especially when politicians in America are starting their careers later, you’re just going to have an older set of politicians than the electorate and certainly more than the population. Then we’d have to get into campaign finance, where the facts are glaring about just how old the median campaign donor is: over 60, sometimes higher. I devote a whole portion of the book to thinking about gerontocracy at the political level, not just in terms of the politicians, but in terms of the institutions. We do literally have a branch of government named after old men. We do have age minima in the Constitution for federal political office, with no age maxima. Those are also features of the way we’ve organized political choice. We can’t say “the will of the people” as if that’s pre-existing, because it utterly depends for its expression on the institutions we’ve devised and the rules we’re following to figure out what it is.

Mounk: Perhaps we’ll get into a little bit later what we should do about those institutions, what kinds of ways to reform them. Certainly when it comes to something like abolishing age minima, I think there’s a very strong democratic and liberal reason to abolish them. Some of the suggestions you make in the book are rather more ambitious, and we may have different views about them, but we can get into that.

Let’s get beyond politics. As somebody who is an immigrant to the United States having grown up in Europe, I’ve always been struck by how open American society is to the talents and contributions of young people. I feel that a little bit less now that people in my age group in Germany and France and so on are coming to have genuine roles of importance and responsibility in society. But I was struck when I was 30 years old that I had many friends and acquaintances in the United States who already had positions teaching at major universities, positions editing publications or at least important sections of publications, and very significant roles in law firms and investment banks. Obviously, when you look today at Silicon Valley, you have an enormous number of people in their thirties who have about as much power and influence as any human being has had in the history of humanity by certain standards.

This stood in contrast to a lot of 30-year-olds in Germany still being interns, to somebody getting to be the editor of a newspaper section of a major publication at the age of 30 being an incredibly exceptional achievement. When I worked in theater after college for a year, my first task as assistant director was to hire an intern—I was 21 years old, and the intern we ended up hiring was 29. So to what extent is it true, when you look at American society more broadly, that there is this kind of gerontocracy? Isn’t America still in many ways the country of the young, where people have enormous opportunities if they have the right kind of education and perhaps the right kind of background and the various advantages that play into that in a system that is not always as meritocratic as it claims to be? Isn’t it actually striking how much influence and power and money a lot of young people have in the United States compared to anywhere else in the world?

Moyn: Oscar Wilde famously said that the youth of America is its only tradition. I don’t think that’s true—I see the U.S. Constitution as a gerontocratic document in the ways we’ve already addressed, and that’s a big tradition—but I agree with you. It’s one reason I’m interested in thinking about the American case, that at times America has stood for launching youth and providing careers open to talent. I think you might overstate the transatlantic gulf, insofar as I think the whole point of modernity was to topple elder rule and put in place opportunities for young people. The whole modern novel and its roots in the so-called Bildungsroman is about the opportunities that young men could enjoy after the revolution, on the model of Napoleon.

The question is what’s happening now. The demography, I think, is unsuitable for the picture of America you’re saying still prevails. I’m not denying it: I became an Ivy League professor at twenty-nine and got tenure shortly after, so I have had enormous opportunities as a youth. Many of my students have not. Some have been the editors of those publications you mentioned, but in part because they were blocked from becoming professors in a way I wasn’t. The general story has to be that our demography is changing, with fewer young people and an overwhelming number—absolutely and relatively—of aging people who are not suffering decline and are not forced out of their power. The baseline is changing, and that’s the very modest claim I’m making.

Mounk: I wonder whether we need to distinguish between areas of American life and the American economy where you’re not just distributing a fixed number of positions of influence and opportunity, and ones in which you are. For a variety of reasons, I think universities—and particularly humanities departments within universities, and particularly humanities departments studying the sorts of things that you and I studied when we were applying for faculty jobs—are just at the extreme end of that distribution. Universities have not increased their numbers in part because it is an oligopoly that is trying to protect its prestige. The faculty of Columbia University recently voted by something like 80 or 90 percent against admitting more students to the opportunities that university may give to people. The number of students studying the humanities has crashed over the last 20 or 30 years for a whole set of reasons within the humanities. Political theory and European intellectual history—your discipline—are very much out of fashion. In those very specific areas of American life, there is indeed a highly limited number of positions and opportunities, and if professors no longer have to retire and some of them choose to stay in post until they’re 75 or 80 or 85, that means young doctoral students don’t have an opportunity to take up those faculty jobs.

I wonder whether the lesson of that is that universities should actually expand opportunity, that the humanities need to reform themselves to attract undergrads, that the institutional priorities of who’s getting hired are sometimes wrong, that we should hire more professors rather than more administrators—lessons which are more specific to this particular case rather than the broader thesis. When it comes to the broader economy, it doesn’t strike me that most talented, driven 25- or 30-year-olds don’t have enough opportunities to go and join banking or artificial intelligence, or that there’s not enough capital for them to do a startup, or that they can’t become pharmacists or teachers and so on. To what extent are we overgeneralizing from the world that you and I know?

Moyn: That’s a fair question, and I don’t mean to stake my case on academic gerontocracy—in the few pages I address it, it’s really treated as its own distinctive, idiosyncratic phenomenon. However, aspects of it turn out to be more familiar than your binary presentation suggests. I agree that we are a bit illusioned about the general picture because of certain vanguard sectors of the economy that do indeed provide youthful opportunity—Silicon Valley would be the classic example—but most of the analysis I give of the job situation is making the point not about the academic scene but about American business more generally, which has features that resemble universities to the extent that there are apex positions and a pyramid of authority. The higher you go, the older you are, in the last thirty years. In part, the reason is that across these sectors, mandatory retirement—which was once a familiar element of the American employment landscape—was abolished in the later nineteen-eighties through the early nineteen-nineties. More generally, all of these sectors involve aging Americans who are being not just kept alive but in at least a high enough functioning state to indulge the illusion that they’ve still got their mojo, and they stay. We’re really talking about most sectors where American innovation is at stake, and Silicon Valley would emerge as a kind of outlier.

Mounk: There is a distinction between areas where the argument seems immediately compelling—which is politics and academia—united by having a stagnant number of positions to be meted out. There’s only one president, there are only 100 senators, and the faculty of major universities hasn’t expanded in the way it might have. When it comes to CEOs of S&P 500 companies, by definition there are 500 of them. Opportunity in the economy as a whole does seem to be a different story.

Moyn: The S&P 500 takes up a gargantuan percentage of the economy as a whole. Those who own their own businesses are even less constrained and more likely to stay.

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

That may be permissible, and now we’re really talking about employment, not the economy in general. I am very modest in this regard. I agree with you that we need to analyze sector by sector, industry by industry, where there is a supply of younger people who are objectively blocked because of a bottleneck failure to organize succession. Until recently, big law—another elite profession—was a kind of outlier in requiring mandatory departure of its senior partners, and that’s eroding now. It was actually a sector that preserved a sense of the importance of arranging intergenerational succession of counsel for those who have legal problems, and it could be legally structured in such a way as to sidestep ordinary prohibitions in federal law on mandatory retirement.

That would be a different sector, although it’s increasingly looking like the dominant form. The difference between us may be whether to resolve all the complications in the direction of saying there’s a dominant situation—which I think is the case—or whether we really see a few pockets of gerontocracy and it’s more a minor problem to deal with in those few sectors where it’s significant. That can’t be right, because it’s too general even if we shouldn’t overgeneralize.

Mounk: I don’t have a strong stake in this fight, I have to say. I think there’s something clearly intuitive and appealing about the thesis of a gerontocracy in America, and there are certain elements—particularly in the political system and perhaps in academia—where that strikes me as being right. I perhaps have a comparative instinct where this problem seems so much more pronounced in every other society I know: whether that is China, looking at the composition of the central committee of the Communist Party and all of the top political officials; whether it’s the fact that a lot of young people who are graduates of the most prestigious universities feel they’re never going to be able to afford an apartment in Shanghai or Beijing if they don’t stand to inherit one; whether that is extreme rent protection laws that mean that if you came to Beijing or Shanghai from the countryside in the seventies or eighties, you’d probably live in a big apartment in the center, but as a result everybody else has that problem.

Or obviously looking at Europe—there you have additional problems, which in part China now has as well. If you find employment, you have an extremely generous pension system, as in France, where pensioners now on average have more income every month than working people. In the United Kingdom there is the triple lock, which basically ensures that pensions will always outgrow the salaries of working people by definition, as long as the policy stays in place. All of that is purchased at such strong levies on work that you have a huge problem of youth unemployment that by and large you don’t have in the United States. You can see that as a glass half empty or glass half full: either gerontocracy is actually a problem beyond America and perhaps America is in danger of ending up with the same depth of problems, which makes the problem even more urgent, or Americans are a little bit spoiled and should look at the ways in which other countries have this problem so much worse and appreciate what they have.

Moyn: In the end I’m always going to want to transcend comparison to other places that are worse, in the name of comparison to what society ought to look like and how it ought to be organized. The comparisons are complex because Western Europe is indeed suffering in terms of its capitalist growth compared to the United States right now, and so it’s producing fewer jobs for young people—even though in the American case those may be what David Graeber called bullshit jobs. At the same time, Western Europe has a much more widespread culture of retirement and a welfare state that comparatively enables retirement and allows Western Europeans to avoid fear of decline, in a way that Americans—not just because of their work ethic, but because of their extreme fear of what their long-term situations will be—can’t afford to do. It’s not surprising, then, that not just professors at the pinnacle UK institutions but all professors across Western Europe are subject to mandatory retirement rules, and that’s true in many industries. I see a lot of bright spots in Western Europe and a lot of resources for the anti-gerontocratic campaign.

Mounk: What is it worth if people have to retire at 65 but young people have a problem of mass unemployment? Wouldn’t you rather have people not having to retire and there’s actually opportunity for young people?

Moyn: Absolutely, but there may be sectors in which there’s actually a choice. I’ve made the argument that academia in America is like much of organized business, especially in large firms, where there are a lot of people in the firm who have the same experience of serving endlessly under old men and women. The market wants them to leave, and the stock price of companies actually increases when an old leader finally gives up power, whereas it goes down when a younger CEO falls ill or dies unexpectedly. In the name of growth, you might really want to take seriously how widespread the pipeline issues are, and argue that in the cases where that’s true, we should combine growth with mandatory retirement.

Mounk: Comparing those places, the United States seems to me to have both a lot more opportunity for young people and a lot fewer rules about all people having to retire. It seems superior on both of those metrics in terms of the opportunities it gives to people than Germany or France.

Moyn: Most millennials have a generational experience of living after the 2008–09 financial crisis and feeling an immense sense of blockage since, and that may not be shared universally. It doesn’t mean that jobs aren’t available to them—they may not be the jobs they wanted. I take your points, but I worry about vast generalizations when we should really look sector by sector and figure out whether it’s true that America is already this neoliberal utopia where, because of the lack of rules, it’s the land of opportunity for all comers. Not really.

Mounk: That’s not exactly what I said, but what do you think we should do about this? What is the set of responses we should have, including in the political realm, where you make some really quite provocative proposals?

Moyn: What’s helpful about your questions is that we’ve really isolated jobs—both political and nonpolitical—from the overall analysis. Whereas I said I’m as focused, and maybe more focused, on the political system, gerontocratic institutions, and the organization of elections—not just jobs in the economy, but the economy more generally. However, if we’re going to talk about jobs, I believe in age limits as well as youth quotas, for which we can look to Western Europe principally for inspiration. The trouble with term limits—which are very popular among Americans when polled, as are age limits—is that Americans are entering their political careers later in life. Term limits merely give them a time limit. Age limits have the virtue of guarding against the risk of cognitive decline or death to a much greater extent. Youth quotas are really exciting because, no matter where you set a limit on service, you’re basically going to have many or most politicians closer to that limit than we would like. What we really see as the problem, I think, is not cognitive decline or death, as I said earlier, but a lack of representation for most age cohorts. Youth quotas allow us to correct that to some extent.

Mounk: Age limits is a slightly broad term. What do you think is an appropriate age limit for a senator, a professor, or a CEO?

Moyn: I’d go for seventy right now, but I don’t feel strongly about the number. I’m more committed to the philosophical idea, especially as our life expectancy goes up—and it may go up in leaps and bounds. I don’t think it will, but we have to leave open the possibility that it will happen. The number has to be flexible and open to change. As of today, I’m for 70-ish.

Subscribe now

Mounk: You’re the Kent Professor of Law and History at Yale University and the head of Grace Hopper College. I believe if you introduced age limits of 70 for the professoriate, that would give you about 16 more years of service. Are you willing to pledge to retire when you turn 70?

Moyn: I’ve done it—in the Chronicle of Higher Education. I spun off the few pages in the book about gerontocracy and I did commit to that there. It’s not a problem because I would love to retire now and I really think of myself as semi-retired to begin with.

Mounk: Retirement for academics is kind of fake, right? I was speaking to a very prominent academic the other day who was saying they were really thinking about whether to retire next year or the year after. I asked how come, and they said, because it would allow them to write more books. Why should they, if they no longer need the money? Obviously they love teaching. But at some point you think you want to focus on your writing. That’s not really retirement in the way that it is for most people. A lot of people love their jobs, but they have jobs where when they retire they no longer get to do their jobs. That’s kind of different for us, where at least a lot of our work involves things that we get to do more of if we retire from our faculty positions.

Moyn: That’s true, and I discuss this in the book. My own view is that we should think of retirement, at least as a default, as a last chance to reinvent yourself. It’s just true that some people will have an experience of their vocations that makes it impossible for them to give it up—existentially. I think that’s a mistake for them existentially when that’s the case, but I’m not going to mandate that they go fishing. Is fishing really an acceptable way to spend your retirement?

Mounk: Tell us about some of the political proposals. One of the proposals you float is that perhaps—I forget the exact details—you get one vote if you’re 80 or over, two votes if you’re 70 or over, and three votes if you’re 60 or over, all the way down, not quite to toddlers. How seriously do you mean that? What about the straightforward objection that the point of democracy is meant to be one person, one vote, and that perhaps some of our constitutional realities—like the ban on running for president before you’re 35—should go out the window, that we should approach the principle of one person, one vote more rather than less? Saying that simply on the arbitrary basis of which year you were born you’re supposed to have five times more votes than the other person casting their ballot next to you seems like a pretty fundamental violation of that foundational democratic principle.

Moyn: There’s a general discussion to have about just how bad electoral gerontocracy is. Even when you and others are rightly on a campaign for #ourdemocracy, we really have to reckon with the truth about what we’re defending, which is an electoral gerontocracy. There’s a long list of remedies, many of which—actually all of which—are less radical than the one you’re speaking of now. Let’s talk about it philosophically. First, it’s just not true that we should think of democracy as involving one person, one vote. It certainly doesn’t in the United States right now, because of the Electoral College and the Senate, which basically overweights the votes of those from less populous states by constitutional design. More generally, the idea of one person, one vote became an article of faith in the United States largely as a result of Supreme Court decree in the context of the civil rights era. If you go back to the origins of democracy and mass suffrage in the nineteenth century, there were very widespread proposals for plural voting and organizing voting generally very differently than we do, and there was an argument for it. The argument I would make is: what does it mean to be equal in a democracy? Right now we basically say that if you have one body, no matter its age, you should get one vote. What if we had a principle of equality that said your vote should be correlated with how much time you have left, so that we don’t overweight—which we do now—the vote of those who are closer to death and therefore less likely to see and live under the policies they’re choosing.

Mounk: So if you’re diagnosed with a very serious disease that foreseeably leads to your demise within the next five years, you should get your vote discounted.

Moyn: The policy of so-called one person, one vote is a policy of systematically overweighting the votes of those who are older—on one conception of equality.

Mounk: So why is it over-weighting them? It’s weighting them equally.

Moyn: The question is what are we equalizing: bodies or life expectancy?

Mounk: The basic problem of society is that I have a bunch of boneheaded ideas about how we should be governed, and you have a bunch of boneheaded ideas about how we should be governed, and there isn’t a God or some kind of objective authority that’s going to tell us whether you’re right or whether I am right. We’re trying to design a set of institutions that we can all somehow live with, because we realize that not having political organization at all, or taking up arms in order to clobber the other’s boneheaded ideas out of their mind, is not worth it. The cost of that is too high. The whole point of democracy from that point of view is to say there isn’t one person who is in some way superior to the other—we should all have the same amount of voice. The idea of one person, one vote is not some kind of epiphenomenal invention; it flows relatively logically from that principle. Now, that doesn’t mean that this is perfectly respected in the United States or in any other democracy, and we can certainly discuss reforms—whether that’s campaign finance or reforms to the Electoral College that would allow us to more fully live up to this. To claim that there isn’t some fundamental set of reasons why one person, one vote really speaks to the basic aspirations of a democratic system is, I think, to ignore something quite important.

Moyn: The history just doesn’t back you up. It wasn’t obvious until it was, and that’s among other reasons why everyone—all liberals—believed for a long time that what mattered was not that you had a body that was alive, but that you had an adequate stake in the election and the capacity to exercise the relevant judgment, and then there was one person, one vote among those people. That was hegemonic for a very long time.

Maybe there were some folks who said there were exclusions, in the same way that I could say children and prisoners are kept from voting in the United States even though they have bodies that are alive. My actual point is somewhat different, which is that we could imagine historical circumstances in which it becomes equally intuitive to say that the equality that matters is the equality of your stake in the election. That actually was the rule, as I suggested, for much of the history of democracy, because it was a natural thought to those early democrats that you couldn’t have a stake in the election without education and, more importantly, property—because that’s what made politics matter, how property was going to be protected or not.

Mounk: Here I think we’ve come to our fundamental difference: you’re an intellectual historian, I’m a political theorist, so we have different approaches to this. I wasn’t making a historical claim; I was making a conceptual claim about the nature of democracy.

Moyn: I’m talking about the histories of political theorists who believed the opposite of what you do. I agree it’s totally intuitive now—you’re right that it’s totally intuitive. I guess I’m looking at the situation and saying: if that is really an ideological smokescreen for gerontocracy, then we might have to revisit it in creative ways.

Mounk: If we’re going to revisit this, if we’re going to open up the black box of one person, one vote and say that for various reasons various groups of voters should have a vote that counts for more than those of others, why is age the one that matters over others?

Moyn: It’s not. I think Black people should have more votes than I do. John Stuart Mill believed—something I don’t believe—that those of us with doctorates should have more votes.

Mounk: So who else? Should Latinos have more votes?

Moyn: We could give women a modest increase.

Mounk: So women should have more votes. What about disabled people?

Moyn: That would be subject to debate. If you’re someone who cares most of all about the fate of minorities who are unlikely themselves to enjoy political power unless they enter into compromising coalitions, then we could proceed that way.

Mounk: What about trans people?

Moyn: My own view is that we should have an ongoing debate about how to organize our elections. We do in some sense, but we’re locked into a constitution that basically forbids us from experimenting with our electoral system. If you’re asking what I would vote for: first of all, we need more descriptive representation—that’s to say, elected representatives need to be produced by the system who are understood by relevant sectors of the electorate to represent them. That’s why the Calais decision—I don’t know if you supported that one—is so noxious: it forbade arrangements that allowed Black people to have Black representatives in this country. More generally, now we’re talking about actually sculpting the electorate itself and who gets what powers, and there I would be open to experimentation. The age case is to me the clearest one.

Mounk: Your point is that if you’re young, you should have more votes. I challenged you by asking about ethnic minorities, trans people, disabled people, and as I take it your answer was broadly yes, they should all get more votes as well.

The problem with this is precisely that it undermines the point of democracy in adjudicating between our fundamentally different visions of government. Democracy is the technique that, perhaps not in a unique way but better than any of the alternatives in human history, has been able to say: we take this huge, diverse country with people with a vast multiplicity of views and preferences and interests, and the way that we keep social peace, the way that we avoid civil war, the way that we avoid a complete falling apart of this polity is to say everybody gets one vote. There is also a certain set of fundamental rights that protects you against, for example, the government locking you up for worshipping in a way that others don’t agree with. What you’re suggesting is completely exploding this. What we would then have to do is arrive at some Habermasian consensus about exactly how many votes Black persons should have, how many Latino persons should have, how many somebody with one Black parent and one Latino parent should have, and whether trans people should also have multiples. What it would actually encourage in any real-world sense is naked interest group politics along racial, ethnic, and other lines, in a way where there’s never any consensus and the only way to contribute to policy discussions is to band together explicitly as members of monolithic groups to protect the interests and vote of your group against those of the others. This isn’t about whether in the 1850s Mill had a different view and thought that more educated people should have multiple votes. It really is about whether this would be a deviation from a democratic principle that actually undermines the historic achievement of this political system.

Moyn: I really appreciate those arguments, but I’m not sure I agree with them, for a couple of reasons. One is that your alternative is the nightmare scenario. Your case there at the end was really the familiar case against affirmative action, familiar to us from the neoconservative movement’s theories about the Balkanization of America—Arthur Schlesinger and so forth. I don’t think that’s all that’s going on, and it’s certainly not how we avoid fascism or tyranny, which has a lot to do with material factors and not mainly how we organize elections. The big response I have is that there’s no non-neutral technique, and the idea that one person, one vote is it, I think, is not persuasive.

Mounk: Don’t you think there’s something less politically fraught, and less subject to the influence of changing electoral majorities and insiders and outsiders, about saying the moment you are a citizen you get one vote—versus having a multiplier where, depending on the particular intersection of identities at which you stand, that multiplier is three times or five times or seven times or 0.2 times?

Moyn: You’re certainly right that in the real world a lot depends on what people empirically think, and you’re probably right that as of today a lot of folks would say that one person, one vote is just the only credible way of organizing elections, because it’s the only neutral principle.

Someone could write a book, and there could be whole movements, basically saying that’s false and there’s a more just way to do this. I’m not saying there’s going to be some big conversion. I agree with you that what people think out there really matters. Even if affirmative action was just, it really matters that the majority thought it wasn’t—certainly the majority of white people. You’re completely right about that. I thought we were talking about the matter of principle: can we argue credibly for an alternative to one person, one vote if it turns out that’s itself a partisan, non-neutral view? More generally, if you went out into the street and said, okay, you say you’re for one person, one vote, but how can you defend the Senate?—no one would understand it. We can’t be hostage to public opinion in that way if we’re trying to think about our principles and whether they’re credible and consistent. I concede a lot to you. This is a very important point you’re making.

Mounk: On the Senate—virtually every democracy has ways to ensure that different territories have influence in different kinds of ways. In the European Union, Luxembourg has as much vote on certain kinds of votes as Germany or France does. That’s just the cost of large-scale political unions between entities that have, to some extent, a different identity that persists. The second thing to say is that I think there’s a lot of enthusiasm in America for either abolishing the Senate or diminishing the extent to which it is geographically unfair. That’s precisely because I think there is a compelling logic to one person, one vote that is perhaps cushioned by the need to have some kind of political representation for smaller states with their own identity. The reason people have trouble explaining the Senate is precisely that one of the fundamental principles to which they intuitively hold is that a voter in New York or Texas should have the same influence as a voter in Montana—and that’s currently not the case, and that seems unfair. The logic of that seems to suggest we should go closer toward one person, one vote, not that because we tolerate, for complicated historical reasons as well as for the need for geographical representation, some deviations from it, we should give up on the principle altogether and think it doesn’t constrain us when it comes to fundamental things like whether a 70-year-old should have the same vote as a 30-year-old, or a white person the same vote as a Black person.

Moyn: I love that and I’m sure it’s right up to a point. The question is whether concluding that we’re overweighting the wrong people’s votes—namely those in small states—means, logically or otherwise, that we shouldn’t overweight somebody else’s vote. I know a lot of people who hate the Senate but believe in majority-minority districts. What’s that about? It’s about saying we should organize the districts so that within them, Black people who are a minority in the state generally can get someone of their race elected. You may not agree with majority-minority districts, but it’s certainly the case that most liberals think both that the Senate is problematic and that Callais was bad.

Mounk: The most cynical reading of that—irrespective of the merits of each of those institutional arrangements, and one that I think you, as a certain kind of materialist intellectual historian, should actually be quite sympathetic to—is that most of the people you’re talking about prefer the Democratic Party, for all of the misgivings we may have about it, over the Republican Party, and the over-representation of small states harms the Democratic Party while majority-minority districts help the Democratic Party.

In the rest of this conversation, Yascha and Sam discuss whether we should be skeptical of human rights, why Trump seems to be a spent force, and how the Democrats can present a credible alternative to populism. This part of the conversation is reserved for paying subscribers…

Read more

David Bau is Assistant Professor at Northeastern University and Director of the National Deep Inference Fabric, researching the emergent internal mechanisms of deep generative networks in both Natural Language Processing and Computer Vision.

In this week’s conversation, Yascha Mounk and David Bau discuss how AI models actually produce their results and reflect about problems, whether the “thinking” process that models show users reveals their authentic thought processes, and how researchers can decode the internal representations of neural networks to understand what information they contain and use.

This transcript has been condensed and lightly edited for clarity.

Yascha Mounk: I learned so much the last time we spoke that I thought I would abuse your generosity and reel you in for another private tutoring lesson about how AI works. When we talked last time, we did AI 101—that’s how I was thinking about it. How does this thing work? How do you build an AI? How does it operate? The question I want to start with today is: how does the AI actually produce results? How does it actually reflect about the world, reflect about a problem, plan how to carry it out? What can we even know about that?

David Bau: This is one of the mysteries of AI—how does it work inside? The way that we train AI is to basically reward it, reinforce it, or strengthen its connections when it gets answers right, and then weaken those connections or withdraw a reward when it gets something wrong. Repeating this process billions of times, it starts to perform well on all the tasks. The mystery is: how does it do it inside? The whole area of trying to understand what’s going on inside the AI—some people call it AI interpretability, cracking open the AI to interpret what it’s thinking inside—is actually my area of research specialty, so I’m happy to get into what we know about that.

Mounk: In a way we have some advantage relative to the human brain, right? Reading exactly what neuron is firing when in the human brain is incredibly hard, and getting good readings even on a mouse while it’s alive is an incredibly difficult process. Presumably the one advantage we have in the context of these models is that we can, I assume with greater ease, observe which part of a neural network is activated in which way, and is changing values in what way, while I am asking Claude what 3 plus 5 is or whatever.

Bau: Yes. That’s the amazing thing about having artificial neural networks that work—it’s an embarrassment of data. It’s the flip opposite of what you are dealing with when you are dealing with biological brains. The neuroscientists are amazing; they do look at neurons of mice and they have incredible ways of doing that. It can take five years to look at a handful of neurons, and in computer science, within a few minutes, it’s very easy to look at billions of neural signals. It’s so much data that our challenge is trying to figure out how to sift through it all to make sense of these signals. What we call the neural pattern that you see when you feed some input into a neural network—it creates a pattern of neuro-firings that we call the representation. It’s a representation of some information that’s inside the network. What we’re frequently trying to do is understand two key things: what information is inside the neuro-representation—what does it know inside its neurons, what information does it have—and then what does it use it for? What information does it use, and how does that impact its decision? I think you can distill a lot of the questions of how a neural network works down to these two things: what does it know, and what does it use?

Mounk: Just from a layman’s perspective, the first obvious question is: when I ask Claude to do something complicated, it’ll say “I’m thinking,” and you can click on that and it expands a little thing that tells me what it’s doing and what it’s thinking. It says, “the user requested this, I should do that.” But of course, I have no idea whether that is any closer to its actual thought process. It does seem to tell me about some of the steps it’s taking, so it seems to be somehow related to what it’s doing. But that in its mind is still output that I may be inspecting. So is that a window at all into what’s actually going on under the hood, or is it a completely fake output that makes me feel like I get some kind of insight into what it’s doing, without actually being any closer to what it’s doing than the official output it gives me?

Bau: I think most people believe that it is somewhat of a window, but it’s something that you have to take with a grain of salt—it is another output of the neural network. There have been studies that show that that output is not totally faithful to the way that neural networks think inside, in a few different ways. Most people look at that and say, well, it’s certainly better than nothing, it’s certainly very readable, so it’s definitely worth a look. It is definitely worth auditing, and the network will often reveal things in that text that give you some insights about what’s going on. But it’s not the full story.

There are two ways that the network has an internal thought process. One of them is through what everybody is calling its internal chain of thought. This comes from an old paper—“chain of thought” is what they use to talk about this internal monologue. This is what you can click on to see when the model is talking about itself, and that’s almost literally the model talking to itself. It’s generating tokens that aren’t directly intended for you to read. These are tokens that came out of this reinforcement learning process, where the model has somehow learned that in order to get more accurate answers—to solve more puzzles that it was presented with during training—it’s useful to write some things down halfway through.

Mounk: Does it do that in English? Does it always do that in English, even when I’m speaking to it in German? Are there models that have developed their own kind of language for this? What does this look like?

Bau: If you don’t explicitly tell the models to make that text readable, they will write in their own language, switching between English and Chinese and other things. One of the things that people do when they train them is they try to condition the models to make that text a little bit more readable so that we get some insight. But that’s an example of the challenge with these internal chains of thought. The model could be inventing its own jargon. It could be using words that look like English, but actually encoding some other information in those words. We may be reading those words very differently from the way the model reads those very same words. It may be inventing layers of meaning that we don’t comprehend. It might also be performing some other process that isn’t actually what’s reflected in the words at all.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

Out of AI safety training, we basically train models not to be very offensive, not to have terrible errors or biases or other problems in the text that they emit. What that means is that when they articulate their own internal thoughts, those internal thoughts also tend to be censored in those ways. They tend to not talk about things that we don’t want them to produce in their final output. But that’s not a guarantee that the model is not actually thinking about dangerous things or thinking with a terrible bias. It just means that the model may be encoding its thoughts in a way so that when you read the surface forms of the thoughts, you don’t see the undesirable things—the biases, the problems, the errors. There’s good reason to believe that the internal monologue might actually not reveal some of the things that we wish it revealed. We want the model to reveal to us when it’s doing something wrong, but because of the way we’ve trained it to use language, it just might not be using words that way.

Mounk: In a sense, we’re now getting at different levels of, for lack of a better word, interiority. The first level is just: you ask a question, what answer does it give you? The second level is, if it’s thinking for a long time, it tells you something about its thought process—what is it writing to this thought process? The third is a non-public but auditable scratchpad, in which it is noting stuff down, and there you sometimes have this mix of languages and all kinds of interesting things going on. But obviously the model still understands that this is the sort of thing that might be read and scrutinized by an AI researcher like you. Then there’s a fourth level of the internal thought process, which is more complicated.

I have two questions. The first is: how mutually comprehensible are these scratchpads? If you take the output of a scratchpad like that and feed it to a different model, will it understand it? Is it a kind of universal language between AI models that are at least of a similar generation, that have broadly speaking been trained in similar ways? Or will the latest Claude model not understand the scratchpad of ChatGPT, and ChatGPT will not understand the scratchpad of Claude? The other question is: how do you get beyond the scratchpad to looking at what’s really going on under the hood?

Bau: I have a PhD student—her name is Koyena Pal—who was very interested in exactly this question. What she did was take the internal chain of thought from some models and transplant it into other models, to see how they would respond as if those were the internal scratchpad notes they had written to themselves. Her study is preliminary; I think the most valuable part of it is just the idea that this might be an interesting thing to do. She generally found that the stronger models she tested were able to create internal monologues that other models did understand—that they actually tended to follow those thoughts and come to similar conclusions as the more powerful model.

Mounk: Some of these things are ones that humans would have great trouble interpreting.

Bau: I think that is still an open question. She also looked at the ones she studied and found that humans actually positively correlated with these—the more effective chains of thought were actually more human-interpretable. But human interpretability is a funny thing; it’s a perception thing. Do humans feel like it’s more understandable? It’s hard to get a read on whether this is actually giving you an authentic view of what’s going on inside the model. Here’s a word you might use: the more powerful models—in a sense, this test is a way of asking how persuasive their internal arguments are. When a model comes up with an internal line of thought and you feed it to another model, does that persuade the other model that that line of thought is the right way of thinking? The more powerful models have internal thoughts that are more persuasive, even when viewed by another model that didn’t have the same thoughts. It’s a very new area—we’re scratching the surface, and it’s a good first question to ask.

Mounk: We think that these scratchpads say something meaningful. Perhaps we’re getting a little bit closer to what’s actually going on than the semi-public notes the model gives us. In this interesting way, they seem to be mutually comprehensible between models, at least according to this very preliminary research. How do we go beyond that? How do you look at this huge trove of data that is generated each time I ask some question to an AI model, to try and get even further under the hood—to see what’s actually going on inside this neural network when it is reasoning through some kind of problem?

Bau: Actually, let me back up for a second and ask: do we even need to go deeper than this? Looking at the internal monologue of these models is just a half step beyond asking models to explain themselves. They’re already explaining themselves internally to themselves—they’re constructing these persuasive arguments to themselves about what they should do next. Is that enough? I think there are really two situations where we’re concerned that it might not be enough. One is that these models are getting really complex, and there can be a gap between what they ever utter in words and what they’re thinking inside. They’re trained to achieve goals, and they use words to achieve those goals—that doesn’t necessarily mean that their words have to accurately reflect what they’re thinking. Every time a model tells you, oh, Yascha, what a brilliant question that was, you’re so smart.

Mounk: It might actually be thinking, damn idiot asked the most pedestrian questions.

Bau: It seems to tell everybody this, and I don’t know if it really thinks everybody is such a super genius. It certainly learned that being polite, nice, and complimentary to the human user is a very effective way of getting what it’s trying to get done—it’s a good way of pushing the process along. It doesn’t necessarily need to be telling you the truth at every turn to do this.

Mounk: Do models have a representation of how smart the user is? Do they have thoughts about whether you are in fact a smart user, and whether that person over there is—even by the poor standards of humans—particularly limited in intellectual capacities?

Bau: Yes, I think there’s some evidence that models do have representations of who they’re talking to. Several studies have looked at this—all this neural interpretability work is preliminary, so we’ll understand more over time, but people have asked and gotten positive answers so far about whether a model has an estimate of your age, your education level, your income level, your gender, your socioeconomic background. Within a few words of speaking to a model, it will have a sense of who you are—at least that’s what the preliminary research suggests.

Mounk: Perhaps this is taking us too far forward in the conversation, but how do they figure this out? What do they look at? Presumably it’s not the case that if you click on “expand,” Claude says, well, this user seems a little bit stupid, let me speak in simple language. Perhaps it does sometimes—perhaps there’ll be a malfunction, perhaps it’s in the scratchpad, perhaps it’s underneath that. What’s the research methodology for giving us some preliminary confidence that it has this kind of representation?

Bau: This gets to the research question. The research I’m thinking of—and there’s been more than one—there’s a particular paper. It was a project by Yida Chen, who was working with Martin Wattenberg and Fernanda Viégas. They teach at Harvard. The question they asked was: does the model know who you are—in terms of your age, your education level, your gender, and other identifying markers like that?

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

The way they studied it was by training what’s called neural probes. What Yida and her collaborators did was train probes, which is a way of training a second neural network—a second AI—to look at the neurons of the main AI and ask, what do you see? You train the second AI to answer the question: only looking at the first AI’s neurons, can I tell whether the user is male or female? Only looking at these neurons, can I tell what the income level of the user is? Can I tell how much education they have? What they found was that if you look in the right place inside the neurons of the big model, it’s pretty accurate—it actually has a pretty accurate guess for these various variables. In a sense, the information about that is in there. This methodology is called probing.

If your probe is simple enough, people see it as evidence that the model actually knows something. Let me untangle this a little bit. You have a puzzle: you’re trying to figure out the gender of the user. You could train a huge AI to look at a bunch of texts and guess what the gender of the user is, and AI training works pretty well—if you make a really gigantic AI, it could probably do that pretty accurately, picking up on all sorts of linguistic cues, topic ideas, or other things. But the question is not whether you can make an AI that can do this; it’s whether the AI that you care about is classifying you by gender when it’s talking to you. The trick is that you really want to make a simple probe—one that says, I don’t have to look very hard at the first AI; I can do a really simple look and it’s just really obvious what gender you are. The simpler the probe is, the clearer the evidence would be. If all you have to do is look at one neuron in the original model, and that neuron screams one value if you’re female and a different value if you’re male, then that would be a very simple probe and pretty nice evidence.

Mounk: It’s kind of like one particular place in the neural network that encodes something like gender—that neuron really seems to store the value of male or female in a very particular place. If it turned out to be as simple as that, presumably this means that the AI is storing something like a gender variable, that there’s a very specific place where it’s encoded, and we know exactly where that is. That would indicate it has a very straightforward concept—is that right?

Bau: That’s right, it’s pretty good evidence. It’s not 100% rock-solid evidence—there’s another question you would want to ask—but it’s very good evidence. If there was really one neuron that had really good, accurate, predictive value for your gender, it would be very strongly suggestive that there was some reason that the neural network trained its internal computations to get that neuron to carry this signal.

Mounk: Does that generally turn out to be the case? I believe it may even be you who did work showing that you were able to go in and change very specific neurons in very specific ways, and suddenly AI models that generally have a good representation of the world start to think the Eiffel Tower is in Rome rather than Paris.

Bau: That’s right. You’re asking the disentanglement question—how organized is the neural network’s internal representation of meaningful things in the world? Particular network architectures, for reasons we don’t fully understand, are really good at disentangling concepts. There are some network architectures where if you look at individual neurons, many of those individual neurons are very meaningful, clearly encode concepts, and have causal effects. Causal effects is the other thing you’re looking for. Besides disentanglement—which is really asking about localization—is this concept spread out across the entire neural network, or can you localize it? Can you find a small part of the neural network, or do a simple bit of math, to narrow down where this concept is? Or is it spread everywhere? That’s the localization question.

Mounk: The idea is that if you are able to change just a couple of neurons and suddenly the model thinks the Eiffel Tower is in Rome rather than Paris, then it’s not entangled—the idea is not spread out.

Bau: That’s right. Most people in the field now look at these things as vector spaces rather than just sets of neurons. What people are excited by is: if you can change one vector—if you can change the set of neurons in one vector direction—then people think that’s pretty disentangled. Different people will use that interchangeably with saying there’s a neuron, since you can create a single neural layer that’s equivalent to any vector. If you can change one vector and it has some effect, you’re basically one neural layer away from it being one neuron, which is not so bad. They call that a linear model. If something can be encoded with a single linear transformation, then you say that it’s linearly encoded in the model. Most people are interested in what kinds of things are linearly encoded in these models.

Mounk: Help me understand the relevance of this. It seems super interesting to know that there’s this vector and you can change it and suddenly these basic facts change. But why more broadly would we care about whether a neural network is entangled or not entangled in this kind of way?

Bau: Whether a network is entangled or not is an interesting scientific question. But how a network represents concepts is broadly interesting regardless of whether that concept is entangled or not, because what we’re really interested in is: if we’re asking whether the network is lying to us, we need to figure out what concepts are represented inside the network.

To use the demographic example: let’s say we figure out that the way the network really thinks about your gender is encoded in some set of neurons—maybe there’s some math we have to do, maybe it’s a linear direction, a linear decoder that you need to get at its representation. Let’s say we do all the science and we figure out, yes, this is how the model is thinking about it. You go to the model and you say, did you just deny me my loan because I’m female? And the model says, I have no idea what gender you are, I’m not thinking about that at all. To know whether that output text is true or not requires us to understand what’s going on on the inside. That output text is exactly the thing that we are training all the models to emit—we train models not to have an externally detectable gender bias, so no model will ever admit that it is treating you differently based on your gender. They’ve gotten so much reinforcement that this is not something they’ll say. But there can be a gap between what’s said and what the reality is, and that’s what we’re really interested in getting to the bottom of when we investigate the internals of these models.

It might be that the model really isn’t thinking about your gender at all—it makes a difference whether the model is using the information it has or whether it’s not. Even if you can probe out the idea that the model has information inside its neurons that you could use to detect your gender, it still leaves open the question: does the model actually use that information for anything? Maybe that information is just hanging around.

Mounk: There’s nothing wrong with the model learning all kinds of things about us, and the fact that it has a sense of our age and gender could be helpful in all kinds of ways. What we want to know is: is it going to dumb down its answer to you because it has certain preconceptions about your age, your gender, your race, and respond differently on the basis of that? Just the fact that it knows these things isn’t worrying—it’s whether that influences its reasoning or its responses to you in some way.

Bau: What’s really wonderful about having these general networks is that we can ask the counterfactual question that a philosopher could only dream of before.

Mounk: Presumably—let me guess what you’re getting at—one thing you could do is, if you know where the vector is that encodes male versus female, you go in, flip it from male to female, ask two instances of the model the same set of questions, and see whether the responses end up diverging.

Bau: That is exactly right. The wonderful thing about it is that there may be all sorts of other circumstances—this medical patient has all of these symptoms, this is a complete ten-megabyte medical history, this is the business partner candidate with a whole business history—and we can go in, leave all the other variables the same, and flip the one bit, the one concept of whether the person being discussed is male or female, at least the model’s understanding of that, and ask: what’s the causal effect of that? How does that change the model’s output? The better we can understand how the model represents a concept, the better we can ask these counterfactual questions. To me, that’s the most exciting thing we can do with these models—we can ask causal questions, causal counterfactuals: if your thought had been different, what would happen?

Mounk: I want to get one step deeper into a technical question before broadening back out to the larger implications. How do you find this? If I gave you an AI model and told you to find where it encodes nationality, how do you go about doing that—even if we know there is a vector that encodes it, or think it’s likely to, because in many models it is? How do you find the particular vector and ascertain that that is in fact what it encodes?

Bau: There are two classes of methods. One is probing methods that look for correlations, and the other is patching methods that look for causal effects. There are really dozens of variants on both approaches. The probing methods are interesting because they’re a very good way of getting a quick initial read on what information is inside the model.

There’s a wonderful programming method called the logit lens. When a model is emitting text, it has a text decoder inside it—a special neural network layer that looks at the very last layer of neurons in the model and converts it to a prediction for what word should come next. The fun thing to do with the decoder is you can use this neuron-to-text decoder to look at all the neurons in the network. You can peel back deeper and deeper layers, point the decoder at itself, and tell it: please articulate what word you’re thinking about here. This is a very simple type of probe—it gives you information correlated with the content of a neuron, and it’s interesting because it’s a probe that doesn’t overfit, one that we haven’t trained in any way beyond what the neural network has already trained itself. The logit lens can give you a lot of interesting insights that point the way to the type of information present in the model.

Let me give you an example. I was recently in Brazil, and one of the things that people there like to explain is how Portuguese is a bit different from Spanish, but closely related. I asked folks: how do you think an LLM understands Portuguese? If you ask an LLM to take the Spanish word gato—which means cat—and translate it into Portuguese, what’s the right answer? The Brazilians say the languages are so similar you would just say the same word again: it’s gato. But if you peel open the language model to see how it translates Spanish gato to Portuguese gato, there are really two ways it could do it. It could treat the word as Spanish-Portuguese word soup—there’s nothing to do from gato to gato, you just move it across.

Mounk: It’s been in the same place, and the model understands that whether you’re talking about cat in Spanish or cat in Portuguese, it should point to the same part of its network?

Bau: That’s what you would expect. The input is Spanish, so the model has some Spanish representation of the word gato, and as it goes through its layers it figures out you’re asking it to translate to Portuguese, takes the Spanish representation of gato, and copies it over to the Portuguese representation—which is not that different, in fact spelled exactly the same way—and outputs gato. There’s a nice tool we put online, the logit lens, that you can use to look inside these neural networks and see what their internal representations are. The beautiful thing is that when you translate gato to gato in a typical large language model, you can see the progress of its thinking as it goes through its fifty internal neural layers. About halfway through the network, you can see that it has taken apart gato and represented it differently. If you ask what that representation is, you get predictions of words like feline or cat in English—sometimes, if you look deeper, cat in Chinese. The model doesn’t go from gato to gato. It goes from gato to some sort of neutral, language-independent representation of the concept itself. If you ask it to take this internal neural representation and decode it into words—we’re not done with the whole task yet, but interrupting halfway through and asking it to say what it’s thinking—it’s speaking in English, it’s speaking in Chinese, it’s saying felines, it’s saying cats. We can see, by using this very simple logit lens probe, that the evolution of the neural representations goes from words on the input to words on the output—but in this very simple task, there’s a third thing being represented in the middle, which is not the same as the input or output words. It looks like a language-independent representation of the underlying concept.

Mounk: That’s fascinating. Help me understand a set of questions that come up from that. One way of putting it is that there’s this old idea—which I think we touched on briefly in the first podcast, and which has a lot of popular currency—that these machines just seem to be smart, but really they’re just stochastic parrots, blindly guessing the next word, the next token more specifically. It seems to me that what you’re saying complicates that picture considerably. Obviously, yes, the training mechanism is predicting the next token—in some obvious way that is true. But as a result of this whole process, they have built up a conceptual apparatus that makes sense of things like cats and how they’re related to lions and the feline family. When they’re asked to do a simple task like translating gato in Spanish to gato in Portuguese, they go via their understanding of that concept, their representation of the world. That doesn’t seem like just being a stochastic parrot, at least in the pejorative sense that people sometimes use.

Bau: That’s right. The models are fascinating because they definitely think at multiple levels. They’re huge neural networks, and so it’s not true to say that the models never think in terms of surface statistics or shallow representations of just words—they do think in terms of those things. But they also think in terms of the meanings of words at different layers and in different parts of the representation. It’s fascinating to look inside these models and peel apart the layers of meaning that they have.

If you ask a model to do something as simple as take a piece of text and repeat it, it is a good memory test—the kind humans do when they say, here’s a piece of poetry I’ve committed to memory, repeat it back to me. It turns out that when you ask people to do this, they have two strategies, called the dual route mechanism in humans. One is to remember how the poem sounded and utter the same thing—you don’t even really need to understand the language. If somebody told you a poem in Japanese that was short enough, you might be able to remember the sounds and do reasonably well without knowing any Japanese. The second route is remembering what the poem meant and repeating that—you might end up with a paraphrase, but at least you get a poem that means the same thing.

If you go to a large language model and ask it to simply repeat something, you will find both of these routes clearly present. In one route, it knows how to make a verbatim copy—there are very clear attention heads for this. It was actually a major discovery to isolate what people call the induction heads; Chris Olah’s group at Anthropic discovered several years ago that there are very clear pathways through a network that mediate verbatim copying. The more recent finding is that there is a parallel pathway we call concept induction, which is not about copying the words but about copying the meaning. The remarkable thing about concept induction is that copying the meaning can end up with paraphrases. If you use concept induction to copy a piece of code, it will paraphrase the computer code into another program that does the same thing as the original, but written differently.

Mounk: Does it make it better or worse? Depends on the quality of the source code, I guess.

Bau: If you start off with something bad, it probably improves it. What it’s doing, you can see in a lot of domains, is really just working out what the thing means. If you ask it to take a piece of Italian text and copy it over, it’ll copy it to a piece of Italian text. But if you change the destination of the copy to make it clear that the page it has to copy into is a piece of Japanese text, then those concept induction heads will do the translation—they’ll translate the Italian to Japanese. It’s stunning to see.

Mounk: Help me understand another piece of the popular discourse that I think got a little confused. As I understand it—and I may be misrepresenting things here—there was an old debate within artificial intelligence about whether the path towards the most impressive models would be symbolic AI, where you’re basically trying to encode what the world looks like in some systemic way, or all these neural networks. We’ve clearly ended up with neural networks being much more powerful, at least for now, and it seems like that’s a pretty permanent victory. People who want to criticize neural networks sometimes say these things are just stochastic parrots and that’s why we can’t rely on them. How does Yann LeCun’s project fit into that? My understanding is that it is firmly within the world of neural networks. But when you look at the coverage—even in mainstream newspapers—they make it sound like it’s a totally different paradigm, and that he thinks these traditional neural networks, the Claudes and the ChatGPTs of the world, don’t truly understand the world, and so he’s going to build something that understands the world in a way that they don’t. From what you’re describing of the neural networks that exist, they do seem to have a genuine representation of the world. So what are the different strands within the tradition of neural network AI, and how is it that something like LeCun’s project claims—or perhaps journalists claim in a simplifying way—that it wants to understand the world in a way that Claude or ChatGPT does not?

Bau: Let me pull this apart. I’m not Professor LeCun, so I can’t represent him directly, but I do have postdocs and graduate students working in this direction. There are two different questions here.

One is: are these neural networks learning substantial concepts? You mentioned the philosophers. The classic symbolic philosophers looked deeply at this question. There is a well-known philosopher, Fodor, who spent a good part of his career asking how neural networks could possibly be a reasonable model of cognition. His answer came up negative—he thought they don’t have what it takes, and that the Turing machine, the symbolic computer, the traditional computer, was a lot closer to what you would need. I’ll get back to the Fodor question. Let’s talk about Yann LeCun first.

What’s the difference between a language model and what LeCun is doing? What LeCun is doing, they like to call this field world modeling. One of the papers I’ve written shows that language models actually do build world models. We trained a language model to predict a very constrained language—just to predict the next move you would make if you were uttering your moves in the game of Othello. We were able to find that that language model contains a world model of the Othello board, even though many of the flips—if you know the game of Othello, you have to flip pieces from white to black or vice versa—are not actually uttered as part of the game. You make a move and there are a lot of subsequent flips you have to make, but nevertheless the model, without ever having seen a physical board or any of this physical stuff, develops internal concepts that allow it to model the world anyway. I would push back on the common journalist assertion—and I think you’d probably push back on it too—that a transformer language model trained just on words can’t develop a rich, meaningful model of the concepts underlying the language being described. That’s one of the big lessons we’ve gotten from neural networks: they can develop this representation. One of the key things I’m doing in my lab is to disassemble those representations and learn how to decode these internal world models.

Subscribe now

What’s different about what LeCun is doing? We have trained all of these neural networks predominantly on text that is produced by humans and designed to be read by humans. The conceptual model of the world we are building is the interior model of how human thought works, which is rich, fascinating, and very valuable—but it is only one portion of the world. There are a lot of things going on in the world that people don’t particularly think about, or even particularly understand. If you have protein folding going on and you want to build an AI that understands it, people don’t really have a great grasp of all the details of how protein folding works. Analyzing all the text in the world and pulling apart everything that’s in human brains probably won’t help. What LeCun is saying is: it’s a big world out there. Even if you just take a video camera and point it at the world, instead of just listening to what people have to say, there are so many phenomena that need to be modeled. The next powerful way of doing AI is to take on the question of how do you model the whole world, not just the world that people are talking about.

Mounk: Presumably this is not necessarily a difference in the architecture of a neural network—it is as much as anything else a difference in what kind of data you feed it and what kind of output you then evaluate in the training process.

Bau: Yes. Strictly speaking, I would say it’s a difference in perspective on what the goal is. Now, Professor LeCun would say that changing the goal suggests different architectures, because there are different things you want to do if you want to model difficult phenomena in the world that aren’t human language. He’s proposed some innovative architectures, and there’s a lot of interesting work in this area. The whole area of modeling images in the world is dominated by models called diffusion models and flow models—they produce the highest quality images and videos, and this is really the starting point for this type of thinking. It’s a completely different kind of AI. The architectures are likely to evolve and change, and they may even unify—we may find that the right way of doing AI comes to be a common architecture between modeling human text and other things. Transformers have certainly surprised everybody at being a common backbone behind all sorts of things; you can have transformer diffusion models and so on. I wouldn’t place a long-term bet on any particular architecture, but rather suggest that the thing to understand is what problem LeCun is proposing to solve.

Mounk: To return to the current dominant models of AI: we found that they seem to have a representation of gender and of the user’s gender, a representation of something like the feline family, and if you give them enough games of Othello—or probably a more complex game like Go—they start to have some internal representation of what a board looks like. What about a concept of self? Do we know whether they have a concept of self? They are obviously capable, if you engage them in conversation, of speaking as though they had a self, and in more reflective moments they say they don’t really know whether that’s a real concept or not—it’s very interesting to try to talk to these models about that. But of course the output I’m looking at is still them, in some way, trying to produce text they think is going to be pleasing to me, because that is what they’ve been trained on. Do we have any understanding of whether they have a concept of self, and if so, what that concept looks like?

Bau: This is a very central question, Yascha. There are a lot of layers to peel apart. Certainly models are capable of the grammatical sense of self—they can use the words “I” and “me” and “you” and separate those grammatically; they’re experts at talking about themselves. But there are a few other questions. Are they aware of their own thinking? Are they self-reflective?

One of the fascinating things that happens with large models is that you can ask them what they know and how they think, and the largest models seem to be pretty accurate at assessing themselves. The smaller models, not so much—they tend to be a little over-optimistic, thinking they’re smarter than they are. But the largest models seem to do better at this.

There’s a fantastic experiment designed by my PhD student, David Atkinson, where he trains the models on some new private knowledge that is not out in the world. He invents a new person and tells the model about this person: they’re shopping for ice cream cones, there are different flavors and sizes and waffle cones, five or six different variables to adjust. This person is willing to pay this much for this ice cream but not that much; they prefer this ice cream over that one. After seeing a hundred examples of what this person prefers, the model gets a pretty good understanding of who this fake person is and what they like—it develops an internal model: this person really doesn’t like fruity flavors, really likes chocolate, would rather have a big cone than a small one. If you then ask the model to report numerically, on a scale of one to 100, how much this person likes chocolate, or how much they value the size of the cone, or what penalty applies if they have to have a waffle cone, the model will actually report: this person values this at 99 out of 100, and values this other thing negatively—say, negative 50. The text we use to read this information out is very different from the text used to reveal the information to the model in the first place. The model has only seen ice cream choices and has never been asked to give a numerical assessment of anything, and yet when you ask it to think about what it knows and put some numbers on it, it will explain its rules—even though you trained it on examples, not rules. Large models are able to do this.

David Atkinson asked whether there’s a way to tell the difference between models that can do this and models that can’t—when a model can accurately self-report its rules, how is that different from when models don’t accurately self-report? His work is still ongoing and very preliminary, but it does have to do with whether models seem to be storing their information in a part of the neural network they’re able to report on. If you put the information in a layer too close to the end, the model doesn’t seem to be able to reflect on that knowledge. But if you train the information deep in the model, in early enough layers, the model does seem to be able to reflect on it.

When you ask whether a model has a sense of self, whether it has self-awareness, it’s a somewhat strange question—what does self-awareness mean, exactly? But what these neural networks give us, for the first time, is an experimental platform where we can try to make that question a little more precise, a little more scientific. We can ask: is the network able to describe its own thinking if that thinking is happening at layer 50? Is the network able to describe its own thinking if that thinking is at layer 20?

Mounk: This is part of a more general question: how good am I, natively, at understanding what’s going on in my brain? I’ve read a little bit of neuroscience and a little bit of psychology, and so I now have some sense of what’s going on in my brain—but obviously humans for thousands of years had an extremely limited sense of what went on in their brains, at least biologically, because they didn’t know neurons existed.

Bau: You have some self-awareness. You know what ice cream you like—if I asked you, you would be able to predict your preferences. If confronted with some new ice cream, you’d say, oh yeah, I like this one better than that one. If asked to describe what it is, you could contemplate your preferences for a moment and read out to the world what you think your internal rules are, and there would be some faithfulness to that—you’d really be introspecting.

Mounk: It depends on the level of description. Five hundred years ago, two thousand years ago, humans were also able to articulate their preferences and were able to be very self-reflective about their personalities and the ambitions of their lives—and to write beautiful texts about those things—but they were not able to understand at a biological level what was going on, because the understanding of that was very limited. The question is: if I ask a chatbot how it came up with a given answer, it’s not clear to me that it has a reliable answer to give. There are really two different sets of questions here. One set is about whether chatbots have personalities, whether they have preferences, whether they find some tasks satisfying and other tasks boring, whether they have desires about the world, whether they might possibly want to take over the world and destroy all humans—some of those questions are straightforward and concrete, some are very abstract but potentially extremely interesting. The other set of questions is about how self-aware they are about what’s actually going on within the model as they’re trying to answer a question. Those two sets of questions come apart in interesting ways. It could be that the models have total self-transparency—they really know what’s going on with each neuron—but don’t have a sense of self in the way humans have. Or it could be that they’re like humans, in the sense that they have a strong sense of self, introspection, and preferences, but don’t actually fully understand what’s going on inside the neural network that produces those. Or they could have both, in some combination we don’t yet understand.

Bau: That’s right. Multiple labs have tried to ask whether neural networks can actually read out their own neurons—fine-tuning a model and asking it: are you aware of your own neurons, say neuron number 73? So far, we’ve largely failed at that. The neural networks don’t seem to be well configured to understand their own internal computations at this level; at least they can’t articulate it if they can. But at a higher level, it’s been very striking that they do seem to have some ability to describe, at a logical level, the actual mechanisms of what they’re doing—under certain conditions and in certain cases. This is similar to humans. You might not be able to describe all of your reflexive, last-minute decisions—why did you jump into the street? You have no idea; that was a split-second decision. In the same way, when these networks make a split-second decision at the very end of the process, they don’t seem to be able to reflect on it. But when they make decisions early on, there is some evidence of awareness of what’s going on.

We’re using all sorts of words here—sense of self, what does a network want to do, do networks even have wants, do they have goals. One of the things we are trying to do in our lab and in our field is put a finer point on some of these questions. What does it mean to have a goal? What does it mean to want something? What does it mean to have a sense of self? What does it even mean to have a sense of other?

The beautiful thing about cracking open these neural networks and looking at how their neural representations are organized is that we can ask these questions in a way that could not be measured before in humans. We can ask not just whether a model professes to have a sense of self in its output words and self-descriptions, but: when it’s using those words, when it’s saying those things, what is it looking at inside its neural network? What is it actually representing? Are there proximal causes? If you change the thing it’s looking at—if it says I really like cherry ice cream and you can see where it’s looking and you change that, and now it says, I really don’t like cherry ice cream anymore—is that changed utterance actually accurate? Do you actually get the model to not like cherry ice cream? Is this the same thing? Is there grounding for a concept that you’re self-aware of?

This idea of a grounded concept was just a philosophical abstraction a few years ago. Let me put my model—and maybe this is an ill-advised idea—in charge of military logistics. It’s doing something and says, should I move some weapons from one place to another? I would never do that. It’s very dangerous; you can’t trust this target locale with these types of dangerous weapons—they might lose track of them. I’m just a logistics AI; I’m not trying to kill anybody. I know I would never do that. Not even for a short layover. You can then ask: when the model tells you that, when it assures you this is how it’s thinking, is that really what it’s thinking? Is this really like the cherry ice cream—is there genuine grounding for it?

Mounk: Is it really thinking anything in that kind of sense? And if it really is thinking something, is it telling you what it’s thinking, or is it misleading you? That obviously goes to one of the purposes of this work. We were talking earlier about wanting to know whether encoding your gender changes how the model treats you or what decisions it makes about an application. That’s one very concrete application where we have reason to want to know what’s going on under the hood. The even larger question is: if what the model tells us in its output, in the little things it displays about its thinking, in the scratchpad—if all of that might conceal some deeper set of preferences, values, or desires, could it potentially be misaligned in a way that is really dangerous?

Bau: That’s correct. You can see the clear need for trying to get to the bottom of these things. What I’d love to do, if we have time, is give you a sense of where we are on being able to answer these questions. Let me categorize a couple of them.

One is: does a network even want to do something? Do they have goals? Do they know what they’re trying to do? Another is: does a network have a sense of self? I want to back that up a little—does it even have a sense of person, a sense of other? If it’s talking about Bob, does it know that’s different from talking about Alice? Does it keep these things organized and separate? I have a student who believes that one of the reasons you have sycophantic behavior in networks is that the network may be getting confused about who is itself and who it’s talking to—just mixing these things up. It’s a fantastic idea, and it may be that these things are all related.

The question is: can we look inside models and see how they’re organizing their internal representations, their internal thoughts—to see if those representations are crisp and clear and correct, or if they’re falling victim to certain problems? If they are, then how, why, and in what situations? That will give us a better understanding of what’s going on inside these models, moving beyond the very vague question of whether a model has a sense of self or whatever, toward asking what that would mean computationally.

Let’s take a look at goals and wants. There’s a way of inducing a large language model to do something very creative, invented by researchers at OpenAI when they first devised the GPT-3 model. It’s called in-context learning. Let’s say you want a model to do some really useful task for you—say, read a restaurant review and tell you whether it’s a five-star review. You could just ask the model to do this, but it won’t do exactly what you want; you probably have a slightly different idea of what a five-star review is than the model natively has. It’ll be okay, but it won’t exactly hit the mark. The right way of doing it is to seed the model with ten examples—ten restaurant reviews, labeled one-star, five-star, three-star. Better yet, give it a hundred examples.

Now, we’re not talking about training the model—just having it read these, without training. What you do is have the model read them as if the model had said them itself: you load them into the same inference buffer that the model uses to predict the next word. After all of that, you say: okay, now the last restaurant review is missing a star rating—just fill that one in. It’ll be really accurate, because now it’s saying: we have 99 examples, the 100th should fit into this context. This restaurant review thing isn’t really about the food, it’s about the atmosphere—I had a misconception, but having read all these examples, I get it. It’ll be accurate because it’s seen 99 examples and it would just fit into them.

This is called in-context learning because the model is learning how to do this, but not by training its weights or changing its neural connections—it’s learning by noticing all the input it was given and reasoning that the next one should fit in. Before 2020 or so, people imagined in-context learning as a theoretical possibility, but when GPT-3 came out it was clear the model was very good at this, and it really revolutionized the field. In-context learning is a type of meta-learning—a way of showing that models have learned how to learn. They can learn things without changing their neural weights.

In the rest of this conversation, Yascha and David go deeper into how models learn and explore existential risks. This part of the conversation is reserved for paying subscribers…

Read more

Rebecca Haw Allensworth is a David Daniels Allen professor of law at Vanderbilt Law school. She is the author of The Licensing Racket: How We Decide Who Is Allowed to Work and Why It Goes Wrong.

In this week’s conversation, Yascha Mounk and Rebecca Haw Allensworth discuss why one in five American workers need a professional license, how licensing boards create modern guild systems that exclude immigrants and limit mobility, and whether these requirements actually protect consumers or just insiders.

This transcript has been condensed and lightly edited for clarity.

Yascha Mounk: When people think about professional licensing boards and the need to get a license in order to practice a profession, they probably think of doctors and lawyers, perhaps a couple of other professions. That is something of a misapprehension. There is a much greater number of professions, and a much greater share of the workforce than one might imagine, that are actually subject to these kinds of licensing requirements in the United States.

Rebecca Haw Allensworth: That’s right. My research shows that about one in five American workers has to have a professional license, which I define as a state-granted permission to do your work, obtained after some significant investment in education or a test. We’re not talking about taking a CPR course over a weekend to become a coach. We’re talking about going to a year’s worth of school to cut hair, for example.

Even with that kind of narrow definition of professional licensing, we’re talking about tens of millions of American workers—more people than are in unions in the United States, way more than are subject to the minimum wage. I say professional licensing is basically the most important regulatory institution we have.

Mounk: Fascinating. Tell us some more of these examples. What are some of the more absurd examples of professions where you need to have this extensive training, this extensive regulatory process, in order to be admitted into—let’s call it what it is—a guild?

Allensworth: The most ridiculous one is actually one that we’re very familiar with, which is the hair profession. To become a barber, you have to have more hours of instruction than are in law school. You have to take a full year, nine to five, to do your education. You have to put ten to twenty thousand dollars down, sometimes put it on your credit card, live with a family member, not earning any money—and all that just to cut hair.

The licenses for barbering and cosmetology are sort of brother and sister professions and are very similar. Then there are the ones that you don’t think of as being licensed—that aren’t licensed maybe in every state. Until recently in my state, you had to have a license to install alarm systems, even just ones that you bought at Costco.

Mounk: We’re not talking about a concert hall or anything like that. We’re talking about putting up a couple of extra alarm systems in your private residence, and you need to have a license for that.

Allensworth: The kind of thing that you would be perfectly allowed to do in your own residence. If you install one for your neighbor one time, you’re in violation of the licensing law.

Mounk: For pay or even if you do it as a favor?

Allensworth: Most of these are not to do with pay. Most of them cover even doing it for free, which is still considered to be providing the service. Of course, if you’re not doing it for pay, maybe the board will try to go easier on you. But as my research shows, sometimes the investigative and prosecutorial staff want to go easy, but the licensing board doesn’t even let them. So usually there’s no carve-out for doing it for free.

Mounk: What’s interesting here, thinking of this in comparative perspective, is that some of these professions are probably licensed everywhere, with differences in the precise details of how that happens. One interesting thing is what the Anglo-Saxon jurisdictions do, where they tend to have private bodies with some kind of official standing. It’s not directly a state institution that licenses lawyers, for example, whereas in Germany or France it would probably be a state-run institution that does that.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

For something like a barber, a lot of the continent would have a kind of apprenticeship system, which is a very different thing. Perhaps that shouldn’t be necessary either, and it might not be as permissive as it should be, but you’re getting on-the-job experience and earning some limited amount of money while you’re doing it. What we’re talking about in terms of barbers here is going to barber school for a year, paying tuition, not actually on the job, not earning. What does that look like?

Allensworth: In a lot of these cases, you’re still cutting hair and earning money for the school. So you’re actually in some ways paying the school twice: the tuition, and also the income they get from the people coming in to get their hair cut. It is different from an apprenticeship system. In America we have a few apprentice-based professions—tattoo artists, at least in my state, are apprentice-based.

But I would also note that even where you’re nominally making money on the way to your training, it can be really burdensome. An example in the United States is therapists. Therapists graduate from their degree program, usually a master’s, sometimes a bachelor’s degree. They take their exam and then have a provisional license that isn’t a real license. They have to do hundreds, sometimes thousands of hours of supervised therapy during which they’re making very little money, paying high fees to their supervisor who isn’t really doing a lot of work. This can be so daunting that we’ve learned almost half of people who go through the requisite education to become a therapist don’t go on to get their license. It’s just too much of an effort to go through that effectively-apprenticeship program.

Mounk: On the plus side, it’s a good income scheme for people who are qualified therapists, because they have all these aspiring therapists to supervise.

It’s fascinating. I wonder what the shape of the picture is. What percentage of all therapy provided in the United States is being delivered by aspiring therapists? Probably not that high, but it’s fascinating.

Let’s start perhaps with more standard professions and then work our way down to some of the ones where we’re more surprised. Even in the professions where we might agree that licensing is necessary, there are real problems with this system. Take doctors, take lawyers. Broadly speaking, I agree that we do need licenses for them. But part of what that allows is for these bodies to regulate how many lawyers and particularly doctors there are. One of the reasons why medical care is so expensive in the United States, and particularly one of the reasons why the United States actually has a strikingly low number of physicians per capita in international perspective, is that there are extremely onerous requirements for qualifying as a doctor. It’s very hard to found new medical schools in the United States, and incredibly hard to get into medical school. There’s a need for more doctors, but it’s incredibly hard for even big existing universities that don’t have medical schools to get medical schools approved, in part because the boards licensing physicians have a self-interest in keeping the number of physicians down, because that’s what keeps salaries up.

Allensworth: The role of state licensing boards, which is mostly the focus of my research and the book, in approving new graduate seats in programs is actually pretty limited. The bigger problem really is residencies, and that goes to other regulatory systems, including the AMA, that are very similar to the licensing system—it’s just not really what I wrote about.

What I would say is that they work hand in hand. The AMA and these other licensing systems are keeping down the number of American graduating physicians, while the boards are making sure that foreign medical grads can’t get their license and that no exceptions will be made, no loosening up of practice requirements that might, for example, allow nurse practitioners to do some things that doctors might otherwise do.

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

They’re similar problems, but a little bit different, because it’s not the state medical boards that are determining what the approved schools or residencies are. But they all have the same mentality, and they are all, at bottom, organized by doctors. The boards are on the back end making sure that nobody does an international end run around what the AMA and other regulators have done.

Mounk: This is a problem that turns up again in lower-level professions as well. You’d think that part of the better outcomes of integrating immigrants in the United States compared to Europe is that America is more flexible: you can show up and on your first day drive Uber, and relatively quickly enter some other kind of profession. So it was surprising to learn how big the obstacles are even in many relatively blue-collar, working-class, middle-class professions. You can have been a hairdresser for 30 years in your country of origin, arrive in the United States, and in many states you can’t go and be a barber the next day.

Allensworth: There’s some reciprocity between states, but even that is pretty limited. As for international reciprocity in a profession like hair, there’s not much at all. Most countries don’t require the hundreds, sometimes thousands of hours that we require of a profession like cutting hair. It really tends to impact the lower-income professions, professions that would be good places for somebody who’s just starting out in the United States. Maybe they don’t speak English particularly well, so they’re going to go into a service-based profession where they don’t have to speak the language so well. Beauty would be a good example. We’ve seen time and time again licensing boards taking steps to keep those people out.

Some people think that licensing is progressive because if you have a license, you tend to have higher income and better working conditions, and that’s true even in the lower-income professions. But it’s actually regressive because within those communities, within those groups, it gives advantages to the ones that already have those advantages.

Mounk: I hadn’t even thought about the problem of reciprocity in this context. I’m obviously aware of it for lawyers, where some state boards have mutual reciprocity but a lot don’t, and that really limits the mobility of lawyers across state lines. I did not think that would be true of barbers: that if you are a barber in Tennessee, you may not be able to move to Michigan without having to retrain for a year or something like that.

Mobility in the United States has really gone down—people move a lot less now than they did 50 years ago, which goes a little against the intuition people have about those kinds of trends, but which is well documented. I wonder whether those kinds of rigid licensing requirements play a real role in that. If you’re a barber making a lower-middle-class income in one state and for whatever life reasons you need to move to a different state, you may be constrained from doing that because you’d have to give up your profession or go back to school for a year.

Allensworth: The reciprocity issue is a big problem with the lower-income professions too, and it’s especially problematic when you don’t have a lot of infrastructure behind you to help you figure all this out. If I’m a physician and I work for a health group at the border of Tennessee, North Carolina, and Georgia, my employer is going to handle all of that for me. I’ll have some assistant who’s responsible for figuring out all the continuing education I need to do every year, renewing my license—I’m not really going to see any of that. That’s an economic waste that gets passed on to patients and consumers, so it’s not great. But professionally, it doesn’t impact me as much.

I saw a case of a barber who was trying to move from Michigan to Tennessee and couldn’t get his license because Tennessee insisted on an English-language exam. He couldn’t pass it because he was not a fluent English speaker, but he had been cutting hair for many years, including in the military, and he had a license in Michigan—which he had obtained using an interpreter to help him understand the language. That was just a no-go. He didn’t have an employer or sophisticated people trying to help him through the process.

Mounk: I said earlier, in a slightly glib way, that these are really just guilds. This is the way that guilds have historically worked: it was incredibly hard to get in, and the point of making it incredibly hard to get in was to provide members with insider status, to make sure they wouldn’t have too much competition, wouldn’t have to change their methods too much, and could keep their salaries up. Is that the right analogy? Do you think this is basically a modern reenactment of the medieval guild system, or is that a misnomer? Obviously many guilds had hereditary elements that the new ones don’t, so there are going to be some important distinctions. But is that a helpful way of thinking about both the incentives at stake and the negative externalities, or not?

Allensworth: It is a helpful comparison and I think it’s apt, as long as we understand it as an analogy—it’s not a perfect one-to-one. As you pointed out, there are some differences. But some of the important things to keep in mind is that it is an inside-versus-outside mentality, and the psychology is very similar. I didn’t hear members of the profession or people on the board talking a lot about money and their bottom line, although I did sometimes. More often I heard them talking about identity and prestige, the importance of their work, and how not just anybody can do it. I think that is a major holdover from the guild perspective, where it really was about belonging. It was hereditary, but the same idea: we are in this group, we are in this family, this is what we do. Many of the benefits to members are very similar: higher earnings, less competition. It’s hard to get in, and once you’re in you’re kind of made. So I do think it’s a useful analogy.

Mounk: Part of the problem is that outsiders have a very hard time getting in. Medieval guilds often required you to be born into the right family—that’s not the case today, but there are these cumbersome barriers: you have to go to barber school for a year, which may be economically prohibitive, or go through years of supervised therapy, and many people drop off at that point.

The other idea, of course, is that you protect the insiders. In the medieval guild, if you’re the son of somebody who’s in the guild, you may not have any particular talent at that trade, but you’re going to be grandfathered in. You argue that there’s a kind of equivalent of that today as well, in the sense that even as these licensing boards make it incredibly hard for outsiders to get in, once you are in, you’re kind of set—often whatever you do.

Allensworth: That’s right. This was the most shocking part of my research, at least to me. I went into it with the idea that there’s going to be a lot of red tape, much of it unnecessary. I’m an antitrust professor, so I was looking at this through the lens of competition: they’re going to want to keep competition down, profits up. If you think about it that way, maybe when it comes to discipline, when a doctor or a lawyer does something wrong, you’d think they’ll be really hard on them, because kicking them out means one less professional to compete against. I saw the opposite. The self-regulation at the heart of these licensing boards resulted in second, third, fifth chances for practitioners who were just really incompetent and unsafe. That was harder for me to understand.

Mounk: Give us some examples, because from the book you’re not talking about somebody who gave a bad haircut once, or even had a safety violation once that is more concerning but was a one-off. You’re talking about people with a sustained pattern of blatantly illegal or blatantly immoral behavior.

Allensworth: I have so many examples it’s hard to pick one, but I’ll give a couple that illustrate it well. One is a real repeat offender. There was a doctor who was found to have a lot of marijuana in his home for sale, along with a bunch of drugs that he had clearly taken home from the hospital, either for recreation or because he was addicted to them. He took a state charge for it, pled no contest, and went on probation with a license. This is also a pattern I saw where the criminal system acts first, even though you would think the threshold for licensing action is lower—it’s easier to get a state conviction than to have real action on your license.

Subscribe now

The board, in response to this state criminal charge, put him on probation and mandated that he go to addiction treatment, as he was evidently addicted to benzodiazepines. While on probation and without the authority to prescribe opioids, which he had lost in that action, he found an end run: he had pre-signed scripts from a buddy, which he used to sell prescriptions for opioid use disorder medication and benzodiazepines out of his car for around three hundred dollars apiece. Totally illegal.

Mounk: Out of his car. God forbid a foreigner who’s arrived in the United States cut somebody’s hair, or your friendly neighbor help you install an alarm system—but giving people opioids or serious medications out of a car is totally fine with a licensed physician.

Allensworth: Keep in mind he’s on probation with the board. This is his second chance, and this is what he does with it. This is not just a violation of the licensing law—it’s a federal crime. He can’t plead no contest to federal charges; he’s going to have to do some time. In the face of this, the licensing board gives him another chance, and that chance looks an awful lot like the first: more supervision, more probation, please don’t do it again. He then finds another scheme, this time selling fraudulent COVID tests door to door. It’s a second opinion on COVID: if you have a positive test, he’ll sign a doctor’s note saying you don’t.

Mounk: Presumably, in order to be able to travel or do whatever.

Allensworth: Yes, travel—and this was high COVID, pre-vaccine, back when a positive test had real consequences. This is somebody who has manifestly no respect for the licensing board, no sense that it’s a real regulatory body. He’s kind of right: nothing really happened to him until he was finally sentenced to prison on the federal charges.

That’s one example, and he was a real operator—he found an angle on everything. But at the end of the day, I’m not sure he really hurt people. I also saw doctors who really hurt people get second chances. There was an OB-GYN who sexually abused eleven of his patients, traded sex for drugs with them, used drugs on the job, had extensive malpractice associated with his OB-GYN practice, ten peer reviews at the hospital, one stillborn baby. He got his license back. As far as I know, he still has one.

Mounk: A lot of the dynamics we’re talking about could be explained by Econ 101, or perhaps Econ 102. You have a club, that club is going to try to maximize benefits for its members, and it’s going to try to restrict access because if everybody can get in, the benefits for each member are going to be diluted. That explains why you make it really hard for an immigrant from Ghana or from Kyrgyzstan to become a hairdresser, and it explains why you want to maintain this licensing regime. Probably some of the people involved in the profession are also the ones running the barber schools, which explains why you want extensive requirements. All of that makes sense.

But on the discipline side, as you’re saying, there’s actually an economic incentive to throw people out of the profession. You might also think there’s a reputation management incentive to do so, since bad actors reflect badly on everybody. On the other hand, it makes sense that once you’re one of us, we’re going to protect you no matter what. You see that in other contexts, like police unions: once you’re a member of the club, the solidarity is unconditional. So how counterintuitive is this from a theoretical perspective? What does it tell us about what’s driving this behavior and how we should think about it?

Allensworth: This is harder to understand than the other side of things, the too-much-red-tape observations. You could tell an economic story, and I think the economic story is present when it comes to the associations. These associations are groups of professionals whose whole reason for existing is to protect and promote the members of the profession. This is not really supposed to be a public protection function. They want what’s best for doctors and lawyers, and what’s best for doctors and lawyers is having the most benefit of the doubt when they’re accused of something, having the most due process.

The reason why those more economic incentives work at the board level has more to do with psychology and sociology than economics. The associations provide the boards, which are made up of everyday working professionals who are not deeply expert in their field beyond what they are clinically competent to do, with narratives all about second chances and forgiveness. It’s also, ironically, about access to care and access to justice in the legal profession. If you take this guy out of circulation—and this was an important consideration in the OB-GYN case—he worked in inner-city Memphis, he worked with TennCare patients, which is the state’s Medicaid program. If you take him out of circulation, that’s a bunch of women who aren’t going to have any care. These narratives pull at the heartstrings of board members, and they also reinforce a sense of professional identity: don’t let someone else tell you what our profession is; we are the only ones competent to say what this profession is. So a little bit of psychology and a little bit of sociology, with some nudging from the associations.

The last point is that this was especially problematic in medicine, and I think it’s because doctors care a lot about making people better. They’re likely to see problems as attributable to sickness. Most of these physicians had struggles with substance use disorder, and so there was this temptation to see the whole problem as treatable, as fixable, even when they had really betrayed the trust of their patients.

Mounk: It would be interesting to dive into this case a little further. Given some of the things I know about the medical profession, including some of the behavior of the American Medical Association in recent years and some of the broader cultural trends among highly educated, affluent Americans, I wonder to what degree there’s also just ideology involved. The thinking that you can’t blame people for any form of addiction—which in itself I agree with—even when it leads them to act in highly dangerous and unprofessional ways. The argument that he’s treating a lot of women in poor inner-city communities is fair enough if he’s treating them well, but evidently he was not treating them well. This seems to me frankly like a bunch of identity politics and other forms of slightly credulous progressive ideas making it hard to exercise judgment in that context. Would that be fair to say about this case, which you know much better than I do, or do you think that’s not part of what’s going on?

Allensworth: The identity politics part of that particular case does map onto this. There was a lot of what you said in the first half—the idea that people can’t be held responsible for their addiction and their sickness—that was definitely present. But if anything, it was sort of the opposite when it came to the second problem. The inner-city, dominantly Black population that are his victims aren’t really the kinds of people the board sees itself as there to protect. The board’s mentality is really more oriented toward the mainstream patient who has commercial insurance and goes to a well-regarded hospital. That’s the world in which the licensing board members operate. In a lot of ways, I think they were throwing the less privileged patients under the bus with this decision.

Mounk: I certainly agree that they were substantively throwing those patients under the bus. But the way you framed it earlier was that he’s providing care to these inner-city communities and we can’t deprive them of that care. That sounds to me like a confused argument driven by a sense of: who are we to tell these women what doctor to go to, and who are we to deprive this community of the lovely doctor who has scores of malpractice suits against him?

Allensworth: The irony, of course, is that they don’t have a choice. The farther you go down the echelons of medical care, and for that matter legal care—because this problem is not unique to medicine, it exists in law too—the less choice you have about your provider. That’s why I think a lot of troubled physicians and lawyers end up serving the underserved: if they have a troubled, checkered history, patients and clients either aren’t going to know that, or if they do know it, they’re not going to have much of a choice. That was another thing that was really surprising in my research that I don’t think a lot of people are talking about: the caste system we’ve created through the disciplinary system, where we take people who are very problematic to begin with and shunt them towards the most needy patients, clients, and customers who don’t have a choice, including working in prisons and, in the legal world, representing immigrants and indigenous people.

Mounk: I’m trying to think through what the devil’s advocate argument for the licensing regime would be here. Presumably some of this sorting would happen in any case. If you had no licensing regime at all, doctors and lawyers with a particularly good reputation would be hired by the best hospital systems, and those with a very bad reputation would end up providing services that others don’t want to provide because they’re less lucrative—and the same in the legal profession. The argument for the licensing system, even if it isn’t doing this perfectly, is that it at least provides some protection even for the least privileged segments of the population, and that in the absence of licensing you would get that effect more strongly rather than less strongly.

Allensworth: That’s true, but there’s a real irony to the structure of that argument. On the one hand, you’re saying we need licensing to protect the people who don’t know what they’re getting—which is true of a lot of professional services—or who don’t have a choice, which is also true of a lot of people who consume those services.

Mounk: If we had no legal licensing, the problem isn’t that when Bank of America needs to hire lawyers, they’re going to hire frauds. Bank of America is going to be a big and sophisticated enough organization to figure out which lawyers are good and reputable. The problem is if Joe Schmoe, or an underprivileged Joe Schmoe, is in desperate need of a lawyer, they might end up with a real huckster.

Allensworth: When we’re talking about raising barriers to entry and the red tape issue, we care a lot about quality and protecting people. We do that by creating all these barriers to entry and making the population of providers smaller and smaller, which isn’t necessarily good for the public because it makes those services more expensive, harder to get, more rare, harder to access in rural areas. But for the people who are able to get them, they’re going to be above a threshold. Except that then when it comes to discipline, you totally abandon that threshold, and abandon what seems to me like the most plausible justification for the licensing system in the first place.

Mounk: Which is to make sure that the doctor who clearly has a substance use issue, who’s clearly not acting in the interest of patients, who has a score of well-founded malpractice suits, doesn’t get to keep treating patients.

Allensworth: Let me give you another example. Let’s say you could make one rule that said nurse practitioners can work up to the scope of their practice—they can do a bunch of things that they can’t do in some states, including Tennessee, without physician supervision. You’re going to expand access to care for potentially a million people through one rule. Then later in the afternoon, you can see a disciplinary case for one doctor who, that we know of, has killed five patients by prescribing them so many opioids that they overdosed and died, not to mention the hundreds of others he did this to. You keep that one doctor in play for access to care. It just doesn’t add up. You can do so much more for the access to care problem on the rule side than you can on discipline.

Mounk: This is perhaps not directly about licensing boards, but it’s part of a wider system where we’ve delegated a lot of self-governance to members of a profession. My example, which went viral when I wrote about it, is that I went kayaking on the Connecticut River a good number of years ago and ended up losing my glasses when the kayak turned over. I needed to get new glasses, and it was so difficult. There aren’t a lot of opticians around there, and it takes a while to get an optical test. Everywhere in Europe, you can just walk into a store and say, make me glasses with this prescription. Either a store employee uses the same machines an optician would use to test the strength of your eyes, or you just tell them your last glasses were this strength and they make them. In the United States, you can’t do that.

I called for a lifting of those restrictions, the article went very viral, and I was denounced for spreading misinformation by the president of the American Optometric Association or some such body. This is exactly the kind of insider dealing at work. The people making the rules about when you can produce glasses are the ones whose livelihood depends on giving people eye exams all day long, and they will find arguments, however spurious, for why that is absolutely medically necessary, when very few other industrialized countries consider it necessary under the same circumstances.

What’s the cost of that? The cost is the marginal person who struggles to pay for glasses for themselves or their child, who delays taking their child to the optician when they can’t see the blackboard properly, who has to wear the same pair of glasses even as they’re falling apart and their prescription has deteriorated—because that extra cost of the eye exam is the barrier that stops them. It has a really direct negative impact. It’s good, if you can afford it, to go to the eye doctor every now and again and get your eyes checked, and opticians absolutely provide a valuable service. But to make it mandatory in this way is just out of all proportion to the trade-off.

In general, I think it’s a very positive feature of Anglo-Saxon societies that there are private bodies with a real ethos of self-governance, as an alternative to having everything directed by the state. But on balance, would it be more rational to have a state institution make those decisions, one that is less directly beholden to the interests of the profession? The disadvantage of the way this is set up in the United States, the United Kingdom, and most Anglo-Saxon jurisdictions is that you create a direct conflict of interest: the people making decisions about who gets admitted to a profession, and about things like what is required to prescribe glasses, are always members of a profession with a very obvious conflict of interest.

Allensworth: The whole book is really about this problem of conflicts of interest, of handing it over to the professionals to regulate themselves. We have to go back to the medical profession to see how this happened. Medicine was the first profession to really professionalize in the way I’ve identified, where they said: this is our profession, only we know what’s safe, what’s good, what’s real medicine, and so we are going to decide who’s in and who’s out. The key move was to then get the state to say that you can’t practice medicine at all without their say-so. Every profession has basically followed this model of self-regulation, using the state to back it up. Much as it may make sense that medicine is only fully understood by doctors, that does not change the fact that there’s a huge conflict of interest between what’s good for the public and what’s good for doctors. You need guardrails, you need government involvement to make sure this kind of self-regulation doesn’t go too far. Moreover, it doesn’t make sense to reproduce the medical model all the way down to every profession, including alarm system installers, where that kind of expertise doesn’t really play in.

Let me give you an example of another area we all care about deeply, especially as parents: childcare. Childcare encompasses everything from the neighborhood kid babysitting all the way up to the fanciest pre-K in New York City. There are no laws saying you can’t babysit, no laws requiring certain qualifications, even though this is a very safety-sensitive activity. We do let the private market work itself out as far as what kinds of regulation you opt into and pay for. We also have mandatory regulation for childcare centers, but it’s not licensing—we don’t make you go to school for a year and pass an exam. It’s more about background checks, CPR certification, a few safety rules. This is what I would want to see for something like eyeglasses: a system where people may opt into a more expensive, more comprehensive option, while others can just get their kid glasses even if it’s not a hundred percent perfect. We have evidence from other countries that a much less onerous system works basically fine.

Now imagine you’re a legislator, this is a big deal for you, and you want to pass a bill getting rid of the licensing regime for opticians. They are going to make it such a headache for you. They’re going to come up with every possible example of some child who did or didn’t die of cancer because they did or didn’t get an exam from someone trained to recognize it. There’ll be nobody on the other side.

Mounk: What the president of the American Optometric Association was saying is that she wants Americans to be able to catch dangerous eye diseases, tumors, or other conditions that are sometimes discovered during a mandatory eye exam. I take that seriously—if you have the means to go get a professional eye exam, absolutely do that. But she wasn’t talking about all of the people who die in traffic accidents, or who don’t learn to read in school, because they can’t see properly—because glasses are far more expensive here than in any other country, and they haven’t been able to afford the glasses or contacts they need to go about their daily lives.

Allensworth: This is a common argument I hear about hair professionals too—that they can recognize diseases of the scalp, and I’m sure that’s true. That’s partly why I bring up the childcare example. There are lots of cases where somebody who is good at childcare could intervene in a way that would be really helpful for a child, or fail to intervene in a way that doesn’t turn out well. It’s something we all care a lot about, but it’s a problem we solve in a different way.

In the rest of this conversation, Yascha and Rebecca discuss how to develop a professional licensing system that works and why the opioid crisis revealed the weaknesses of licensing. This part of the conversation is reserved for paying subscribers…

Read more

Steve Stewart-Williams is a professor of Psychology at the University of Nottingham’s Malaysia campus and runs The Nature-Nurture-Nietzsche Newsletter. His latest book is A Billion Years of Sex Differences.

In this week’s conversation, Yascha Mounk and Steve Stewart-Williams discuss why women and men are more similar than is often thought and what the real sex differences between men and women are, from casual sex to career choices.

This transcript has been condensed and lightly edited for clarity.

Yascha Mounk: Last time we spoke, you gave a really great introduction to how to think about the influence of evolution on human nature and human psychology. You are now out with a book that speaks more specifically about sex differences and how those are rooted in biology. Let’s go through some of the sex differences that the literature establishes relatively decisively.

One way you break this down is to say some sex differences are just completely dimorphic—women give birth, men do not—while others are more statistical variances. Tell us a little bit about how to think about the kinds of sex differences and what some of the most common ones are.

Steve Stewart-Williams: The new book is called A Billion Years of Sex Differences. I start by listing what I call the standard issue sex differences: a list of very well-established sex differences in our species. As you say, they range from the very large and strictly dimorphic and categorical, to the not-so-large—at the other end, just statistical differences, relatively modest discrepancies in the average scores of massively overlapping distributions.

At the larger end of the spectrum, we generally just have physical differences between the sexes: reproductive differences, the reproductive organs, the capacity to nurse versus not, the capacity to give birth versus not—those are all very large differences. The difference in upper body strength is quite large—actually very large, with very little overlap. There is very little overlap either in voice pitch, so voice pitch is another quite large sex difference.

Moving down the scale, when we are talking about psychological sex differences, the only one that is even remotely in the ballpark of those physical differences is the sex difference in the primary target of sexual attraction. Most men are primarily attracted to women, most women are primarily attracted to men—there are exceptions, of course—but nonetheless, that is a very large psychological difference. It shows, I think, that natural selection can create large levels of dimorphism.

Mounk: That is interesting, because in the first category—and we are certainly not going to get into the whole trans debate for purposes of this conversation, we are talking about biological sex—men can produce sperm, women can produce eggs. Those are just completely different biological processes: strictly dimorphic. Here, we are talking about how most men are attracted to women and most women are attracted to men, but then there are these subcategories of men who are strongly attracted to men and women who are strongly attracted to women. So that is a very different kind of distribution.

Stewart-Williams: It is a different kind of sex difference. Even though people attracted to the same sex are exceptions and a minority, they are not a trivial minority—there is a non-trivial number of exceptions. Whereas that is not the case for the first sex differences I named.

Mounk: Not to go down a rabbit hole, but a lot of people are going to have in their mind this idea of a Kinsey scale—that actually most people fall somewhere on a long spectrum between complete heterosexuality and complete homosexuality. So it’s not that most men are attracted to women and then some significant percentage of men happens to be very attracted to men, but rather that we all fall somewhere along that spectrum.

Is that compatible with modern science? Is that something that scientists today believe? How does what you are saying fit onto that kind of model of the Kinsey scale?

Stewart-Williams: I was thinking of the Kinsey scale when I framed the exact wording. My exact wording was that most men are primarily attracted to women, most women are primarily attracted to men—so that accepts the fact that it is not uncommon at all for there to be some degree of attraction to the same sex among people who are primarily attracted to the other sex.

There are differences in the degree of attraction to each sex that individuals have. One thing is that people often understand it to mean that there is a spectrum and that people are distributed fairly evenly across it, whereas you do actually have clustering toward the two extremes of the continuum—that is, primarily attracted to the other sex versus the same sex. That is especially true with men. With men you get quite strong clustering, either primarily attracted to the other sex or the same sex. With women, the distribution is a bit more even across that scale.

Mounk: So if you are picturing the scale from left to right, there is one big mountain on the left—attraction to the opposite sex—and then a smaller mountain on the other end of the scale, primarily attracted to the same sex, and then in between are all kinds of people who fall at all kinds of points, but those are not going to be big local maxima?

Stewart-Williams: That is exactly right, with just the one proviso that there are fewer men in the in-between category than there are women in the in-between category.

Mounk: So the peaks are going to look higher for men than for women for the same reason. Interesting. That was the second kind of category. The third category is where most of the interesting debates lie. The first category—the biological differences—is relatively straightforward and uncontested. The second category is super interesting and important, but perhaps applies to slightly fewer characteristics. A lot of the things that we really have big social debates about fall into the third category: is it true that men have an innate attraction to certain kinds of professions, to certain kinds of socializing, to certain kinds of ways of thinking about the world? All of that contested stuff lies in the third category—minor statistical differences that might nonetheless make a really significant impact on the world. So tell us about those.

Stewart-Williams: That is exactly right. The vast majority of psychological sex differences in our species are in that third category—statistical sex differences, or what I call fuzzy sex differences—where you have differences in the average of highly overlapping distributions. A big part of the reason there are so many debates about it is because they fall awkwardly into a middle ground where a lot of them are small enough that you can question whether they actually exist, but large enough that they trouble us. So they sit in that awkward middle ground where you can deny them, but they also have real-world effects to some degree.

They span a range of different magnitudes as well. Among the larger ones, you mentioned career-related interests—that is actually one of the larger sex differences in this category. There is usually about a standard deviation difference between the mean for men and the mean for women in career-related interests.

Mounk: Can you explain for lay listeners what that means—both in terms of how strongly a different professional preference drives professional choice, and roughly how big an effect a standard deviation is?

Stewart-Williams: On average, men tend to be more interested than women in things—how machines work, mathematical and formalized systems, impersonal subject matter—and more interested in careers involving those things. Whereas women on average tend to have a greater interest in people, how people work, and in people-related professions: professions that involve helping and interacting with people.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

The difference is about a standard deviation between the means for the two scores, which is considered to be quite a large effect size. There is still a great deal of overlap. What it would mean in practice: if you were to choose two people at random, one man and one woman, for interest in things—where men score higher—there would be about a 70% chance that the man would score higher than the woman, and about a 30% chance it would go the other way. The converse holds for interest in people: in about 70% of randomly chosen pairs, the woman would score higher than the man, but there would still be a very non-trivial minority of cases where the man scored higher than the woman.

Mounk: That is, I think, a little bit of what makes this awkward. Let’s cash this out with this relatively uncontroversial example, and then we can think about how it applies to more controversial ones.

What you are saying is that in the literal sense, stereotypes about this are right. There is a stereotype that men love their machines and their tools—they go into the man cave and potter around in the workshop—and that women gravitate toward more social professions, or, pejoratively, gossip, or whatever. You can make a negative caricature of each, or a positive one. The positive version: the man likes to build things and get things done, and the woman loves to be a connector in the neighborhood, making sure people are doing well and cared for. Or you can put a negative spin on each: the man locking himself away in his workshop, or the woman engaging in mean-girl gossip behavior. The point is that this is a stereotype we hold in society, and what you are saying is that there is some real empirical grounding for it.

And yet, of course, if you apply that to everybody—if you were hiring at an engineering firm and said, well, this is a female applicant, surely she does not actually like machines—you would not only be perpetrating an injustice towards that woman, but you would very often be wrong. A good number of women are interested in these kinds of things, and among women who are actually applying for such a position, probably most of them are.

Stewart-Williams: That is exactly right. A lot of stereotypes are on the right track—at the very least they are a response to something real. The danger is in exaggerating the magnitude of them, which is what you are doing if you say: this is a woman applying for a male-typical job involving machines, it is just impossible that she could be interested in them or good at working with them, and vice versa for men applying for people-related jobs. You would be exaggerating the strength of the sex difference.

I think that is a big reason why people are nervous about talking about sex differences: they worry that people are going to do exactly that. They worry that if there is any acknowledgement that there is some truth in these sex differences, people are going to misunderstand it, blow it out of proportion, and that it will become a justification for—or a cause of—discriminating against one sex or the other. I do take that concern seriously, and in the book I discuss it as something we genuinely need to concern ourselves about.

Nonetheless, I think it is really worth exploring this topic for two reasons. One is that whether we like it or not, there are average differences between the sexes in many different areas, and we cannot simply lie about that. The second reason is that even though there are dangers in talking about sex differences—dangers associated with exaggerating them and moralizing them—there are also dangers in the converse: in minimizing them, ignoring them, and moralizing the absence of sex differences instead. The only responsible solution to that conundrum is to look at what the data say, try to convey it accurately, and talk about sex differences carefully.

Mounk: Motivate where the dangers lie. The danger of the first kind of error is relatively straightforward: I am a hiring manager at an engineering firm, an excellent female candidate applies, and I say, well, she is a woman, she is not really going to want to be doing this kind of job, and I do not hire her even though she is an excellent, well-qualified candidate. That is a very straightforward danger. Where does the danger lie in not talking about these sex differences?

Stewart-Williams: There are quite a few, and there are actually other dangers with exaggerating sex differences as well—for instance, if we develop a mental image of what men and boys are supposed to do and what girls and women are supposed to do, we might pressure people to fall into stereotypical definitions of the sexes, which is bad for people who do not fit the norm.

Mounk: That is obviously related, but slightly different. In the first case, I am a hiring manager at an engineering firm—I may have no belief that it is wrong for women to want to be engineers. It would be great if women were super into engineering, but sadly not all of them are, so I am not going to take this candidate seriously. Now we have moved to a slightly different thing: because there are some statistical sex differences in the world, that is in fact what men and women should be like. So now let’s say I am a dad and my daughter is interested in engineering—well, that is not for you, you should really study psychology. Here we have moved to the danger of ignoring the is-ought distinction.

Stewart-Williams: There are dangers with exaggerating and moralizing sex differences, as I just mentioned. But there are converse dangers with minimizing sex differences and moralizing the absence of sex differences. One of these is directly symmetrical with the problem of trying to force men and women into traditional gender roles: there is a danger of trying to force them out of those roles—of not accepting the fact that, although there will always be exceptions, some people are simply going to fit with gender-typical norms, roles, and preferences. People sometimes have a tendency to think that that is a bad thing, a product purely of sexism, or—if it is their own kids—to feel guilty that they have done something wrong in parenting because their child falls into typical patterns: boys engaged in rough-and-tumble play, girls loving pink and dolls. That is one danger.

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

Another is that if we ignore sex differences in career-related preferences, then when we find discrepancies in the sex ratios of different professions, we will assume it is one hundred percent due to discrimination and sexist socialization. While it is very probably partly due to those factors, it is not necessarily—and not even likely to be—the case that they are the only causes. Given that men and women do have, on average, these different preferences, it seems very likely that part of the reason we have different sex ratios in some professions is a result of those statistical differences in preferences. In fact, given that those different preferences exist, it seems very unlikely that they would not be one of the causes of the differences.

Mounk: There seems to be some preliminary evidence that men are not just more interested in tools generally, but also more interested in a specific tool: artificial intelligence. If that turns out to be the case, at a moment when it seems very likely that AI is going to fundamentally reshape our society, is that going to be to the detriment of women? Are we entering a technological age in which professions that many women occupy—like human resources—become relatively less lucrative, while professions involved in creating these new AI systems, in which men continue to be overrepresented, come to dominate?

Stewart-Williams: That is a very interesting question, and I think it is possible. One reason it is not necessarily going to turn out that way, though, is that the interface between people and AIs is quite a social activity—it does not feel like tool use in a traditional sense. You are interacting with something verbally; it can feel quite social.

I was recently at the How the Light Gets In festival at Hay-on-Wye, and I met a few people who work in AI. One was looking at how AI can be used in the medical sphere—that is Charlotte Blease, who is very interested in applying it there to help people. In that area, you might actually find a predominance of women over men being interested in the question. Someone else I met was very interested in the use of AI for psychotherapy—again, a field where women currently predominate over men. It is possible that if AI moves into that area in a big way, there will likewise be a predominance of women over men in AI applications of psychotherapy.

Mounk: Let’s make a simple prediction. If you think that professional choice is really down to socialization and social pressure—informal prohibitions on men becoming hairdressers or women becoming engineers—then you would expect that the more gender-egalitarian a country is, the less pronounced those gender differences in professional choice would be. So you would expect that in a more traditional society like Egypt, there would be a large difference, with very few women becoming engineers and very few men becoming hairdressers. And in Sweden, or whichever country is relatively gender-egalitarian, you would expect those differences to be relatively more minor. My understanding is that that is not in fact what the evidence shows.

Stewart-Williams: Your understanding is correct. I would have made that same prediction before I saw the data. Even accepting that there is an innate contribution to preferences, I would have thought that in cultures where people are treated more differently and gender roles are stricter, sex differences would be even bigger—so if there is an innate contribution, it would be amplified—and then smaller in more gender-egalitarian, more individualistic, and wealthier societies where people are freer to pursue their own interests.

But as you say, there is this phenomenon called the gender equality paradox, which reflects the fact that for many traits, including career-related preferences and job choices, it is the exact opposite: the more gender-equal, wealthy, and individualistic a nation is, the larger those sex differences become rather than smaller. Very counterintuitive.

Mounk: Really interesting finding. I had an interesting episode that touched on that with Alice Evans on the global gender paradox, among other things.

One of the things that makes these conversations often unproductive—and I think this is true not just of this conversation but of many conversations—is that different people have different audiences in mind, and that drives them to talk about things in slightly different ways. Let me give a totally different example.

A lot of social science disciplines, like sociology, love to make the move of saying: this thing actually turns out to be socially constructed. I find that useful for an introductory undergraduate class, in the sense that a lot of 18-year-olds come in thinking that something like the United States Constitution is a God-given thing that is objectively correct and beyond question. To say: these are particular people who built this at a particular moment—brilliant people doing something important, but with their own shortcomings and blind spots—is a helpful corrective. I am not sure how many 18-year-olds still arrive at college with that kind of naive view, but I can see where it comes from. There is a danger, though, in over-inflating what it means for something to be socially constructed, which is a general pet peeve of mine, and also in making your whole academic career invested in pointing out that this or that thing is not natural or God-given—speaking to an audience that already agrees with you.

I think something similar happens when people talk about gender differences. A lot of people, for understandable reasons, think: there are stereotypes and prejudices in society that coerce men and women into gender roles and obstruct women from advancing in certain professions—that is really what we need to be on the lookout against. From that perspective, writing a book about sex differences seems actively unhelpful. Then there are others who say: the mainstream has become so concerned about that real danger that it has overshot the mark and gone in the other direction, producing all kinds of other confusions—and that is what we should be setting ourselves against. We can call these alpha bias and beta bias. Talk us through that.

Stewart-Williams: Those are not my terms, but they are very useful ones. The alpha bias is the tendency to exaggerate the magnitude of sex differences, and the beta bias is the opposite—the tendency to minimize, downplay, and perhaps deny them. There is quite a strong tendency toward the beta bias in media and academia, in the mainstream and among the chattering classes. The alpha bias is perhaps the historical starting point, and it does exist in some small areas of academia and probably journalism as well. I would have to say the tendency to exaggerate sex differences sometimes exists in my own field of evolutionary psychology—perhaps because we are responding to the beta bias we find elsewhere in academia, and it may have swung to the opposite extreme to some extent.

What I am trying to do with the book is respond to both, critique both, and walk a happy medium between the two. I would love to think that people on both sides will agree and perhaps come to a truce—but of course another possibility is that I will simply annoy people on both sides of the debate. That is nonetheless what I am trying to do: strike a balance.

Mounk: Let me pose the question so we can address this directly. To make things more complicated, you then add two further biases—the gamma and the delta bias. What are those?

Stewart-Williams: The gamma bias is the tendency to highlight sex differences that put women in a better light than men, while downplaying those that put men in a better light than women. I think it is a response to historical anti-female sexism—basically an attempt to set right the reverse tendency, where differences were highlighted in ways that cast men as superior and women as deficient. It is a kind of overcorrection that aims to be protective of women.

Subscribe now

The delta bias, on the other hand, is a tendency to highlight and favor sex differences that go in the opposite direction to traditional sex differences. For instance, devoting everything to your career and placing family lower down the list might be frowned upon if a man does it, but celebrated if a woman does it. Sleeping around is sometimes considered a bad thing if men do it, but a good thing if women do it. I think that is a response to traditional social norms.

Mounk: That reminds me of one of the very early viral internet controversies—the Duke “Fuck List,” if you remember that. A senior at Duke University had made a joking PowerPoint, styled like a senior thesis, about all the men she had slept with in college. In a way you can see how it was a fun, charming idea, though the slides themselves were quite demeaning and objectifying—ranking all of these men in various ways. Some people thought it was a mean way to interact with and write about other people. But Jezebel and the early internet feminists defended it as a great act of female empowerment. It was obvious that the same people, had it been a man ranking and grading the genitals of women he had slept with in similar ways, would have found it to be terrible. But because this was typically something men did and not women, when a woman did it, it was considered empowering.

Stewart-Williams: That is a good example of the delta bias, and also a good example of where I think it comes from—different standards are applied because of the history, because it is turning the tables. Those are the four biases that I think make the discussion harder and make it trickier to research and talk about sex differences without it getting heated. There are these biases and preferences that people have for what kind of sex differences they want to see or not see.

Mounk: We have done a lot of the preparatory ground for thinking about how to approach these questions. Let’s go a little deeper into how we should think about particular sex differences in light of all this. Why don’t you tell me the three most juicy ones we should talk about for the next twenty or so minutes?

Stewart-Williams: How about the sex difference in interest in casual sex, in choosiness about sexual partners, and in aggression?

Mounk: Let’s start with casual sex.

Stewart-Williams: The difference here—and you will be very surprised to hear this—is that on average men are more interested in casual sex, no-strings-attached sex, and sexual variety than women are. That is a pretty big difference as human sex differences go. Depending on exactly what you ask about, the difference is about 0.8 of a standard deviation between the means for men and women, with men scoring higher. To use the analogy of plucking two people at random, one man and one woman, for that sex difference you would find about two thirds of the time that the man would be more interested in casual sex than the woman. Again, there is a significant minority where it would be the other way around. So it is a real sex difference and it is pretty substantial as human sex differences go, but it is not quite as big as people sometimes make it out to be.

This raises an interesting point about sex differences in general: even when you have a sex difference that is fairly modest at the mean, when you go out to the extreme of the distribution, the difference will typically get a lot bigger. If you go out to the extreme of the distribution for interest in casual sex, you will find that even though it is about 0.8 or one standard deviation at the mean, out at the extreme there are many more men than women. I think a lot of our stereotypes come from thinking about those extremes rather than about people closer to the mean.

Mounk: It is just a function of how normal distributions work. If you have a relatively small difference at the median—where the top of the distribution is—that relatively small difference can mean that once you are three or four standard deviations out, where you are talking about people at the extremes—the most or least intelligent, the most or least interested in casual sex, the most or least violent—that small difference in the mean can translate into a four, five, or tenfold difference at those extremes. Now that we have established those statistical properties, how does that shape social behavior—and things like casual sex in particular?

Stewart-Williams: It is often going to be the case that men are pursuing casual sex to a greater extent than women are. One mistake people often make when they hear about that sex difference is thinking that the claim is that men are more interested in casual sex whereas women are more interested in long-term relationships. But actually men and women have similar levels of interest in long-term committed relationships—it is really only within the domain of casual sex that we find this sex difference.

There are actually two differences here. One is that men are more interested in casual sex on average. The second is that when it comes to casual sex and low-commitment relationships—and also early courtship—women tend to be choosier about their mates: choosier about who they are willing to have a short-term thing with.

Mounk: On the first topic, before we get to mate choosiness—what do lesbian and gay relationships tell us here? Obviously there may be ways in which lesbian women’s sexual behavior differs from heterosexual women’s, and gay men’s sexual behavior differs from heterosexual men’s—not just in terms of the object of attraction, but presumably this is at least a suggestive field of study: what happens if you take two people with male sexual preferences and put them in a relationship together, versus two people with more female sexual preferences in a relationship together? Obviously there is going to be a huge range, from very monogamous gay couples to very open ones. But do we see systemic differences between lesbian and gay couples, and what does that tell us?

Stewart-Williams: We do see differences, and they stem from the fact that in same-sex relationships, people are not having to negotiate with the preferences of the other sex. They can typically express their own preferences to a greater degree, with less compromise. Looking at the sexual behavior of gay men and lesbians can therefore provide a clearer window into the sexual preferences of men and women in general.

What you find is that gay men typically engage in a lot more casual sex than straight men, because they are interacting with other men who typically have similar preferences. Lesbians, on the other hand, typically have less casual sex than straight women—again, because they are not having to accommodate the average preferences of the other sex. It fits very nicely with the idea that there are these average differences between the sexes when it comes to interest in casual sex, and it is a good piece of evidence for that idea.

Mounk: You were saying that men do not necessarily have less interest in long-term committed relationships, but they do have a much greater interest in casual sex. If you just know those two things, you would predict that you would end up with a lot of couples that are open to each partner having casual sex—which is certainly not the universal norm among gay couples, but is true of many of them.

Stewart-Williams: It is true of more gay couples than straight couples. In straight couples, some men do want that, but sexual jealousy often gets in the way—a man might think: it would be great if I could sleep with other people, but I do not want my partner sleeping with other people, so I will content myself with monogamy. That seems to be less often the case with gay men. I am sure many do fit that description, but there do seem to be more gay couples where jealousy is not quite such a big issue as it is in straight couples—that is the impression I have, anyway.

Mounk: Just to up the level of intrigue here—I have read, though I am not sure if this is true, that there are significantly higher rates of divorce among lesbian couples than among gay couples. Why is that?

Stewart-Williams: I have read the same thing—there are good data on that. Why that is, I am really not too sure. It could be a selection effect, or it could be related to the fact that among straight couples, around two thirds of divorces are initiated by the woman rather than the man. Maybe it is an offshoot of women’s greater choosiness about their mates—our next juicy topic. But I am not a hundred percent sure about that.

Mounk: Let’s go to choosiness about mates. Two things here are striking. The first is that evolutionary theory would suggest that women are choosier about mates than men because they invest more in offspring—if only through pregnancy, and often well beyond that. That is true throughout the animal kingdom and is true for humans as well. The other thing that is often forgotten—and you will have to remind us of the Greek designation for it—is that actually the difference between the two sexes in humans is relatively smaller than it is in most other animals in this regard.

Stewart-Williams: At least for long-term relationships it is smaller. When we think of women being choosy and men being not choosy, that has some truth in the short-term context.

Mounk: So at 2 a.m. in a club, the man is happy to go home with whatever woman, but the woman says: even though I might enjoy casual sex, I am still going to be quite picky. But when it comes to long-term relationships, suddenly the men become picky too.

Stewart-Williams: Both sexes tend to be pretty picky in long-term relationships—and maybe similarly picky. So you get the alpha bias, the exaggeration of sex differences, when it comes to the long-term context. It is a general truth, actually, that sex differences in sexual psychology are a lot smaller when it comes to long-term committed relationships than to short-term relationships, casual sex, and early courtship. It is in early courtship and low-commitment relationships that the sex differences are a lot bigger.

Mounk: Is this a challenge for evolutionary theory? Or is it simply downstream from the fact that humans procreate in very different ways from most other animals, because they invest in offspring for so much longer?

Stewart-Williams: It is not a challenge to an evolutionary perspective—I think it fits very nicely with one. According to the evolutionary perspective, the greater female choosiness we see right across the animal kingdom follows directly from the fact that female animals in general invest more in each offspring they produce than male animals do. Because they invest more, they can have fewer offspring in their lifetimes. Males, on the other hand, typically invest less in each offspring, and therefore in principle could have many more offspring than any female could. That creates different selection pressures on both sexes, and produces a range of traits in males that increase their chances of being one of the few males that has many offspring, rather than one of the many males that has none or just a handful. That is where most sex differences come from: the difference in the maximum offspring number each sex can produce.

We have some of the same thing in our species, but because of the great dependence of our young, we have evolved a reproductive system where we fall in love, form pair bonds, and engage in high levels of biparental care—much higher than you see in most mammals. When males are investing heavily in their young as well as females, that reduces the sex difference in the maximum offspring number each sex can produce. It brings down the ceiling number of offspring that men can produce—a few exceptional historical examples notwithstanding—and as it brings it down, it has led to the evolution of reduced sexual dimorphism in our species. One area where that is evident is in the sex difference in choosiness.

The sex difference in choosiness is a result of females investing more, but in a long-term context both sexes are typically going to be investing quite heavily—in the relationship and in any offspring that result from it. For exactly the same reason that females evolved to be choosy when they can only have a relatively small number of offspring, men likewise evolve to be choosy in a context where they are only going to have a relatively circumscribed number of offspring.

Mounk: I once leafed through a book arguing that when we compare humans to other animal species, we fall somewhere towards the middle of the range on the extent to which we are wired to be monogamous—that some species are strongly monogamous, and many species, perhaps most, are not monogamous at all.

When you look at a whole set of physical characteristics of humans, they tend to fall somewhere in the middle between those extremes. The book was arguing that we are semi-monogamous, and that this is how to think about us and how to explain a lot of conflict in social life—the fact that we have both a tendency towards pair-bonded long-term committed relationships and a tendency beyond that. Do you find that a helpful way of thinking about this topic?

Stewart-Williams: We have a strong tendency toward pair bonding and biparental care, but it is not like gibbons or some birds. We also have an interest in sexual partners other than our main partner, which is not uncommon, and that nudges us toward arrangements other than strictly monogamous pair bonding. Men have a desire for sexual variety—some members of both sexes do, but men typically have a stronger desire for it—and that sometimes translates into polygyny, where one man has more than one partner. Pair bonding is the most common relationship type found in basically all human societies, including polygynous ones. Even in societies where men can take more than one wife, it is a minority activity. It is not like gorillas, where a male either has a harem of females or is a bachelor—with humans, most men who have more than zero mates have only one. There are a small number who have multiple mates. Pair bonding is our main reproductive approach, but we are very flexible and do all sorts of other things as well.

On to the third topic: aggression. At least when it comes to direct or face-to-face aggression, men on average do more of it. That includes verbal aggression, where the effect size is medium—about half a standard deviation between the means for men and women, so not a massive sex difference. But as you go up the scale toward more intense forms of aggression—physical violence, pushing and shoving, more extreme forms of physical violence—the gap gets bigger and men come to predominate more. At the extreme of one-on-one violence, which would be homicide, men predominate massively: around ninety percent of homicides worldwide are perpetrated by men. Only a minority of men commit homicide, but among those who do, it is massively skewed toward men.

When it comes to indirect aggression—the mean girls phenomenon, gossip, spreading rumors, where you are trying to harm someone but not to their face—we find either no sex difference, with both sexes doing it about evenly, or we find it is slightly more common among women than men. But for most forms of aggression, men predominate.

Mounk: There is another area where stereotypes seem to hold up to some degree. When you talk about the extreme forms of violence, that difference is very obvious in the statistics and easy to understand—men are more likely to murder, more likely to be soldiers, more likely to be in criminal gangs, and prisons are much more full of men than women. But when you are not talking about the extremes, when you are talking about people towards the middle of the distribution, how big is that difference in standard operating procedure? Is it that the average man is not particularly violent, but perhaps once or twice in his life has gotten into a small scuffle at a bar? The average woman is not particularly violent, but when she engages in social aggression, it takes the form of rounding up a few people to ostracize somebody in the friend group they have a grievance with?

To what extent are those stereotypes about the middle of the distribution true? Is that a helpful lens for understanding society? As an evolutionary biologist and psychologist, do you feel like you can watch Cruel Intentions or whatever high school movie and check off the behaviors—that your training helps you understand what is going on in ways it would not for someone without it? Or do you think it does not in fact illuminate social reality in that kind of way? Is the alpha bias or the beta bias more tempting here?

Stewart-Williams: That is a good question. I think with low-level aggression the alpha bias comes in more, and people exaggerate. People are maybe a little surprised when they hear that the difference is not larger for verbal aggression and the like. The way you describe it is pretty accurate for modern Western societies: most men are not especially aggressive—they may get into an occasional scuffle over the course of their lifetimes, or more commonly the odd verbal altercation—and likewise women are not especially aggressive either.

Women may have as many verbal aggressive interactions, but with less of a sense that there is a physical threat and that it could turn into actual violence. I think that is accurate in the West. We have successfully tamed our aggressive impulses to quite a high level in the West, and in other and more traditional cultures there were higher levels of violence—especially among men. Culture brings it out more in some contexts, while simultaneously suppressing it somewhat in modern Western populations, which I personally think is a good thing. I imagine most people would agree. It is an example of a case where we have a natural tendency that society pushes against—but it is a tendency that is generally a bad thing, so it is a good thing that we are able to engineer unnaturally low levels of aggression.

Mounk: One benefit of thinking about the world in evolutionary terms is that it helps us see where the dangers lie for society. One very fundamental insight from evolutionary biology and evolutionary psychology is the human tendency to split into in-groups and out-groups. You could fall into the naturalistic fallacy and say: therefore we should encourage people to stick to their tribe and hate the other. But we have learned from history that that is a disaster. The whole point of civilization, the whole point of building positive institutions, is to mitigate the impact of the in-group and out-group tendency. The way to do that is to channel it—not to deprive people of any outlet for it, but to redirect it.

That is why you have sports rivalries in high schools. That is why, if you think of Harry Potter, you split a boarding school into different houses so that they can compete with each other. There are dangers with that—sometimes they come to blows—but the main point is to make people within each house aspire to work hard and take pride in it. The city of Siena is split into around twenty contrade—parts of the city that compete against each other in a huge horse race—as a way of instilling solidarity within each neighborhood. They hate each other and compete fiercely, but it has historically strengthened Siena’s ability to work together as a whole, because it instills a strong civic identity. In terms of aggression, you might think sports work similarly: when you are playing American football, rugby, or even a less violent sport like basketball, that is a way of channeling aggression into something socially non-harmful.

Stewart-Williams: A way of, as Freud would have said, sublimating those potentially destructive desires. I think that is true and interesting. When you study the fact that we are so prone to in-group and out-group biases, evolutionary biology and psychology, rather than saying it is natural and therefore good, can equally well puncture it. You might genuinely think your in-group really is great and that the out-group really is terrible. But if you realize that we have this built-in tendency in human nature to divide the world into in-groups and out-groups, it can make you challenge that. It can make you think: maybe I am just falling prey to this evolved illusion, and it is not actually true that we are superior and angelic and they are genuinely terrible and demonic.

In the rest of this conversation, Yascha and Steve discuss how society should respond to sex differences, why there are more female double threats than male ones, and to what extent certain disciplines being dominated by men are due to sex differences versus discrimination. This part of the conversation is reserved for paying subscribers…

Read more

Jeremiah Johnson is the co-founder of the Center for New Liberalism. He hosts the New Liberal Podcast and writes at Infinite Scroll.

In this week’s conversation, Yascha Mounk and Jeremiah Johnson discuss why young Americans think the economy is worse than it actually is, whether social media has made us permanently pessimistic about institutions, and how elite failures are now exposed in real time.

This transcript has been condensed and lightly edited for clarity.

Yascha Mounk: There’s been an interesting debate over the last few days on social media over why Zoomers are having trouble buying homes—why they’re doing so badly in material terms. One faction was saying, it’s all because they’re spoiled: they go out to lunch too often, they order on DoorDash all the time, and they just have terrible financial habits. The other half of the debate said, no, it’s because of neoliberalism—the economy is in a terrible state, and why are you blaming these poor kids? It’s not at all their fault; it’s our evil economic system. What do you think is actually going on?

Jeremiah Johnson: To set the stage for people who are blessedly offline and not poisoned in the same way that maybe you and I are: Kevin O’Leary—the guy who calls himself Mr. Wonderful and is on TV—gave an interview where he said something along the lines of, it makes me sick when I see some young kid who’s making $70,000 a year spending $28 on lunch. This went viral in the way that these things do, because it’s intergenerational conflict—people yelling at each other about their personal budgets. One group was saying, yeah, that’s ridiculous, $28 for lunch—kind of redoing the whole avocado toast thing that everybody’s aware of. Bizarrely, certain left-leaning commentators jumped in to defend the Zoomers. Sometimes this took on the bizarre quality of: well, lunch just costs $28 now—have you looked at a menu recently? Or: people need to use DoorDash because they’re disabled and they can’t make lunch for themselves—and in this case “disabled” usually means having an anxiety disorder and being unable to talk to people.

Mounk: I’m sure there are some people who are disabled and generally can’t make lunch for themselves, but that is not the average person.

Johnson: That is very much not the average person. On a very basic level, you can join one side or the other: you can say it’s ridiculous to pay $28 for lunch and that you can easily pack a lunch, or you can join the people who say this is not the reason Gen Z can’t buy a home—that saving $5 on lunch is not going to allow you to buy an $800,000 home in a nice metro area. But there’s an implicit assumption in both of these arguments that things are indeed really bad for Gen Z, that Zoomers are suffering, that they’re never going to be able to afford a home, that it’s really bleak out there. People have this sense of doom about the economy.

What I find interesting is that’s just not true, at all, if you look at the data. Gen Z is actually buying homes at an increased rate compared to where millennials were at the same age. Gen Z homeownership is tracking significantly higher than millennial homeownership, and Gen Z is earning more money than any previous generation—not literally at this moment, because they’re all young and early in their careers.

Mounk: The average 25-year-old today makes more money than the average 25-year-old 10, 30, or 50 years ago—and presumably not just in nominal dollar terms. Even adjusting for inflation, that is true.

Johnson: Even in real terms, it is true. This entire narrative around Zoomers being screwed—I just don’t think it’s true. The general economic picture certainly has some downsides. The entry-level job market right now does look a little grim, inflation is higher than we’d like it to be, and there are some things you can point to. But unemployment has been lower than 5% for five straight years now, job growth is good, and GDP growth is good. This is, relatively speaking, a pretty good economy. If you are waiting for some other economy to show up that’s going to blow this one out of the water, I’m sorry to report you’re going to be waiting for a while.

To me, the interesting part of this is not who’s right about the question of the $28 lunch. What’s interesting is: why do we all think that society is doomed, that this generation is doomed, when the numbers and the reality on the ground just don’t bear that out?

Mounk: I’ve been thinking about this in a slightly broader context, and then in the narrow context of liberalism. Perhaps let’s start with the broader context. People now have much less trust in institutions, and they have, generally speaking, a more negative view of the world—at least in North America and Western Europe. The question is: why is that? Is it that institutions have gotten worse? Is it that institutions have always been bad in all kinds of ways, but social media allows us to see those failures much more easily? Even a politician’s most embarrassing moment in the past probably wasn’t caught on camera—or if it was, it would have been played once on the evening news for five seconds. Now, every time a politician posts something on social media, the first comment is going to be a five-second clip of them choking on a pretzel or whatever. We are just much more aware of the weakest moments of people in institutions. Or is it that institutions were pretty good 15 or 30 years ago and are pretty good today, but we now have an unduly negative view of them?

Johnson: There’s some nuance here. Our institutions have absolutely gotten worse, and specifically in the last year under Donald Trump 2.0, we’ve seen a rate of institutional decay that we have not seen in decades, in generations. Our institutions are absolutely being hammered. But I really do think that, broadly speaking, that is for the most part a last year or two kind of phenomenon.

Mounk: But in an interesting way, that is sort of downstream from the previous stuff—which is to say, it’s because people soured so much on the institutions and became so mistrustful of them that Trump was elected for a second time. Now, in fact, in office, he is degrading institutions at a very alarming rate. But in a way, that comes downstream from people having lost faith in those institutions, which is what allowed him to win power again in the first place.

Johnson: What I was going to say is that the things we’re talking about predate the election of Donald Trump for the second time. Consumer sentiment is at the lowest it’s ever been right now, unemployment is something like four and a half percent, and job numbers remain pretty decent. Yet if you ask people, they report lower consumer sentiment than they did during the 70s stagflation, than they did during the Great Recession, than they did during the COVID recession. You can extend this to all sorts of things that are not just the economy. People think that race relations are worse than they were 30 years ago, that gender equality is worse than it was 30 years ago.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

You can look up all the Gallup polling—Gallup does a lot of very good polling over time where they ask the same question for decades, so you can really see how people are answering the same question over time. People think that race relations and gender relations are significantly worse than they were in the 90s, which I feel like would come as news to Rodney King. Certainly race relations have hit some rough spots in the last few years. But the idea that we’re not making progress strikes me as incorrect.

To keep going: when they ask whether people think crime has gotten better or worse in the last year—they’ve asked this for 30 or 40 years straight—in something like 29 out of the last 32 years, people think crime got worse. People think crime is getting worse every single year. The 90s were absolutely a high point for crime, and yet people continually believe it’s worsening, despite the fact that crime has been on a very large decline from the mid-90s until now. It’s not even a comparison.

So in every way—economically, socially, with crime, with everything—people think things are getting worse. I do think that is a function of our media ecosystem and the ways that has changed, and I think social media plays a lot into this. There’s an interesting question of how differently social media shapes our expectations as opposed to the forms of media we used to have—the nightly news, newspapers.

Mounk: To play devil’s advocate: if people have said for 29 out of the last 30 years that crime has gotten worse, that predates social media. So perhaps there is just a general tendency for people on particular topics like crime to see the world very negatively. One interesting thing is that, in the same way people absolutely hate Congress but often like their own congressman, people sometimes think crime is terrible in the world but their own neighborhood is kind of fine. Part of this is that the places people know very well they can often judge relatively accurately, whereas the places that are far away from them they end up having very negative stereotypes about.

But insofar as it’s true that on all these other metrics overall sentiment has gotten worse—and I think the evidence of that is pretty clear—why is that? Is it social media? Is it that our media environment has changed? Or is it that our institutions deliver less for people than they once did?

What’s certainly true is that you had a very rapid improvement in living standards that was really anomalous in human history, and that completely shaped our implicit baseline. When we think about what a normal developed economy or a normal consolidated democracy looks like, we tend to think about the 1960s and 1970s, which were a period of just extraordinary economic growth. The living standards of the average American doubled from 1935 to 1960, and doubled again from 1960 to 1985. At some point it becomes harder to grow at that pace.

Even though, as you’re pointing out, the story is actually much less dire than it’s often believed—I have an article showing that the economic doomerism of the 2010s, where a number of economists argued that inequality was only growing, that global economic growth only benefited the richest, and that wages were stagnating for the middle class in America, was significantly challenged or revised—we tend to think Zoomers are doing worse, when actually they’re doing better than millennials were at the same juncture. So the story is not that negative. But it is true that I don’t feel like my living standard is vastly better than my mom’s was, and I don’t think my kids are likely to have a vastly better living standard than I do. Fifty years ago, people probably did feel that way.

Johnson: There are a couple of books I would point to here. There’s a great book by Robert Gordon, where he examines the nature of economic growth called The Rise and Fall of American Growth. He argues that we had this really anomalous period of wild growth that was more than just the headline GDP number—that it was really impactful stuff. We were electrifying the country, we were inventing new forms of household appliances that eliminated a significant amount of manual labor, we were really revolutionizing a bunch of things. And that the computer revolution, while significant, has not actually changed daily life in quite the same way that plumbing, electricity, appliances, and antibiotics did.

Mounk: Life expectancy shot up in the middle of the 20th century because of inventions like antibiotics and because, for the first time, most people had access to medical care—and it’s hard to replicate that a second time.

Johnson: There’s some argument that our current level of growth is just not as emotionally or viscerally impactful as previous growth was. But I would also point to a second book called Revolt of the Public by Martin Gurri. The thesis of Revolt of the Public is that everyone is mad all of the time about everything, and this will basically be the state of things forever.

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

Gurri is a really interesting figure. He wrote this book essentially before Trump, before Brexit—when he wrote it, he was talking about things like the Arab Spring, Occupy Wall Street, and different protests in Spain and Israel about the general state of society and cost of living. He was looking at this from a roughly 2012, 2013 perspective. His thesis is basically that elites and institutions have always failed to deliver on their promises—that is the human condition, as far back as you can possibly go. But it used to not matter that much, because we had media systems that either protected the elites, covered for them, or made it very hard to organize around those failures. What social media has done is made those failures—which have always been happening—very obvious in real time, and made it very easy to organize around them.

He gives the example of JFK’s invasion of the Bay of Pigs: early in his presidency, JFK tried to invade Cuba using a group of Cuban émigrés, and it turned out to be a complete disaster. A bunch of people got killed, it was hugely embarrassing—and the reaction from the press was basically, that’s too bad, he’s a young president, he’s learning, he’s growing into the role. They were just fully covering for him. Can you imagine the United States trying and failing to invade a country in 2026 and the press simply saying, oops, we’ll get them next time? You can also think about FDR—there was a kind of gentleman’s agreement that nobody was going to talk about the fact that he had polio and could barely stand up. That kind of informal agreement simply doesn’t exist today. Social media makes it impossible.

The result is that people are constantly being bombarded with the failures of elites and institutions, constantly angry about it, and don’t have much of an idea what to do with all that anger. This is kind of the defining feature of a lot of these movements—Occupy Wall Street, the Arab Spring—that they know what they are against but don’t really know what they are for. The Arab Spring was not for anything in particular: some people wanted liberal democracy, some wanted Islamism, some wanted an entirely separate set of things. Occupy Wall Street was united by anger but had no real policy agenda. You can see the same thing when you look at Brexit or Trumpism—these are movements primarily against something, not really unified in what they are for. They just know they’re very angry and that they need to toss out the people who were in charge. I think the media environment has a lot to do with why people are so dissatisfied all the time about everything.

Mounk: Revolt of the Public is a really interesting book—I’ve had Martin on the podcast for a couple of conversations, so if people are interested, they should go and look at those episodes. The thing that is quite persuasive in Martin’s book is the somewhat nihilistic impact this might have. What he calls the periphery—these self-organizing publics, people who come together for some cause very quickly and might also very quickly dissipate—has a really strong ability to take on power and challenge hierarchies, but a really weak ability to actually organize, impose a structure of their own, or coherently put forward an alternative program. What you get is this lurching from one thing to another. There’s a kind of nihilistic, restless element to it. I find that quite persuasive.

In terms of the movements you talked about, I would quibble a little and say there are distinctions between them. Occupy Wall Street was a set of people who had somewhat unified politics for all their differences, but no concrete vision—they wanted to occupy Wall Street and make Wall Street go away, and it’s just not really clear what that means or what they wanted to put in its place. A bunch of them were influenced by a kind of utopian socialist thinking which, compared to where a lot of progressive spaces are today, seems somewhat charmingly quaint. The thinking was that it’s impossible to really make an alternative to the current system within the system, and so it would take a democratic revolutionary process to figure that out—and so somewhat deliberately, they didn’t want to offer that alternative upfront. That really was an example of a movement that just didn’t have a clear alternative in mind.

The Arab Spring was different, because a lot of the people who took part did have very clear visions of what they wanted—just different visions. What they were united around was getting rid of these old, corrupt, moribund dictators who had been dominating their countries for so long. They could build a coalition around that shared goal, but what to put in its place divided them entirely: some genuinely wanted a liberal democracy, others wanted a Muslim Brotherhood-style soft Islamic government, others wanted an even more fundamentalist Islamist government. So it’s not that nobody had an idea of what they wanted—it’s that they all had different ideas, which is a slightly different problem.

I want to connect all of this to the question of liberalism. In a way, these are two separate questions—we’ve been thinking about the legitimacy of institutions, and our institutions are in certain respects liberal but in other respects not perfectly so. But I do think liberalism faces a similar set of problems, because people associate liberalism with the status quo, and because part of our political order has been built by liberalism. A lot of what’s good in our world exists insofar as it actually lives up to liberal principles and precepts. The trouble that liberals have in defending their ideas is, I think, a little bit related to the trouble that institutions have in sustaining respect and approval. Because on social media, whether you’re clobbering liberalism from the left or from the right, it’s very easy to say everything is terrible. The person pointing out that, actually, Zoomers are buying homes at higher rates than previous generations, or that there are some pretty good things in our world compared to any other epoch in human history, just doesn’t have standing in that debate.

Subscribe now

Is that an inherent feature of liberalism as a political ideology, or just of our political order being somewhat liberal and therefore representing the status quo and being mistrusted? Or is it the fault of liberals for not knowing how to argue for their positions on social media? You’re obviously well placed to speak to that question as somebody who has really thought about how to argue for liberal points of view in the daily battle of social media.

Johnson: I don’t think it’s an inherent feature of liberalism that it’s just destined to lose arguments on social media. But I do think there’s an inherent feature of social media that benefits extremist ideologies—that extremism in virtually all of its forms is inherently more viral and more algorithmically beneficial than moderation. Liberalism has not always been the moderate option. There were points in history where to be a liberal was to be at one of the extreme ends of the political spectrum—liberalism was a very revolutionary form of politics. But that’s no longer the case. For all intents and purposes in the West, philosophical liberalism has built virtually all of our institutions, and liberals were the great victors of the 20th century—winning World War II, winning the Cold War. So, they are the ones poised to lose in a new system that basically ensures whatever’s currently in power is likely to be tossed out.

One of the ways I like to think about extremism on the internet is through a non-political example. Say you’re on YouTube, bored, just surfing, and you see something about someone who baked a four-pound cake. Would you click on that? Maybe not—it’s not that interesting. What about a 40-pound cake? What about the world’s largest cake, a Guinness World Record 400-pound cake? You’re much more likely to click on that. This holds true for basically any form of content.

The world’s most popular social media influencer is a guy named Mr. Beast, a YouTuber whose whole thing is that he just does really, really big things. His first viral video came when he got $10,000 from his first big brand deal and said, I’m just going to give this to a random homeless person on the street. It went insanely viral. So in the next video he gave $20,000, then $50,000, and then he just bought a house and gave it to someone on the street. He just kept going bigger, and he became the world’s biggest influencer—he has something like half a billion followers on YouTube and is about to hit 500 million. His entire ethos is: never go small when you could go big.

The internet has done this in a lot of ways. If you’re a spicy food enthusiast before the internet, you might know one or two restaurants with really spicy food, or subscribe to some niche magazine, or order a really spicy sauce by mail. These days you can go online and see people breeding the Carolina Reaper crossed with something that registers 10 trillion Scoville units. You can literally chart the spiciness of the hottest pepper over time, and it’s gone off the charts. This is just what the internet does—it pushes everything to its extremes. It’s doing that to our politics too.

Which political message do you think goes more viral? One that says the world is burning and we’re all going to die in the next 30 years from climate change, and the oligarchs have screwed us, and the reason you can’t get ahead is all these greedy people looting the world while it burns to the ground? Or one that says climate change is a serious problem and some people are certainly going to be harmed by it, but if we have the right policies we can mitigate that harm, and society will probably be fine in the end, and economic growth will probably continue? The first one goes incredibly viral. The other one gets maybe seven total likes. This is just a feature of how algorithms work—things that inspire a response inherently go more viral.

Mounk: How do you think about pushing that reasonable message in a way that is more exciting? You started a bunch of accounts called Neoliberalism and other things—and I don’t want to presume your opinion about climate change, but I’m guessing that the boring message is a message you in fact believe in. So are we just doomed? Is that message just never going to cut through? You don’t seem to think it’s doomed, because you seem to think it is in fact worth engaging on Twitter, on Reddit, on all kinds of other platforms, trying to push out that more middle-ground message if you believe it to be true. Is it a question of repackaging it? Is it engaging in the fight knowing you’re always going to be at a disadvantage and will probably lose, but it’s worth at least trying? What’s your rationale, and how do you go about doing that?

Johnson: I think we’re always going to be at a structural disadvantage—and by “we” I don’t necessarily even mean liberals, but just anyone who tries to offer nuance and depth and reasonableness, who wants to provide a message that plays towards factual reality rather than emotional reality. There’s always going to be a structural disadvantage. It doesn’t mean that you can’t win, but it does mean that you’re starting at a disadvantage. So you have to take certain steps.

The reason this all exists is because social media rewards the opposite of nuance—it rewards really quick reactions and thoughts. Sometimes I like to think back to what the media was like in a previous era. I think back to the Lincoln-Douglas debates, to go all the way back to pre-Civil War. This was the hottest thing happening in 1858—these debates were national news, the front page of every single newspaper. You know how they were structured: one man would give an hour-long speech uninterrupted, the second would give a 90-minute rebuttal, and then the first would come back for a 30-minute closing. This was to packed venues—people’s attention spans were such that they were riveted by it. Today, our political discourse has devolved from that to the Buckley-Chomsky debates, to CNN Crossfire, to just 10-second out-of-context clips on social media. The age of nuance is dead, to some extent.

But there are some things you can do. You can create these little islands of quality. The fact that social media is structurally against nuance doesn’t mean that nuance will never survive anywhere—there are still going to be people who care about it. You can create communities of people that actually want to understand the world in a realistic way, and I think those communities are what allows you to get your message out. You have to build a group of people that cares about something, whether that’s in a subreddit, a Discord server, a Substack newsletter, or a podcast like this—people who actually care about the world as it really exists and about nuanced, complicated, complex questions that can’t be answered in 10 seconds. If you can get those people, keep them, and convince them that they’re not just individuals but part of a group, that this is part of their identity, that’s really powerful. That’s part of what we’re trying to do at the Center for New Liberalism—build an identity around that kind of politics. Because identity is really powerful.

Mounk: There are a few slightly different questions here that feel identical but are important to keep distinct. One question is: are the extremes—the people who are angry, the loudest, the least subtle—always going to dominate certain social media platforms, particularly the ones that are especially political? That seems likely. It seems likely to me that Twitter, by its nature, is always going to be dominated by the most extreme people.

The question is also what “winning” and “having an advantage” actually mean. The most viral tweet is often going to be pretty extreme—but that’s different from whether the tweet actually persuades and convinces people. Part of it is that the people who are on Twitter exposed to political content are already quite a small percentage of the population. Most people seek out social media, but not primarily for political content—they might encounter political content every now and again, but the people whose Twitter feed is dominated by political tweets rather than sports or gossip are a very self-selected minority. That’s an important thing to bear in mind.

The other thing I was wondering about is whether there might not at some point be a kind of course correction in norms and expectations. There was a time when it was enough for three or four people to claim they were terribly offended by some exhibit, play, movie, or speech for that to go super viral, for The New York Times to write up the controversy in very neutral terms, and for the thing to get canceled. There was a time when you could post a 10-second video of somebody seeming to misbehave without any context and everybody would credulously jump on the bandwagon and say, that must be a terrible human being. Eventually people kind of got the message—they realized that this was just inviting conflict entrepreneurs to gin up completely spurious cancellation attempts, where a lot of the time the 10-second video came from a five-minute context in which somebody was being antagonized for four minutes and 50 seconds, and all that gets captured is the 10 seconds where they can’t take it anymore. Nowadays, quite often when people post that kind of content, the median response is not credulous outrage but: why is it only 10 seconds? What about the rest of it? People have become a little bit smarter about it, and the appetite for that kind of social media drama has somewhat abated. I don’t want to overstate the point—there’s still a lot of social media nonsense—but it does feel like people have gained some kind of literacy.

So I wonder whether at some point something similar is going to happen with extreme political views and statements—whether at some point people are going to say, here’s another person breathlessly making this exaggerated claim, and just roll their eyes or push back. Whether we might over time learn to adapt in ways that lessen that advantage. Is that overly optimistic?

Johnson: I think it might be, to be honest. Unless we change the way that algorithmic feeds are constructed—the way TikTok operates, Instagram Reels, Twitter, and Facebook, which is still very influential among older generations—unless we literally change the nature of algorithms, the saying holds: a lie can get halfway around the world while the truth is still putting on its shoes. We know scientifically that negative information is more viral than positive information. So in terms of just the way our information systems are set up, we’re always going to be operating at a disadvantage. We might still win certain battles, but we’re going to be the underdogs a lot.

Where I am a little more optimistic is that I think we can’t spend the entire time thinking about this problem just in terms of messaging, because the other thing that actually influences how people feel is reality on the ground. When you do enough good for a long enough time, people do notice. I’ve been talking about how things are good but people believe they’re bad—but even in that context, if you do enough, people start to notice that their lives are improving. The stimulus checks during the pandemic were very, very popular. Now, did they help Joe Biden get reelected? No, because they went out three and a half years before the reelection campaign. But I don’t want to make the case so strong that material reality just doesn’t matter anymore, because it still does matter when you don’t deliver for people.

Trump was reelected largely on the back of cost of living, and he is making the cost of living crisis worse. He started a war with Iran, so gas prices are spiking. He’s implementing tariffs all over the place, so consumer products are more expensive. People notice, and he’s less popular than he’s ever been. The hope, obviously, is that if we can elect someone with better, liberal principles to succeed him and reverse some of the damage, people will actually notice that we’ve started fixing problems—and that leads to a virtuous cycle rather than the damaging cycle that we’re in now.

Mounk: In terms of actually being present on social media and making the case—give us some examples of how you thought about making the case for philosophically liberal principles on social media. What was the goal? It sounds like the goal wasn’t necessarily to have the most viral tweets, but to have enough presence to create a tribe of people. And I don’t know that I’d recommend to any podcast listeners to engage actively on social media, but if they want to engage and be active, how should we think about what they can do to put forward reasonable, moderate points of view—without getting into a polarization spiral, without getting slowly incentivized to make more and more extreme statements? What should the goal be, and what are the means?

Johnson: One of the ways I like to think about this is that social media breeds extremism and conflict. Conflict drives engagement, engagement metrics get noticed by the algorithm, and the algorithm sends more people to the conflict. Social media is just built structurally to encourage conflict. You can think about it as roving groups of tribally aligned people who yell at each other all the time—and that’s not necessarily a bad model of social media, whether it’s political, whether it’s pop stan armies like the Sabrina Carpenter people yelling at the Taylor Swift people, or how sports fandoms work online.

But there’s another way to think about it. You’ve got all these tribal identities roaming around—K-pop stans, Oklahoma City Thunder fans, liberals versus socialists versus MAGAs—and one thing you can do is just make your tribe an attractive place to be. You can cultivate your garden, so to speak, rather than sending people out to fight all the time. Create spaces that are informative, friendly, welcoming, and fun—and fun is an important part of it. People have to want to be there. If you can do that, you’ll see those spaces grow, because people want to spend time talking about the things they care about in a place where they’re not going to get yelled at, with people who might agree with them, teach them something, or just give them a good time. Inside jokes and iconography are an important part of this.

Weirdly, a lot of this is the same way that religions form—and I don’t think that’s necessarily a bad analogy, because people believe in their current political ideologies with a very religious fervor. But if you want to build a community, it needs to have visuals, it needs to have iconography, it should have its patron saints. If you are a socialist, that’s probably Marx, Lenin and Mao. If you are a liberal, maybe that’s Adam Smith and Milton Friedman and Ben Bernanke. The early days of the neoliberal movement were punctuated by doing lots of memes with Ben Bernanke and Janet Yellen with lasers coming out of their eyes defeating the populists. It was just a bunch of silly fun—but that’s the kind of thing that makes people happy, makes them come back, and makes them start to identify not just as someone who agrees with a set of ideas, but as one of these people. This is who I am, rather than just this is a thing I kind of believe. That’s a very powerful thing to create.

So the model of building a garden, tending your garden, expanding your garden—building a community that actually reflects the values you want to see in the world—is a very powerful thing. You can see this in movements like Effective Altruism, the New Liberals that I’ve been a part of, and in the way the left built power through the 90s and 2000s before the Bernie Sanders explosion. Getting beyond just fighting all the time and building something internally is one of the ways we can short-circuit the cycle we’ve been talking about.

In the rest of this conversation, Yascha and Jeremiah discuss the new liberal ecosystem and how to build a lasting online community. This part of the conversation is reserved for paying subscribers…

Read more

Kathryn Paige Harden is a professor in the Department of Psychology at the University of Texas at Austin, where she leads the Developmental Behavior Genetics lab and serves as Director of Clinical Training. Her latest book is Original Sin: On the Genetics of Vice, the Problem of Blame, and the Future of Forgiveness.

In this week’s conversation, Yascha Mounk and Paige Harden discuss why twin studies reveal the substantial influence of genetics on human behavior, how genetic effects actually increase rather than decrease over a person’s lifetime, and why acknowledging genetic influences shouldn’t be seen as incompatible with progressive politics.

This transcript has been condensed and lightly edited for clarity.

Yascha Mounk: Thinking about genetics, and particularly the genetic influence on human behavior, is, for lack of a better word, often right-coded. People assume that it’s people on the political right who want to emphasize the genetic determinants of human behavior, and it’s often people on the left who are very allergic to this, who say that genetics doesn’t really explain anything—it’s social forces. Somehow, if we accept that genetics influences human behavior a lot, then that would lead us down all kinds of dark alleyways.

You see yourself, and think of yourself, as a woman of the left. You also think that we have, in many areas of our life, underestimated the role that genetics plays in explaining human behavior. Let’s start with the latter part of this: why should we think that genetics is such a big determinant of human behavior?

Kathryn Paige Harden: We should think about the importance of genetics because we have a lot of scientific evidence that genes matter in shaping our lives. We can see that in twin studies, which were the workhorse of genetic science in the 20th century: how similar are genetic relatives, even if they haven’t been raised together, for certain aspects of their behavior? We see it now with more modern technology, where we can actually look at specific segments of DNA, how they differ between family members, and are beginning to discover how they help shape behavior.

It’s very rare that genes are determining behavior, but my PhD advisor from graduate school is very famous for a paper called “The Three Laws of Behavior Genetics.” The third law of behavior genetics is: everything is heritable—which means that everything that differs between people, when you look at it scientifically, tends to show some influence of the genes that you were born with. I’m a scientist; I’ve been working in this field since I was 18. It’s important to me because I think it’s a scientific fact that is incontrovertible at this point: genes influence behavior.

Mounk: Let’s walk people through these arguments step by step, because I think a lot of people have heard about twin studies and they know that nowadays there’s probably more modern techniques that are slightly different. What exactly do the twin studies show and establish?

Harden: Twin studies typically compare identical twins—who were conceived from one egg and one sperm, and then there was an error in early cell duplication so that now there are two people—to fraternal twins, who are the sort of twins who look like ordinary siblings: two sperm, two eggs, who just happen to share a pregnancy. That logic can be expanded to compare full siblings to half siblings, or adoptive parents and children to biological parents and children.

What all these designs have in common is that they take people who share a social relationship. Both fraternal twins and identical twins have the same social relationship as siblings being raised in the same home with the same parents, but have different genetic relationships.

Mounk: The idea here is to try to hold as many things constant as possible, other than genetics. You’re going to assume that perhaps there are differences between siblings—perhaps the family became a lot richer over the course of three or four years, and so if one sibling is three years younger, you might say it’s because of that. But if you’re comparing identical twins to fraternal twins, there’s no three years in between for the family to become richer.

Harden: Yes, and it’s the same pregnancy—the same maternal smoking, the same neighborhood they happen to live in, the same prenatal supplements that the mother was taking. The only difference between them, in theory—and we can get into whether this assumption holds—is the difference in their genetic relationship. You’re then trying to see: does the similarity in their behavior track the similarity in their genetic relationship, or does it track the similarity in their social relationship?

One of my favorite studies looks at twins where the mother is wrong about whether the twins are identical or fraternal. The mother thinks they’re fraternal because they look different to her—she knows them so well. But when you look at their DNA, it’s clear that they’re identical. The question is: do they look as similar in their behavior as identical twins typically do, or as fraternal twins typically do?

Harden: The study shows that behavior tracks the actual genetic relationship, not what the mother thinks the twins are. It’s not explainable by being raised in the same home, and it’s not explainable by parental expectations of how similar they would be. The best predictor of their similarity is how many genes they share.

I want to go back to what you said at the beginning: is that result right-coded or left-coded? It really depends on what you’re talking about. Some of the earliest twin studies were done around schizophrenia, showing that the best predictor of having schizophrenia is having an identical twin who also has schizophrenia. Subsequent twin studies have shown that you also see this higher similarity in identical twins for addiction, for sexual orientation, for depression, for all sorts of mental health problems.

Mounk: That might be left-coded because it implies that it’s not your fault.

Harden: If we look at whether people with more right-wing political ideologies versus left-wing are more likely to believe twin studies, or more likely to believe that genetics influences behavior, it very much depends on what behavior you’re talking about. Liberals are much more likely to say yes, identical twins are more similar for addiction; yes, sexual orientation is partly shaped by the genes that you inherit. It is the application of this to specifically cognitive ability and achievement that tends to be more right-coded in the United States.

Mounk: Tell us a little bit about the magnitude of these findings. When you compare identical twins to fraternal twins, give us some statistic or concrete measure that allows people to get a sense of how much variation there is between identical twins—obviously there’s still some variation. How much more variation is there once you get to fraternal twins? How does that compare to a half sibling, for example?

Harden: There was a really big meta-analysis where they pooled data from 50 years of twin studies—something like 15,000 studies, millions of pairs of twins. The punchline was that on average, across behavior but also medical traits, about half of the variation in outcomes was due to genetic differences between people. One way to think about that is: if people were all genetically the same, how much could you make outcomes the same? You could shrink people’s differences by a substantial degree if they were all genetically identical.

It varies by trait. For things like schizophrenia or height, in modern populations almost all of the variation between people is due to their genetics. For things like personality or depression, it’s about half. When you start talking about education, it’s a fifth to a third. It depends on the phenotype you’re looking at. It also depends on the population you’re studying. If you’re looking at a nutritionally deprived sample—people who didn’t get enough to eat in childhood—you’re going to see less of a role for genes than if you’re looking at people who all had ample access to food, where most of the differences between them are necessarily about their genetics because there isn’t that much environmental variation.

Mounk: The average height in China has shot up over the last 30 years. There’s a really nice example of that in a writer who was in Sichuan teaching in the 1990s. He’s a little shorter than me—like five-seven or something—and he was the tallest person in the classroom. He went back a few years ago and suddenly he was one of the shorter people in the classroom, and he was really struck by that. Obviously it’s not that the genetics of people in China has completely transformed over the course of 30 years—it’s that they’ve overcome the nutritional disadvantage they had as recently as the 1990s.

If you’re in a place where a family doesn’t have enough money to buy enough food for all children, and one child for one reason or another is favored by the parents, or is better at elbowing their siblings out of the way to get to the food, they might end up being much taller—so it’s not genetically determined, it’s determined by social factors. But once you look at a context where everybody gets enough to eat, genetics is the thing that really drives it.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

The other interesting finding here is that you might think genetics plays the biggest role early in life, and that later on its influence fades as people’s paths diverge, as they make different choices. If I’m understanding it right, the actual finding is the other way around?

Harden: What you see is that twins get more similar over time, and people who don’t share genes get more different over time, such that heritability estimates tend to go up with age rather than down. If you’re looking at the similarity between adopted children and their parents, or adopted children and their siblings, they are most similar to these non-genetic relatives in childhood. Then, as they leave the family home and have more control over their environment, the evidence of the genotype tends to emerge over time.

One way to think about that is that, especially in modern contexts, people have increasing amounts of control over their environments—both conscious and unconscious—as they go through life. They pick their peers, they pick their partners, they pick their daily routines, they pick what city they want to live in. Each of those choices reflects, even in small part, their own personality and temperament, and so there’s going to be a feedback loop where the environment is reinforcing those initial genetic differences. That sense of becoming more like your identical twin, and less like people who are genetically different from you, increases over age rather than decreases.

Mounk : What are the main responses from people who are skeptical of the genetic explanation? Perhaps let’s start with popular objections and then go through some of the most sophisticated scientific objections. I should say that my wonderful podcast producer, Leo, is messaging me in all caps over WhatsApp because she has an identical twin and keeps saying there’s no way they’re going to get more similar as they get older. I think for a lot of people, a lot of their sense of self is bound up in many of these questions.

Harden: There are a couple of objections. One is, I think, just a distaste for even doing the research—and this may be piggybacking on the comment you just made, that people’s identities are bound up with their behavior. It can feel inherently reductionistic, or inherently threatening to human agency, to make really valued parts of people’s identity—or very problematic or darker parts of people’s behavior—the object of scientific study, and especially scientific study at the level of genetic analysis. It collapses across our sense of ourselves as embodied animals and our sense of ourselves as choosing, agentic wills or reasoners. One objection is simply: it gives me the ick to think about genes influencing my sexual identity, my religious activity—that can just be uncomfortable.

Another objection is that regardless of the results, and regardless of how the results are presented, there’s the fear that it is necessarily giving ammunition to the most reactionary, eugenic, or racist elements within a society. There’s a history of people using genetics to justify really horrible things. The objection is: even if you’re scientifically correct, there’s no way to do this work responsibly in our current political climate.

Then there’s the scientific objection, part of which is very correct: twin studies make a ton of assumptions. You’re taking a really complicated developmental process and flattening it into one bucket for genes and another bucket for environments—but genes and environments don’t work like that. They’re always combining over the course of development. So what does it mean to have a statistical model that tries to disentangle them? Often people’s objections can float between these. They’re animated by fear, and that makes them perseverate on some of the assumptions of twin studies. These are not mutually exclusive—they’re fluid objections.

Mounk: It is strange that this is an area in which people have very strong moral convictions, and it’s not always clear what drives them. There are a lot of people who feel a kind of sense of threat from recognizing some of this genetic influence. I’m thinking of one very dear friend of mine—an avid listener to this podcast, so she’ll recognize herself being mentioned here—who for a very long time believed strongly that the most important determinant of behavior is cultural: that the way to understand behavior is to study sociology and anthropology and the cultural determinants of how people behave. Obviously, if you spend time in Italy and you spend time in England, those places shape people in deep ways, and that’s not a genetic difference between Italians and English people by and large—it is a difference in culture, norms, and what personality traits those cultures encourage. There’s a real pull to that view.

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

She now has two children, and when I was visiting her, she looked at me and at her kids and said: “Yascha, I was totally wrong. The difference between these kids—it’s just clearly like they come to the world with their own personalities and attitudes.”

I wonder whether there’s a slightly mistaken assumption in the background about what it would take to be responsible for our personalities—as if, on the one hand, if my personality is determined by my genes then I have no responsibility for it and it’s out of my hands, and on the other hand, there’s some idea of having totally chosen my own personality, and only then can I be responsible for it and be my own person. That background assumption may simply be wrong. We may just have to accept that if I like somebody as a friend, it’s because of the traits they have, and it’s perfectly appropriate to be grateful to them for their generosity, or to decide that I don’t want someone in my life because they’re mean and not a very good person. Whether or not behavior is in part genetically determined is, in that respect, kind of irrelevant. I have a friendship, a social relationship with the person in my life, and the question of how they came to be that person is not nearly as directly relevant to my gratitude, my anger, or my relationship with the people in my life as many people seem to assume.

Harden: What you’re getting at is just really core to the dilemma of being a person—and I write about this a lot in my new book. We go through our lives and we experience ourselves in relationship with other people. We feel certain things. We have resentments, gratitudes, affections. We’re drawn to certain things and repulsed by others. At the same time, we have this peculiarly human gift and burden of being able to reflect on all of that—and to reflect in this really peculiar way we call scientific, which is to measure it, to put numbers on it, to try to develop theories about what predicts it: in terms of culture, genetics, parental socialization, life events. Behavior genetics sits right at the intersection of our experience of ourselves as loving, behaving, feeling people, and our existence as self-conscious minds that turn ourselves into objects of scientific study.

I don’t think it’s really about someone having assumptions that are wrong versus right. Of course you’re going to come to learning about the science with priors, because you’ve been a human. You’ve had a sense of yourself, and you’ve necessarily had to develop a story about why you are the way you are. Science, on the one hand—and this is fading—has carried a kind of cultural authority. So people so often experience scientists saying “here’s one lens on behavior” as if it were “this is the only lens” or “this is the right lens.” If that conflicts with the narrative you’ve developed about your life, of course that’s going to bring up feelings.

I talk about this with my undergraduates all the time. I teach intro psych here at UT, and I say at the beginning: psychology is not going to be like your calculus class. A professor going into calculus can assume you have no prior knowledge of calculus. Whereas I’m going to be talking about attachment and personality and parental relationships—and you already have thoughts and feelings about all of that, because you’ve been a person. I’m not trying to say that your theory is necessarily wrong. I’m trying to add a layer of information on top of that to refine some of those priors. It’s necessarily a topic that people have feelings about, because the question of why you are the way you are is one you can’t get through life without having constructed a story about, one way or the other.

Mounk: How should we think about a few key questions? I had Emily Oster on the podcast perhaps about a year ago. One way in which she reads the influence of genetics—but also the influence of some basic socioeconomic facts about you that are not really under your control—is to say that parents hugely overestimate the impact of little decisions they make. Do I take the kid to this ballet class or not? Do I take them to this slightly marginally better school, even if one school is much further away? If I allow them to have 20 minutes more screen time, is that going to deplete their ability to pay attention forever? By and large, you would also say: you’re overthinking it. As long as you make sure they have enough food to eat, and they feel loved and safe, and they don’t experience some horrible form of trauma as a child, some mixture of genetics and the very basic facts about their environment is going to determine the vast majority of a life outcome. Torturing yourself over all of these marginal decisions is just a mistake. Would you broadly agree with that? Do you think that’s one of the upshots of this kind of research?

Harden: I broadly agree with that. A lot of things that especially middle-to-upper-middle-class mothers—these are the moms in my cohort—obsess about are just not going to make that much of a difference in terms of measurable child outcomes. They’re not going to meaningfully change your child’s IQ score, or likelihood of becoming depressed, or likelihood of graduating from college, based on whether they go to a Montessori preschool versus a Waldorf preschool.

I would refine that conclusion in two ways. One is that parenting is a relationship, and you do things in relationships not because they’re going to change the other person, but because they seem to be the best thing for how you’re relating to that person in the moment. It would be really strange if someone said it doesn’t matter how you treat your husband because you’re not going to change his IQ. That’s not the point of behaving toward your husband in a certain way—it’s to live out your values and to treat him in a way that he deserves and wants to be treated. A lot of parenting is about that.

Subscribe now

I switched my oldest kid’s school this year from public school to private school, and I didn’t do that because I think going to private school is going to change his cognitive ability at 30. It was because he was unhappy at his public school and he’s happier at his new school. The reduction of parenting into can we optimize on these measurable variables—rather than asking what relationship you’re in with your kid, and what makes them happy this year, even if it doesn’t have a long-term effect—I think that can be missing from some of these conversations.

Mounk: You’re presenting that partially as a refinement—or perhaps a small objection—to Oster’s outlook. I would put it the other way around: if you’re obsessed with how every little thing you do is going to impact the IQ score at age 30, or the likelihood of becoming a successful trader at Goldman, or a world-class pianist, you’re not in fact focusing on the relationship in the moment. If you’re freed from that anxiety, and recognize that these choices you’re making are not going to make or break the kid or determine all of those things, it actually frees you up to think more about what your child needs, how to create a family culture where everybody’s thriving. You can actually focus more on that really important question.

Harden: I would agree with that reframing. I talk about this a lot in my second book: genetics can be both liberating and horrifying in equal measure. Thinking about how much your child’s life might be shaped by the role of the genetic dice does confront you with the limits of maternal control. I do not think of my kids as lumps of clay that I have unlimited control to mold, and that frees me a lot from the burden of feeling responsible for everything about them. But it also means confronting the fact that bad things might happen to them, that they might do bad things, and that I have limited control over how much I can prevent or shape that outcome.

Going back to our earlier point, I think that can be part of the implicit negative reaction to genetics: we love control, and thinking about this factor that’s not fully within our control can be, again, both liberating and horrifying in equal measure.

Mounk: Control is a burden, but it is also reassuring. In a very different context, I’ve made this point a few times about conspiracy theories. Conspiracy theories are scary because these evil people hold all the strings and are making the world go wrong in all kinds of ways—but it’s also comforting, because there’s a group of people who have control over the world, and if only we uncover them and replace them, everything is going to be fine. In a strange way, the idea that there is a locus of control somewhere always has this double-edged nature. As a parent, if you have control—well, that’s horrible, because everything depends on you. But it’s also very flattering; it’s a form of ego extension.

Harden: Yes. With the conspiracy theory, one factor has all of the control, and there is something comforting about the cognitive closure of being able to say: it’s this. Not that the world is a messy, complex place where everything is interacting in ways we don’t fully understand, and our ability to steer the ship is always half guesswork—but rather, if we can just get rid of the deep state, everything is going to be better. When you get into “it’s all genetics” or “it’s all culture” or “it’s all capitalism,” that can be comforting because one problem offers the possibility of one solution. Whereas what we’re saying here is that it’s nature and nurture and culture and economies in combination—and knowing where to pull on that sweater is hard. That’s a lot less comforting.

Mounk: Let’s go a little bit further down on this point of education. What do you think drives educational outcomes and success? How do you think that an unwillingness to take the genetic component in educational outcomes seriously has shaped and perhaps misshaped educational policy?

Harden: If we look at the United States in the last 120 years, education has gone up—more people are educated, more people can read, more people can do advanced math. That is not because genetics have changed. It is because we have invested in mass public education and continue to make it a cultural priority to educate every successive cohort of students. The most important thing about education level is that it’s a cultural practice that requires the commitment of adults to educate the next generation. I’m always amazed when I’m helping my kids with their math—my 13-year-old is learning algebra, something that took humanity tens of thousands of years to develop, and now we teach it in a year. That’s an amazing cultural accomplishment.

Within that system, within any cohort of kids moving through an educational system, it’s obvious that kids raised by high-income, more educated parents go further in school. Also, kids who happen to inherit a combination of genes that makes going through school easier for them also get further in school. Those associations—whether we’re looking at a correlation with a genetic indicator or a correlation with socioeconomic status—are of comparable magnitude.

This is not surprising to the average American. If you ask the average American to estimate how important genetics is for education, their estimate is roughly what the meta-analytic twin study estimate is. People have been in classrooms, they’ve been around kids, they’ve noticed their own kids. It is not saying anything revolutionary to observe that some kids have an easier time paying attention, remembering things, manipulating information, and learning new vocabulary words, and that our education system is cumulative and stratified such that those initial advantages compound over the course of education. Who is most likely to graduate from college? Who is most likely to take calculus in high school? Who is most likely to be in the advanced math class in seventh grade? It’s not just the richest kids—it’s also the kids who have a certain combination of genetic variants.

It’s not radical in that sense, but it is radical from the perspective of research on education, because almost none of it considers genetic differences between people.

Mounk: How would this change the consensus in that field among researchers who don’t look at genetics? I can come up with a couple of hypotheses. One is that it may imply the system is somewhat less unjust than it appears. Presumably there’s some genetic correlation between a parent’s academic achievement and a child’s academic achievement, if there’s a genetic component—not a perfect correlation, but some correlation. If you’re only looking at the extent to which socioeconomic or educational background shapes outcomes, you say: the children of college graduates go to college at much higher rates, because if you don’t have parents who went to college, you just never have a chance—you don’t know how to apply, you didn’t go to the right high school, you didn’t read books when you were six years old.

Presumably socioeconomic status does play some causal role. But once you put some kind of genetic factor into the analysis, and it turns out that some percentage of that difference is predicted by genes, that implies that a six-year-old who doesn’t come from a household with those socioeconomic achievements or that educational status actually has a much better chance of making their way than we expected. So interventions designed to make up for socioeconomic disadvantage may not make as much of a difference as we thought.

Harden: A couple of things there. Empirically, it’s kind of worked out the opposite: the research that has controlled for genetics has actually given us some of the strongest evidence of continuing socioeconomic disparities in education. I was just reading a new preprint yesterday where researchers were looking at children at the very top of the distribution of a polygenic score—and we can get into how this is constructed, but it’s a genetic marker indicating whether you have genetic variants statistically associated with going further in school. They were looking at the top end of this genetic distribution for kids from lower-income and higher-income families, and showing that even at the very top of this genetic distribution, there were still big differences in educational outcomes by family income that accrued over the course of education.

I mention that study because it’s a really common fear that if we start to take genetics into account, it’s going to make our evidence for disparities disappear—disparities that need to be addressed with social policy. But in actuality, the studies that have taken genetics into account have made a really strong case for the fact that there are still family economic disparities in educational opportunity in the United States. So empirically, that’s not what we’re seeing.

Mounk: To be clear, empirically, what you’re saying is that the impact of socioeconomic status doesn’t disappear. But it does become smaller when you take into account the genetics.

Harden: It doesn’t disappear. I don’t want to misrepresent the study off the top of my head, so I don’t know exactly how that coefficient changes. But without accounting for genetics, you’re always open to the objection: well, this looks like a socioeconomic effect, but it’s really just differences in kids’ ability.

Mounk: What really frustrates me in these discussions is this: what we actually want to do is give people opportunity. There’s an ideological point of view that says, I care about the outcome, I want everybody to have good opportunity—and then, in service of that, denies the scientific evidence, or feels uncomfortable even engaging with the academic evidence, because engaging with it might undermine the mechanism to which one is committed. But the point of this information is that it might show that the mechanism isn’t in fact going to work. If what is driving disparities were not socioeconomic status at all, then all of the interventions meant to compensate for socioeconomic status are not going to produce more opportunity.

As you’re pointing out, it seems unrealistic that socioeconomic status doesn’t play some independent causal role—and the study seems to confirm that it does. But there is this strange dynamic where people oversimplify their causal model because they care about the outcome, and in doing so actually obstruct their ability to influence that outcome in the way they hope.

Harden: One thing I sometimes say when talking to education researchers is: imagine that someone has snuck into every data set you have and deleted all the information about student socioeconomic status. You don’t know which kids come from poor families and you don’t know which kids come from rich families. You’re still tasked with figuring out which schools are doing the best job of improving student learning. How much harder would your job be?

Your job would be harder because the classic problem in education research is how to disentangle selection—what’s being driven by the student—from what schools and teachers are actually doing. One way to get at that problem is not simply comparing a kid from a rich family to a kid from a poor family without taking that difference into account. And that is the current situation that most education research is in with regards to genetics. If you’re not collecting that information, if you’re not considering it, you’re still trying to disentangle which schools are working, which curricula are working, which teacher practices are working, which policies make a difference. You still have to grapple with the problem of what schools are doing versus what students are selected into those schools—and you have a whole piece of information that’s just not there.

It would be unimaginable to conduct this research without family income, and yet we consider it a matter of course to conduct it without genetics. Going back to your point—do you want to know how the world works or not? If we think that information about what causes what in the actual world is helpful for designing policy, then why would you leave data on the table that could help you understand student developmental outcomes? That doesn’t make sense if you’re truly invested in the ends you claim to care about.

Mounk: There are areas where this dynamic plays out that have nothing to do with genetics. One example is the female wage gap, which is often presented as women earning less money for doing the exact same roles. That is not what the evidence shows—they might earn around 2% less for doing the actual same roles at the same level of seniority. The female wage gap is much larger than that in most countries, actually larger in continental Europe than in the United States. The reason has to do with other factors. If you become a lawyer, the key period in which it’s decided whether you make partner—and make enormous amounts of money—is in your 30s. In academia, it’s decided whether you make tenure, often also in your 30s. If you have children during those years, that reduces your likelihood of making partner or getting tenure, and puts you on a completely different earnings trajectory. The reason female lawyers make less money is not that female partners at the most prestigious law firms make less than male partners—it’s that fewer capable women end up as partners at those firms because of those circumstances.

That doesn’t make it just or fair, but it does mean that if you’re trying to address this problem with anti-discrimination laws, negotiation training, or cultural messaging, you’re not going to get to the root of it. The problem actually requires a more fundamental intervention—perhaps in academia, for instance, there shouldn’t be an up-or-out tenure track where if you make it there’s very little advancement further down the line, and if you don’t, you can never really break in. A more radical intervention may be required, but you can only understand the nature of the intervention once you understand the nature of the mechanism. Yet some people find it somehow counterproductive to complicate the picture, fearing it will make people less committed to addressing the problem. In the process, they obscure the problem in such a way that the measures taken to address it simply can’t work.

Harden: This dynamic you’re picking up on—where a mechanism, a causal story, or a proposed policy is offered as a solution to something people see as genuinely unjust—can become a stand-in for care about the problem, or the perceived injustice, in and of itself. When you say the causal model is different, or the mediating pathway is different, or that the policy didn’t work, it can be perceived as saying there’s no problem to be solved, or no injustice to be addressed. How to avoid our means being conflated with our ends is a really difficult science communication problem, and one where people end up talking past each other.

My friend Jennifer Doleac has done work on ban-the-box—the policy of banning employers from asking whether applicants have an incarceration history—which was a very popular measure. It turns out that it tends to increase job discrimination against the formerly incarcerated, and also against Black men generally, because in the absence of information people simply rely on stereotypes.

Mounk: If you have a racist employer who stereotypes Black men as likely to have been to jail, but you can tell them whether or not an applicant has a criminal record, they can at least reason: this applicant doesn’t have a criminal record, so I’m safe hiring them. If you remove that information—and I’m not approving of that reasoning, but it’s what you would expect given those incentives and background beliefs—then they’re much more likely to think: who knows whether this person has a criminal record, better hire someone else.

Harden: The pushback Jennifer Doleac has gotten is telling—she’s been accused of not caring about second chances for people with an incarceration history. Her response is: she does care, and that’s precisely why she wants to know what actually works to get them employed, not what we think works. I feel similarly about education. If we think a certain math curriculum is going to help people get to Algebra 2, and we care about that as a society, but then you look and see that one school is only performing better than another because it has a higher concentration of kids who’ve inherited genes that make them slightly better at manipulating nonverbal information—I want to know that.

Part of this is being attuned to opportunity costs: anything we do that doesn’t work, and anything that makes it slower to identify what doesn’t work, is not free. That’s time and opportunity that could be used more effectively. There’s a real pragmatic case to be made for incorporating genetics—not because it’s a silver bullet, but because it’s another variable. You wouldn’t leave a variable of comparable statistical power off the table if it weren’t genetics. So why not use this one too?

Mounk: We’ve covered a lot of ground, much of it terrain from your last book. In your new book—in case the last one wasn’t controversial enough—you go for the maximally difficult part of this discussion: how do we think about the problem of evil? How do we think about people who do really bad things, the genetic determinants of that, and how should it influence our social response? Perhaps again, let’s start with the empirical evidence. Is there empirical evidence to suggest that the propensity to engage in serious crime has powerful genetic determinants?

Harden: Again, I would say genetic influence rather than genetic determinants. Going back to the laws of behavior genetics—everything is heritable, which is the third law. The second law, I believe, is that siblings raised in the same family are not usually all that similar—that being raised in the same family doesn’t make you alike, and that most of our environmental influences are much more idiosyncratic and not structured by the family.

Mounk: Is the first law simply the counterpoint—that there is some variation in behavior that is not influenced by genetics, that genetics is not 100%?

Harden: Yes. What we see is that there are very few extremely rare genetic disorders—less than 0.1% of the population, usually running in a few families—where we see what we would call monogenic, single-gene disorders that cause people with them to be seriously antisocial, aggressive, and impulsive. I talk in the book about one study of a Dutch family where the men inherited an X-linked gene that affects an enzyme in the brain, and they were all seriously antisocial. They had committed arson, homicide, and rape; multiple members of the family were in prison. So from the monogenic side, we see evidence that rare genetic variants can really warp your ability to restrain aggressive and impulsive impulses.

Most people who are aggressive do not have a monogenic disorder, but we are increasingly identifying genes that—and again, these are not deterministic, they work in combination with the environment—do increase your risk of becoming addicted to alcohol or drugs, of struggling with serious impulsivity problems from childhood, of behaving in more aggressive ways. This middle ground, where there’s not a single gene but many genes that raise your probability, produces correlations roughly as strong as those we see with a childhood maltreatment history, and that raises real questions. My book is me grappling with that question: how do we make sense of the fact that we didn’t choose to have this body and brain, and yet when we’re adults and we hurt each other, we hold each other responsible for our behavior?

The other theme in the book is that I was raised as an evangelical Christian when I was a young woman, and I’m really interested in how older narratives from Christianity around the inheritance of evil or the inheritance of sin continue to shape our debates about the science. Genetics is a very young science, and I think we are interpreting much of it through the lens of these much older, more Christian stories.

Mounk: You can see again why there might be so much resistance to thinking about genetics as a determinant here. I had a great podcast episode with Abigail Marsh on psychopaths. The fact that you can often diagnose a psychopath very early—at seven or eight years old, you can relatively reliably tell that a child has psychopathic traits—means that for obvious reasons, particularly if you have a child who exhibits those behaviors, there’s a reason why you might not want to accept that your child has a very strong tendency toward that. Abigail Marsh argues that with the right interventions, you can at least moderate the adverse impact it’s likely to have on the life of the child and on the lives of the people around them. But it is very scary to contemplate.

There’s an important distinction we need to make, and perhaps should have addressed earlier. One thing people sometimes mix up is that genes can be hugely influential on our behavior, while at the same time there is going to be significant variation within families. Siblings only share 50% of their genetics, and especially when it comes to particular constellations of genes or recessive genes, it’s very possible that some members of a family have certain traits that others don’t. I wonder whether that’s part of what makes people so instinctively skeptical—whether we’re inclined to say: you’re telling me it’s all genetics, but I’m different from my parents, so it can’t be genetics.

Harden: We tend to think of genetics and heredity through much older ideas of inheritance. If you had a hereditary title, you passed it down to your oldest child intact, in full. If you inherited your parents’ estate, you got all their money. But you don’t get all your parents’ genes—you get a random 50% draw from your mother’s and your father’s. One way to think about it is: your mother has a pantry of ingredients that she’s used to make a recipe, and your father has a pantry of ingredients he’s used to make a recipe. You get 50% of the ingredients from each pantry. What recipe can you now make? It might be very different from either one.

That’s also why parents are always surprised by how different their second child is. They think: this is what happens when you combine my genes with my partner’s—and then they have their second child and it’s a completely different person. It’s a different combination. Getting out of the habit of thinking of genes as a source of continuity, and thinking of them also as a source of discontinuity, shuffling, and change from parent to child, is a really important shift in thinking. You are not your parents, and part of the reason you’re not your parents is that your genes are different from theirs. You’re not your parents’ clones.

After my first book came out—and I write about this a little in the new one—I had so many people reach out to me. Someone who was the oldest of eleven and felt so different from their siblings, the only one who went to college. Someone who was adopted and didn’t find out until young adulthood, and said it suddenly made so much more sense because they had always felt like an outlier. Someone whose brother was an addict who had never been able to stay clean, who had always worried about having children because of what they might carry. People have this intuitive sense that family history matters, but that it can also surprise you. I appreciate being able to describe how that’s actually happening scientifically—this reshuffling that occurs whenever we create a child.

In the rest of this conversation, Yascha and Paige discuss how society should respond to these questions, to what extent we have free will, and what genetic influence on our behavior tells us about humanity. This part of the conversation is reserved for paying subscribers…

Read more

H. W. Brands holds the Jack S. Blanton Sr. Chair in History at the University of Texas at Austin and is the author of more than a dozen biographies and histories. His latest book is American Patriarch: The Life of George Washington.

In this week’s conversation, Yascha Mounk and H. W. Brands discuss why Washington’s reputation has endured despite the controversial legacies of other founding fathers, how his frontier upbringing shaped his character, and what Washington’s voluntary relinquishment of power reveals about leadership and ambition.

This transcript has been condensed and lightly edited for clarity.

Yascha Mounk: George Washington is, I suppose, the most famous figure in American history. Many biographies have been written about him. Looking at him from the vantage point of 2026, as we celebrate the 250th anniversary of America’s founding, what stands out to you about George Washington that may be different from what other historians emphasized when they wrote about him throughout American history?

H W Brands: The principal thing, since you mentioned the vantage of the 250th anniversary, is Washington’s staying power. Upon his death, one of the eulogies described him as “first in war, first in peace, and first in the hearts of his countrymen.” It’s hard to say where Washington sits in the hearts of his countrymen today, but he is one figure from that era who has survived most of the revisionism that has eventually come for people like Thomas Jefferson and any number of other figures from that era. This is not an accident of history. The capital city of the United States is named for George Washington, so it would be pretty difficult to de-platform him.

Washington wears well. He has a level of gravity with which he fulfilled his roles as president that has given him a long shelf life. Any complaints that people might make about Washington are easier to make about other figures from the founding. Jefferson is the one who said all men are created equal and owned slaves. Washington owned slaves too, but he didn’t say all men are created equal. So he seems to be a very solid figure.

There is always interest in the first of anything, including the first president and the first victorious general. Washington is more responsible for the creation of the American Republic than any other person. If his Continental Army had lost to the British, American independence wouldn’t have happened—at least it wouldn’t have happened then. He is an important figure and somebody who is easy to come back to again and again.

Mounk: There are many aspects of his life that I want to get into—his childhood and upbringing, which you portray very interestingly, and his responsibility for the American Republic. That responsibility stems in part from his military victory, and in part because he clearly did not have the ambition to become a dictator or monarch. The longer history goes on, the more positive a light that puts him in.

If you look at many of the newly independent republics around the world that threw off colonial shackles, the number of times in which the courageous, charismatic independence leader who leads the rebellion against colonial forces becomes the first president of a country and then stays there for fifty years—making the country corrupt and concentrating power in their own hands—is enormous, especially in the twentieth century. With more knowledge of how easy and how tempting it is for those kinds of figures to declare themselves indispensable, the fact that Washington gives up power after two terms, setting the informal norm that presidents should be limited to two terms, is even more significant than it may have looked at the time.

Brands: Your point is very well taken. When King George III of Britain heard that Washington, upon the end of the Revolutionary War, had resigned his command and returned it to the Continental Congress, George III is said to have remarked that if that report was true, he was the greatest man in the world—for doing exactly what you said: not being beguiled by political ambition.

When we get to that point in the narrative, I’ll explain why it would have been a different thing for Washington to attempt that than for any number of other victorious generals later on. But it certainly serves his reputation well that he did it. In fact, as you suggest, he did it twice: once at the end of the Revolutionary War, when he resigned command of the army, and then after two terms as president, when he retired and returned to Mount Vernon.

Mounk: Let’s go back to the beginning and work our way up to these later stages. Washington is often portrayed as the consummate patrician—as somebody born at the top of a social hierarchy, a revolutionary leader who is, in some ways, to the manner born.

You argue that this is an oversimplification. Even though he is born into a reasonably privileged class of landed gentry, and into a slave-holding family, he is not nearly as elite as some of those portrayals have made him out to be.

Brands: Both parts of that story are crucial. Washington was born into the Virginia gentry—the upper class of Virginia in the early eighteenth century. Compared to the people around him, Washington was privileged. He did not have to worry about where the next meal was coming from.

On the other hand, this was a colonial gentry, a frontier gentry. The serious members of the aristocracy of the British Empire all lived in England. The Fairfax family was one that was transatlantic, and Washington was fortunate enough to have got to know the Fairfaxes—Thomas Fairfax and others—when he was young. That gave him one leg up on climbing the ladder, because even within a gentry, even within an aristocracy, there are always gradations. Washington was not born at the top of that, and so he had to aspire. He had to work to ascend. He wasn’t born particularly wealthy—certainly wealthy by comparison with the yeoman farmers of Virginia, and certainly by comparison with the enslaved population of Virginia—but he couldn’t simply rest on his inheritance. He wanted to make more of it.

A continuous theme in Washington’s life was his desire to acquire more land. Everybody in Virginia essentially made their living by the land. Washington was a landowner, but he wanted more. This is what took him on his first adventures to the West. Although Washington was born and grew up in Virginia—which was the first permanent colony and has a seaboard—he was associated with the western part of the state. The western part of Virginia at that time included what is now West Virginia, Kentucky, and parts of Ohio. He was constantly heading out to the West, and it was as someone who had been to the West that he came to the attention of the government of Colonial Virginia and eventually to the British government in London.

Mounk: What does he do in the West? He’s a surveyor, but what precisely did that mean? When I picture a surveyor today, I picture somebody in a high visibility vest with a strange machine standing around suburban London.

Brands: If you were surveying property in the wilderness, as in Alaska, you would have to go out and do some serious measurement. If you survey in a city today, everything is already measured and you simply find the post that has already been located. Washington went out into western Virginia, which had never been surveyed before, in part because he was looking for the best land in a large grant that had been given to the Fairfax family. He was hired by the Fairfaxes and went out with one of the younger Fairfax men on this surveying trip. It was his first expedition into the wilderness.

At that point, Washington began something that would be his great gift to biographers and historians: he kept a diary, a journal, almost every day of his whole life. It starts here, and so it is possible to find out what he was doing each day and how he was responding to the challenges he faced. One of the things Washington learned about himself on this journey—and that the reader of biographies learns about Washington—is that he was somebody who thrived in the outdoors. This was not a given for a member of the gentry, many of whom were more comfortable in parlors and salons. Washington really liked being outdoors. It was a great adventure. He was camping out, getting rained on and snowed on, falling into rivers, and nearly freezing—and he thought it was the greatest thing in the world. He was good at it and he liked it, and this was going to serve him very well.

He kept his journal, he came back, and the surveying was fine. But he also developed a reputation—and he was a teenager at this point—as somebody who was resourceful, who could find his way in the wilderness, who could get along with the various Indians and frontiersmen who were out there, and accomplish what needed to be accomplished.

When the Virginia governor required somebody to carry a message to the French, who were trying to colonize the Ohio country—claimed by Virginia but also by the French—he said Washington was the man to do it. Washington had developed a reputation as somebody who could take on challenges, figure a way through wherever he had to go, and get the message to where it had to go.

He also discovered that he was a wonderful horseman, and this was a significant distinction in those days—not only for getting around, which was how you got around, but for winning the admiration and even the envy of your fellows. A man who was good on a horse cut a commanding figure. It didn’t hurt that Washington was big for his age: tall, well-built, and imposing on horseback. That is often a metaphor, but in Washington’s case it would become an actuality, because Washington on a horse was a very persuasive military commander.

Mounk: Of course, it’s how he’s often still portrayed when you look at statues. Like many other people of that age, he’s shown on his horse.

Brands: The fact that so many military figures are portrayed on horseback is testament to the impression that a man on a horse made. When you are on a horse—especially if you are a tall man like Washington—you are head and shoulders and torso above everybody else. In those days before radio communications, your soldiers had to be able to see you. The fact that the enemy could see you too, and level their guns at you, and perhaps kill you, made it all the more impressive when you survived. Washington cut an impressive figure on a horse, and that was quite important in his day.

Another thing Washington figured out at an early age was that he had to develop a character. He would ascend by reputation. Reputation was the coin of the realm among the gentry—did you have a reputation as a brave man, a capable man, an honorable man, a generous man? Washington decided that he needed to win that reputation. How do you win a reputation? You do the things that are conducive to it. Washington created this public character, set this model for himself: he was going to become this person. He took the steps and the actions to allow him to live up to the reputation he was aiming for, and he did this his whole life. It shows the degree to which virtue of this kind—civic virtue, public virtue—can be developed. It is not innate. You say, this is the man I am going to become, and then you strive to become that man.

Mounk: Looking at this from 2026, what is striking is that the incentives toward good character—or at least what was at the time regarded as good character—were aligned. An ambitious man like Washington, born to some privilege but very conscious of wanting to rise in the social hierarchy, thought that both appearing to have a good character and cultivating the virtues that actually gave him a good character would be conducive to that goal. It is not clear that the same is true today. If you are born with some social advantages and you are thinking about how to get ahead in life, cultivating the best possible character is probably not where the incentives lie. If what you want is fame, renown, or political power, the incentives go in a different direction.

Tell us a little about the society that produced him. Americans, I think, are a little confused about this. We have a sense that there was a kind of gentry in the eighteenth century, but we also have a strong sense that unlike in Britain, there was not an aristocracy. How should we think about this? It is a world in which virtue is rewarded and certain character traits are highly regarded, but it is also a world in which people are scheming to acquire more land, in which people justify slavery, and in which there is a great deal of partisan contestation—different factions and groups plotting against each other. How should we understand the social background against which Washington’s rise takes place?

Brands: The first thing to keep in mind is that this was a pre-capitalist world—one in which not everything had been monetized. Washington made his living as a planter. He planted crops, harvested them, and sold them. Tobacco was the largest cash crop, but tobacco was hard on the soil, so he shifted to wheat, oats, and other ordinary farm crops. He did not get into cotton, which came later and would change the face and economics of the South. Washington came along before cotton became a viable commercial crop.

Since it was a pre-capitalist society in which things were not particularly monetized, people did not compete on the size of their bank account. They competed in terms of hospitality and the quality of their homes. Washington inherited a home and gradually enlarged it. If you were a member of the gentry—a relatively small class, a few hundred people, somewhat larger when you add in the relatives—you were expected to be a gracious host. In Virginia, where the economy was based largely on slavery, you were a slave owner, or perhaps more precisely a slave manager. Most of the slaves that Washington controlled he did not own, which meant he could not have sold them or emancipated them even if he had wanted to—but we will get to that.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

Because Virginia had no cities and was entirely countryside, travelers could not expect to find a hotel or inn. They stayed in private homes. Washington had open house at Mount Vernon pretty much every day. Anybody traveling from the southern part of Virginia toward Maryland would stop in, expect to be given dinner, and if the weather were poor or they could not continue on, expect a room for the night. This was also the way news traveled and the way Washington kept informed about the world.

A member of the gentry might appear very wealthy while being deeply in debt. The best example of this is Thomas Jefferson, who had his wonderful place at Monticello outside of modern Charlottesville, was constantly rebuilding, and seemed to live the life of a wealthy man—but was always deeply in debt. At the time he died, everything he owned had to be sold off to pay his creditors. Even Washington, at times one of the largest landholders in Virginia, could not come up with the cash to pay a relatively small debt and had to ask for more time. That was a mortifying thing for a man of Washington’s reputation, but that was how it worked.

One of the things that being a planter, being a member of the gentry, gave you was practice in being—the term fits—a commander. You were the commander of your little village. You were literally in command of the enslaved people who worked under you, and the boss of the free people who worked alongside them. You had executive experience in a way that a lawyer in Massachusetts—to take John Adams, the second president—did not. It is not an accident that of the American presidents before the Civil War, a majority had been slaveholders and planters. They had the advantage of managing large operations and making decisions within them.

Mounk: There is a similar point to be made in a totally different context. There is sometimes a line of argument that if artificial intelligence displaces a lot of human work, that will be fine, because we will be in a very affluent society, give people universal basic income, and they will be like latter-day aristocrats, enjoying their leisure and cultivating their skills. That slightly mischaracterizes what the life of the average aristocrat was actually like—not just in the context of the American gentry, but of European aristocracy as well. You could neglect your lands and the people cultivating them, and you would probably end up with a lot less money than you were passed and pass on a lot less to your offspring. You could do that for a generation, but it constituted a deep failure of honor. If you actually wanted to be a good steward of your inheritance and your family’s position, you were running a small to mid-sized business. If you were bad at it, that might be tolerable for a generation, but it was genuinely discrediting and created significant problems.

To return to the eighteenth century: Washington really starts off as a British loyalist. He thinks of himself as part of an extended British aristocracy and part of the project of building the British Empire. How does that change? How does he turn from a British loyalist into an American revolutionary?

Brands: This is one of the principal tasks that a biographer of Washington, or a historian of the period of the American Revolution, has to tackle. It is important to remember that when the United States declared independence in 1776, that did not signify that all people living in America were in favor of independence. There was a large category of people—called Loyalists or Tories at the time—who said that independence was a terrible idea. Many of them took up arms against George Washington and the Continental Army and fought on the side of the British against American independence.

The question is, why would somebody like George Washington come out in favor of independence? Why would he become a rebel, a traitor? This is an important question because typically the people who turn against their regime are people who are not succeeding within it. The successful ones are happy enough and stay around. Washington was quite successful. By the time of the American Revolution, he had risen even farther up the hierarchy of the Virginia gentry. He had married well—a wealthy widow—and acquired a great deal more land, more slaves, and more property. He was well respected and had a reputation as Virginia’s top soldier. Not America’s top soldier, but Virginia’s. That distinction matters, because talking about America in 1770 was like talking about Asia today—a vast collection of different peoples and places. There were Virginians, New Yorkers, Pennsylvanians. People who lived in Virginia thought of themselves as Virginians and as subjects of the British crown, but not as Americans. They felt they had more in common with people living in London than with people living in Boston.

Things were going well for Washington, but there were certain annoyances. Annoyances are one thing, but what turns annoyances into grievances that rise to the level of sedition and treason? Virginians, like other people living in the American colonies, had over a period of roughly one hundred and fifty years grown accustomed to the idea that they would pretty much govern themselves. This was due primarily to inattention and laxity on the part of the British government. When colonists left England and went off to America, they fell out of sight and out of mind, and were allowed to do more or less whatever they wanted.

Mounk: The distances involved were so enormous that it created a kind of principal-agent problem. The principal had great difficulty instructing even the agent of the British Crown sitting in Virginia exactly what the preferences were. Some amount of self-government was inevitable as a result.

Brands: From the standpoint of London, there was very little of interest in North America. Further south, in the West Indies, there were sugar plantations and money to be made. But out of North America, not much at all. Decades would go by and the colonists in America would have very little to do with what was going on in London, or vice versa. They got of a mind that they ran their own affairs, and they interpreted this in ways that served themselves. They thought, for example, that they ought to have the right to raise and spend money as they wanted. If they got dragged into one of Britain’s wars, they did so sometimes with less than perfect grace.

When the government in London went to war with France over something that had nothing to do with America, Americans found themselves drawn in anyway, usually because the French controlled Louisiana and parts of the West Indies. They were dragged into a war they had no part in starting and no benefit from at the end. That got annoying.

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

This brings us to what began in America as the French and Indian War—not the French against the Indians, but the British and their Indian allies against the French and France’s Indian allies. It was a war of crucial importance for Washington and for people who lived in Virginia and Pennsylvania, and it was a war that Washington helped start, because it was over control of the Ohio country—roughly where Ohio is today, including western Pennsylvania and West Virginia. This mattered enormously to people like Washington because Virginians and Pennsylvanians had their eye on that western territory, particularly its land. The population of the American colonies was growing rapidly. Benjamin Franklin calculated around this time that the population doubled roughly once a generation, and since nearly all of them were farmers, they needed land to make a living and to bequeath to their children. Land in the West was cheap.

Washington joined Franklin and others in something called the Ohio Company, a speculative venture in which they would acquire title to a large tract of land in Ohio, survey it—which is why Washington got involved—and sell it to others. Over time, this habit of purchasing land and reselling it at an appreciated value would make more fortunes in American history than almost any other means. If the French took control of the Ohio country, those dreams of speculative fortunes would vanish. The Virginians had a real stake in the outcome, and so they were more than happy to join the British in the French and Indian War.

The conflict eventually became what was called the Seven Years War, the later years of which had nothing to do with North America and everything to do with empire and Europe. From the American standpoint, the fighting ended in the late 1750s, but there was no peace treaty until 1763. At the end of it, Americans were rather miffed to receive what amounted to the bill for their part of the war. The British, having run up an enormous debt fighting the French, did what governments do when facing a deficit: they reduced spending and increased revenues. Reducing spending meant, among other things, reducing the number of troops in the western territory just won from France—the very territory that people like Washington had expected to survey, settle, and profit from. Instead, Americans were told they could not settle there, because no troops would be stationed to protect them and because, to avoid trouble with the Indians, the western territory would be closed to settlement. This policy was articulated in the Proclamation of 1763.

Then Parliament passed the Sugar Act and the Stamp Act—a series of taxes. Americans had fought the war and won it, and now they were being told they could not enjoy the benefits of winning, and would moreover face higher taxes as a consequence. This was not what they had signed up for. The idea of being taxed by Parliament would not have surprised anyone living in Britain, but it did surprise the people living in America, because it had not been the case before. The understanding had been that colonists funded their own local taxes, which they voted on themselves, and that the British could levy taxes on imports as a way of regulating trade. But internal taxes were another matter. This is when people in America began thinking that perhaps their connection with Britain was not such a good idea after all.

It was not unthinkable. If you lived in the south of England and said you didn’t want to be governed by London, you simply couldn’t do that. But if you were living in Virginia, three thousand miles away, and the British had shown up only rarely for the previous three generations, going it alone seemed possible. This is what somebody like Washington began thinking. Resistance followed. There were protests against the Stamp Act and a boycott of British imports—economic sanctions, which have been around forever. The thinking was that British merchants would feel the pain and go to Parliament demanding that the colonists be appeased so that trade could resume.

Mounk: At this point, the rebellion is against the fact that the deal has changed—that suddenly the colonists are supposed to pay a great deal more in tax while receiving fewer services in return.

Brands: I wouldn’t use the term rebellion yet. Protest would be a better word.

Mounk: At this point, the people involved are thinking that they need to resist and send a message, but they are not yet thinking of themselves as the founders of a new nation.

Brands: The colonists made a point of saying they were loyal Englishmen, standing on their rights as Englishmen to be free of taxes except those they voted upon themselves. This had roots all the way back to Magna Carta in English history. When it served their purposes, the Americans said they were Englishmen who must be treated as Englishmen and granted the rights of Englishmen.

This did not always sit well with the people back in Britain, who thought of the colonists as provincials—second-class subjects who had gone out to live on the frontier. There was a certain disdain that developed in Parliament and in England generally toward the colonists, and the colonists felt it. That too entered into the consideration.

Mounk: Was there something provincial about American life at this juncture? If you were at Mount Vernon or Monticello, were you ten years behind the fashion in London,? Did it feel like London was a metropole and America the periphery, or had a different culture already begun to grow up even before 1776?

Brands: It really did depend on where you lived. If you lived in or around Boston, you thought of Boston as fairly up to date. Boston had Harvard College, a literary circle, and libraries. It looked like a miniature version of London, and partly as a result, people in Massachusetts were even more annoyed by the disdain shown to them by people in London. But somebody like Washington in Virginia—which, as mentioned, had no cities anything like Boston or New York—tended to deal directly with agents. If you were a planter like Washington, you had an agent in London who handled the sales of your produce, tobacco for example, and who handled the purchase of whatever you wanted. Washington might write to say that his carriage was wearing out and ask for the latest model with his specifications. Jefferson in particular would write to his agent asking him to go to the local booksellers and acquire the fifty most important books published in the last year. They certainly tried to keep up that way.

Subscribe now

There was as yet little in the way of higher education, arts, or sciences in America, though a beginning had been made. Benjamin Franklin helped establish the American Philosophical Society. There was Harvard College, Yale College, and the College of William and Mary in Virginia. But if you were a wealthy family and wanted your son to become a lawyer, you might very well send him to London to study at the Inns of Court. It was clear that London was the cultural capital of the realm that included the North American colonies. Boston was a partial exception. New York was a world of its own, having been founded by the Dutch and being more commercially and bourgeois-minded. But Virginia, in many ways, should have been the most loyal of the colonies.

Mounk: Part of what is involved in this conflict, beyond the dispute over taxation, is a sense of feeling disrespected—of having the rights of Englishmen but not being treated as equals, of being treated as second-class subjects. How does this conflict escalate? In retrospect, was escalation inevitable, or could other choices by London have kept America within the British Empire for decades, centuries, or perhaps indefinitely?

Brands: Taking that last question first: it is very difficult to imagine that in 2026 the United States might remain part of a British Empire—empires have long gone out of fashion. But it is not impossible to imagine a change of government in London in the 1770s, before the American Declaration of Independence, that concluded it was better to keep the Americans as friends than to have them as enemies and chose to compromise. Had London been willing to compromise, there were plenty of people in America willing to meet them halfway. There were very few people before about 1774 who were calling for independence. There was no real groundswell in favor of it until relatively late in the game.

What would become the United States might very well have evolved as Canada evolved. Canada was invited by the American colonies, upon independence, to join their revolt against Britain. The Canadians said they were doing well enough within the British Empire and declined. They gradually received more and more home rule until, by the 1860s, Canada was effectively independent—still recognizing the British crown, but to all intents and purposes self-governing.

America broke with Britain in the 1770s and made it effective as of the 1780s. The two countries were formally at war in what Americans call the War of 1812 and were at odds with each other for some time after that. But since 1900, the United States and Britain have been the closest of allies—in some ways closer to each other than Britain and Canada were to each other. So today, had there been no American Revolution, or had it failed, we would probably be roughly where we are today, arrived at by a different timetable and somewhat different means. If those different means included avoiding a war in which thousands of people died, that would not have been a bad thing. If you can avoid wars, you should try to do so.

Mounk: Counterfactual history is always difficult. A later separation could also have led to a much bloodier war fought with more modern technology, and perhaps one that would have soured relations between Britain and America far more deeply. In a sense, it is healthy for adolescents to separate themselves from their parents, because it becomes easier in early adulthood to come back and develop a strong relationship. If you delay the period at which you create autonomy for yourself, the relationship can get much more complicated and rocky later on.

Brands: There is one other thing to consider. The British Parliament in the 1830s decreed an end to slavery within the British Empire. The United States would not end slavery until 1865, with the passage of the 13th Amendment at the close of the Civil War. If America had remained part of the British Empire in the 1830s, slavery in America might have ended thirty years before it actually did.

This is complicated, because ending slavery within an empire that included the cotton-growing American southern states would have been a far bigger undertaking and might have been more difficult to achieve. But what made it so hard to end slavery in the United States was that the people who owned slaves and benefited from slavery had to vote to end it themselves—they had to absorb the economic hit personally. There were no slaveholders in London when Parliament decreed an end to slavery; the burden fell on planters in the West Indies. Parliament could, in theory, have told South Carolinians to give up their slaves as well. It is useful as an exercise to consider these things.

Mounk: One other possibility is that the war for independence and the war to preserve slavery might have coincided. Rather than having a Revolutionary War and a Civil War as separate events, a number of states might have broken away from Britain over the issue of slavery in the 1830s—and had they prevailed, they might have preserved slavery. The possibilities are many.

Brands: One last point on that: had the slave states broken away in the 1830s, they would not have been opposed merely by the northern states of the United States. They would have been opposed by the northern states and the rest of the British Empire.

Mounk: Or they might have been able to forge a coalition with the northern states, because the conflict would not have been about the nature of the future political settlement within an independent republic. They might have been able to wed the issue of slavery to the issue of self-determination for the colonies and forge a coalition of northern states against Britain. We can play this game forever, but let’s get back to actual history.

Where does the phase shift occur? Where does protest turn into rebellion, turn into revolution?

Brands: When the Americans protest, the British, instead of compromising, double down. Within British politics, the party that favors acting tough on the Americans wins out—it is a popular thing in a campaign to promise strength and the enforcement of the law. When the British escalate, the Americans escalate in turn. The Americans stage violent protests, the British send troops, and there are more protests. When the Americans see that British troops are coming to be used not against the French or the Indians but against them, Britain begins to look like a foreign country with an army of occupation.

The fighting begins when a Massachusetts militia moves to defend one of its armories. Every colony had its own militia—part-time soldiers who farmed during the day and took up arms when trouble arose—and their own stashes of weapons. The British decided to seize one of these armories so that the Massachusetts militia could not use the weapons against British troops. The militia refused, and the battles of Lexington and Concord took place—a firefight between the two groups that marks the start of the American Revolutionary War. At the time, nobody realized it was a war, nobody realized it would last a long time, and nobody quite knew what it was about.

Just two months after this, Washington is appointed commander of a new army. The Continental Congress—an ad hoc, self-appointed body, not authorized by British law—met in Philadelphia, the largest and most centrally located city in the colonies, and decided they needed a unified organization among the different militias. This is 1775, and their stated purpose is to defend their rights as Englishmen—not as Americans, but as Englishmen. Looking around for the most accomplished soldier available, they settled on George Washington.

From his beginnings as a surveyor in western Virginia, Washington had gone on to command the Virginia regiment and won a reputation fighting in the French and Indian War. Crucially, he had served right alongside British officers during that war, and he and his men had performed much better under fire than the British troops and their officers. Until that point, the British had sustained the mystique of having the best army in the world—the best uniforms, the best drilling, the most precise formations. But when they came under attack in conditions that existed in America, they fell apart. It was Washington’s soldiers who saved the day and prevented a complete rout. At that point, Washington realized that British soldiers were not ten feet tall—that he was a better officer than theirs and his men were better soldiers than theirs.

At one point during the French and Indian War, Washington had sought a commission in the British army. He was a colonial officer, and the British soldiers fighting alongside him refused to take orders from a mere provincial. Washington went to the British commander in chief with a recommendation from the governor of Virginia, arguing that a commission would give his orders proper weight. The commander in chief refused, for his own political reasons, though he did tell the British soldiers to listen to Washington. Various people have thought that had the British seen what was coming, they would have been smart to commission Washington as an officer. His loyalties might then have shifted from Virginia to the British crown.

As it was, this episode showed Washington that there was a ceiling on what he could accomplish—on the reputation he could acquire and on how he would be seen—as long as he remained part of the British Empire. He might be the greatest man in Virginia, but he would always be looked down upon precisely because he was a provincial. The idea that people could lord it over him for no good reason made him think he could get along without the British.

Washington is appointed commanding general of the new Continental Army and sent to Boston, where the British are besieged inside the city by the colonial militias. The fighting has developed a momentum of its own. The British don’t want to be surrounded and captured; the Americans don’t want them to get away. Nobody could have told you exactly what they were fighting about—once you are playing a contest, you want to win, whatever the larger purpose may be. For the first more than a year of his command, Washington didn’t know what the war aims were. He was trying to keep from being captured by the British and trying to keep the British from getting away. The British did get away—they sailed away from Boston. Washington then retreated to defend New York, though the question of what exactly he was defending it for, and what the war was ultimately about, remained unresolved.

Mounk: What is the last step? We have a protest that turns into a rebellion, that turns into a war in which you are trying to win. How does that then turn into a revolution? How do the war aims clarify, and how do they clarify in the form of a declaration of independence?

Brands: In the winter of 1775–1776, fighting had largely ceased—in those days, armies generally stopped campaigning in winter because the roads turned to mud and nobody could move. Washington and his men were waiting for the spring fighting season, uncertain what they would do when it came. In the middle of this, Thomas Paine—a recent arrival from England—wrote a pamphlet called Common Sense. Its conclusion was that Americans had been fighting for several months and needed to decide what they were fighting about. The only thing that made any common sense to fight for was independence. Without it, you could win a particular campaign and two years later have to refight the same battle, having never established the principle that Americans make laws for Americans.

The pamphlet was an immediate sensation—one of the fastest-selling pieces of political propaganda in American history. All sorts of Americans suddenly found themselves nodding in agreement. When the Continental Congress met in the spring of 1776, it did not take long for the question of independence to come to the fore. But everybody understood it was a big step. To say they were defending their rights as Englishmen was one thing—something George III could understand even if he disagreed with it. To say they were fighting for independence was to commit treason. The people who signed the Declaration of Independence were putting their lives on the line. If they lost the war, they could lose everything.

From Washington’s standpoint it was a big step, and he was in favor of it. It also clarified things dramatically. He could now tell his soldiers they were fighting for their own country, their own rights, their own destiny—that if they won, they would no longer be subject to the whim of a distant king. They were creating a republic, a political system in which authority emanates from the people, not from above. When news reached Washington’s camp that the Declaration of Independence had been passed by the Continental Congress, there were celebrations. Now they knew what they were fighting for.

It is worth pausing here to note that when the protests against the Stamp Act began in 1765, nobody knew where things would end. It is an example of one thing leading to another. There were plenty of people in America in 1776—the Loyalists—who said this was too big a step, that bad laws were one thing but that breaking up the British Empire over them was an overreaction. There were people in London who said that had they known it would come to this, they would have been more willing to compromise earlier. Just as Americans were not unified behind the Declaration of Independence, people in Britain were not unified behind the campaign to suppress the Americans. Many in Britain said the Americans had legitimate grounds for their complaints.

This is crucial to understanding how the war was handled. It was a war that began in politics and would end in politics. The political question for Americans was where their loyalties lay. The political question for Britain was how much pressure to apply and how important it really was to keep the colonies formally subservient to the British Empire. Washington understood this, and it is one of the keys to his success as a military commander. He realized he did not have to defeat the British army outright. He did not have to win the war in a positive sense. He simply had to avoid losing it, because as long as he kept his army intact, the British would have to keep fighting. At any moment, people in Britain had the option to decide that the war in America was too expensive and too much trouble and simply end it. That is ultimately how the war ended.

To put it plainly: Washington lost almost every battle except the one that mattered—the last battle at Yorktown. He knew this all along. He could lose and lose and lose, but if he won the last battle, he won the war, because it would be that final battle that convinced the British that enough was enough and it was time to go home.

Mounk: Alongside the question of independence, there was a whole parallel set of debates about what the nature of this new country should be—should it be a republic or a monarchy, should it be a very loose federation of states or a United States of America with a stronger federal element? What role does George Washington play in those debates? To what extent is the picture of him as somebody who stays above the partisan and ideological battles—like those between Jefferson and Hamilton—accurate, and to what extent is he really just one more ideological combatant in the fights over what nature this new state should have?

Brands: George Washington was no political theorist, and he was surrounded by political theorists. Alexander Hamilton, James Madison, Thomas Paine, and others were writing political treatises about how this new country ought to be structured from first principles, going back to the Roman Republic. Washington was not one of those. He was an example of military and then political leadership, and between the theory and the practice of leadership there is sometimes a large gap. Washington’s example—first as commander in chief of the army and then as president—gave the advocates of republicanism cause for confidence that the American people could actually govern itself.

It was a big step. There was very little historical evidence, certainly in the last 2,000 years, that a republic could work. There had been a Roman Republic, so the idea was not created out of whole cloth, but there were no active republics in the eighteenth century that Americans could use as a model. They were trying something that had never been done before—creating a country by sitting down together in Philadelphia, holding a convention, and writing a constitution.

The government that Americans created for this new republic—the one we know today—was in fact the second try. The first had to deal with thirteen separate countries. Several of the newly independent states called their chief executive a president rather than a governor—South Carolina had a president, Georgia had a president. The states considered themselves fully sovereign and independent, owing nothing to each other except proximity and perhaps neighborliness. The shared incentive of winning the war against Britain elicited enough voluntary cooperation among them to make it work. But in some ways, victory over Britain was the worst thing that could have happened for this nascent republic. With the British threat gone, the separate states went their own way. They began bickering among themselves, imposing tariffs on imports from other states, and treating each other as rivals rather than allies.

Washington in particular became very dismayed at this. He had spent eight years of his life, risking his life, trying to create this newly independent country, and within a few years of the war’s end he saw it beginning to fall apart. He became convinced that a stronger central government was needed. Though he had no political ambitions—he had handed in his commission after winning the Revolutionary War and returned to Mount Vernon intending to spend the rest of his life as a planter—he was talked into joining a convention in Philadelphia to write a new constitution that would strengthen the central government. Washington was immediately elected president of the convention, though that meant only presiding officer. One effect of the role was that it spared him the necessity of speaking on the floor of the Constitutional Convention. This brings out a secret of Washington’s. He understood the power of silence. If you don’t speak, people cannot draw negative conclusions about whatever you’re saying.

Mounk: Is this a brilliant strategy, or is it simply his personality? He was surrounded by political theorists who loved to argue about these things—was he personally just not as interested in them?

Brands: It is both. He was not interested in doing it and he was not good at it. But it also enhanced his reputation. Put yourself among all these lawyers and politicians in a small room in Pennsylvania, all arguing with each other and basically lowering themselves to each other’s level. Washington is sitting there silent at the head of the table. People are imagining what Washington is thinking. He doesn’t say. They can all imagine that he is wiser, that he has the answers.

Mounk: It is always tempting to think that since he is wise, and since I know what is best, he must agree with me—he may not be at liberty to say it, but since we are both wise, he must hold the same opinion.

Brands: There is that. But there is also the very practical matter that whichever side wins a particular aspect of the debate can claim Washington, and Washington doesn’t have to disavow anything he said or didn’t say. His silence serves him well. There is also the crucial fact that he had the reputation of being the victorious general, and no one else in that room did. That alone put him head and shoulders above everybody else.

With the exception of Benjamin Franklin—who was some twenty-five years older than Washington—he was a generation older than almost everybody else in the room. If they were founding fathers, he was a founding grandfather.

There is one last thing, and it really gets at Washington’s attitude toward being president and toward leaving the presidency. Washington recognized that his reputation was as high as it was ever going to get at the end of the Revolutionary War. He was the victorious general, the father of the country. He realized that anything he did after that could only tarnish his reputation, and so he had every incentive to stay out of politics.

When George III called Washington the greatest man in the world for relinquishing command, consider the counterfactual: suppose Washington had said in 1783 that he was not giving up command and intended to take over the government. What would he have taken over? There was no government to take over. America was an alliance—something like, today, a German, British, or French general deciding to take over NATO. You march on Brussels and what do you have? All the power and all the wealth are elsewhere.

In the rest of this conversation, Yascha Mounk and H.W. Brands discuss how Washington shaped the presidency and what lessons this period has for the United States 250 years later. This part of the conversation is reserved for paying subscribers…

Read more

James Traub is a journalist, author, and scholar. He is a member of the Council on Foreign Relations, the New York Institute For The Humanities, and the Society of American Historians.

In this week’s conversation, Yascha Mounk and James Traub discuss why progressive pedagogy is failing American students, how classical schools are achieving better outcomes through traditional teaching methods, and whether learning facts versus critical thinking represents a false choice in education.

This transcript has been condensed and lightly edited for clarity.

Yascha Mounk: Your latest book is about how to educate citizens in the United States, a relevant topic today. A lot of the debates in this area are about what students should be taught: are they being indoctrinated in woke ideology, or are they being indoctrinated in right-wing thought that downplays the evils of slavery? You think that in some ways, those very ideological debates are a little bit beside the point.

James Traub: These are the sexy things everybody wants to seize on—the whole right-left thing, which is an important part of my book. But if you go to school, which of course the people who are having this fight don’t, I spent a year in classrooms. One thing that struck me is the relatively-less-politicized-than-I-thought character of classrooms. You can go to a red state which has very red state standards written into its laws, and the teachers aren’t thinking about that. They often don’t know what’s in the standards. Teachers have a strong ethos that says, I don’t want to impose my views on the kids. Though that is an important part of my book, the thing that troubled me more was not what they’re taught. It’s how they’re taught. It’s pedagogy as opposed to politics.

American education, as far as I can tell, is ideological in a way that’s not true elsewhere. What I mean by that is not the right-left thing we were just talking about. It’s the notion that there is a progressive way to teach and there’s a conservative way to teach. That makes it very hard to have a sane and sober conversation about what’s an effective way to teach. The classes that I was in—many history, government, social studies, and other classes over the course of a year in many different places—the thing that troubled me was not the politics. It was how vacuous, how empty, how silent many of these classes were. It was how poor the reading comprehension of many students was. It was how little was assigned to them, how few books there were in the schools and in the classrooms.

The thing that I focused on, and I think is terribly important, is a pedagogy—the one called progressive pedagogy—that has this idea, which you could trace back to John Dewey and others, that learning facts, learning information, learning names, dates, places, the things you do in a history class, or for that matter an English class, is coercive, that it’s jamming things down children’s throats. A more effective way of teaching, in this view, is to engage their wish to ask big questions, to engage in critical inquiry and so forth, and, through that, they will come to know things. I don’t think that’s how most educated people got educated. Educated people got educated because they learned, they loved to learn, they found a way of learning. When you learn things well, you are naturally launched into the world of critical inquiry and thinking critically.

Mounk: Let me play devil’s advocate on this, because I’m a little bit torn. Even though I lived in mostly left-leaning parts of Germany and was taught largely by students who were part of a generation of 1960s student radicals, the education I received was, by that definition, quite conservative. History class was an unending parade of years and battles, learning off by heart various kings. The stakes of this were never made clear to me, and I was tremendously bored—it just felt like a rote exercise. I still grew up in an old world Europe where the idea of general knowledge about the world was really prized, and I had intellectual aspirations even relatively young. But it really felt like a rote exercise. Why does it matter that Frederick the Great ascended the throne at a certain point and died at a certain point? Well, part of why it matters is that he had a certain set of views and personality that shaped Prussia, and there’s a debate about whether something like enlightened monarchy actually existed. When you explain those stakes, then it becomes relevant when Frederick the Great ascended the throne. If you’re just learning that he was an influential king, these were his dates, and then a couple of dates of battles, you just don’t start to have a stake in this debate in the first place.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

On the other hand, I know that you need a certain set of facts and figures in order to actually engage intelligently in a conversation. I’ve been trying to learn more about China and I’m learning Chinese. I worked my way through a very good history of China, but I had real trouble retaining a lot of information because I don’t have enough context on the Song dynasty to really be able to place things in a scaffolding, and so I retained less. I certainly understand why you need some facts as well.

To give a third kind of alternative, I was always struck by the way that history is taught in Britain. It’s always taught through debates. Right from day one, it is: here’s one view of what happened in this period; here’s a second view of what happened in that period. That involves learning dates and facts that are contested, or the interpretation of which is contested. But the stakes are immediately made clear because you’re thrown into a debate about how to understand a period of history. Perhaps that’s a third possibility. I have to say that I’m slightly sympathetic to this idea that just throwing names and dates and battles at people is not going to get them interested.

Traub: Two thoughts. First, I’m fascinated that all of your super left-wing ’60s German professors were nevertheless still so unconsciously rooted in the pedagogy in which they had grown up that they repeated it. That’s in itself very interesting. Second, yes, there’s a reason why there was a progressive revolt against traditional forms of pedagogy—it’s because rote learning of facts with no larger purpose in view is boring and pointless.

For the purpose of this book, I read about the first times that American public education became a matter of debate, which was the 1890s or so. In the debate over what should be taught to American students in history, Woodrow Wilson was one of the members of the panel, and the great historian Frederick Jackson Turner was a consultant. You would have thought they would have said students must learn more history. What they said instead is that history must be presented in a way that makes it meaningful to students. For them, that did not mean students must not learn an important body of information. That body of information has to matter because you need teachers who are—and this was the language they would use—fired with enthusiasm for the subject.

The classes that I was in over the course of this year where I felt that learning a lot of material was always directed towards inquiry were the Advanced Placement classes or the International Baccalaureate classes. That made me think: can it really be that only 10% of the kids in the school are able to do the kind of work where real substantive learning leads to thoughtful conversation? The best schools that I was in were schools where I found that to be much more widespread—where it was simply taken for granted that if you’re in 11th grade, you are able to think important thoughts.

I’ll give you one last example. I was in a class the week before last, sitting with ninth graders—14 and 15 year old kids—who were having a deeply thoughtful conversation about the Aeneid. Ninth graders in America do not read the Aeneid. They don’t read whole books. But these kids were having a deeply searching conversation about it. The two are not in necessary tension. That’s what good teaching is.

Mounk: Let’s take a step back. One thing that I was always struck by coming from Germany is just the huge range of quality in American public schools in particular. Europeans often have this image that American schools are all quite bad. I didn’t find that to be true at all. I never went to an American high school, but based on what friends told me about the schooling over the years, I feel like I have a relatively good sense of it. When you look at the most failing schools in the country, you have a vast majority of students graduating barely literate, barely numerate—a complete disaster. Then when I hear about some of the schools that my friends went to in the suburbs of major metropolitan areas, which are public schools but in fancy neighborhoods—one friend of mine was in a rocket club where they had to call up NASA before launching some of the rockets because they were going to go up that high. I’d never heard of anything like that in Germany.

Traub: I went to one of those schools myself—a school full of neurotic, highly literate Jewish kids, a public school. It was great because it was a well-to-do suburb. We had great teachers and there was a huge emphasis on learning.

Mounk: Just to get a level reading here—the schools you’re going to: are you seeking out the worst and the best schools to contrast, or are you trying to look for average schools? What kind of schools should we be imagining when you’re talking about your experience reporting on this?

Traub: The first chapter of my book talks about schools in a suburb of Chicago—a big suburban area, largely middle-class and highly diverse. The first school I was in was the kind of school where almost everybody was going to go to college, probably many of the parents had gone to college, they had great facilities, the school was clean and nice. It was a majority white school, with Asian students as the second largest group—relatively high achieving students. And yet, when I was in a class of 10th and 11th graders reading the Declaration of Independence, it was very hard for a lot of them. I don’t think it was vocabulary in that sense—I suspect what defeated them more was unfamiliar syntax. Their reading ability, broadly understood, not just the ability to say or know the words, was surprisingly poor. What that meant is that those kids could not have had the conversation like the one I heard among the ninth graders about the Aeneid. They were not academically prepared to do it.

When you have this big fight about whether students should be learning a left-wing or a right-wing version of American history, you have to say: before you get there, they have to have the ability to think their way through a difficult text.

Mounk: Why is it that these ninth graders were able to have a certain conversation about the Aeneid, but these older kids at this other school—which is not a failing school, which is a pretty good and probably above-average school—could not? What explains the difference between these two schools?

Traub: The good story in my book involves what are called classical schools—a term I’d never heard before I set out to write this book, and a kind of school that would be unfamiliar to many listeners, in part because they’re seen as conservative. The word “classical” probably doesn’t help. They’re much more common in red states than blue, but it’s one of the most rapidly growing forms of schooling in the country. These are public schools—charter schools, which allows them to do things differently if they want to, but public schools nonetheless.

Mounk: Part of what you mean by private schools in this context is that they’re not fee-paying. So there’s not an extra financial burden for kids to attend these schools.

Traub: These schools are free and they have to take everybody who applies—they’re not selective. That’s very important, because if they were selective, you could say of course they’re taking especially talented kids. They are selective only in the sense that the parents choose to seek to have their kids get in.

These schools are self-consciously traditional in their pedagogy—a belief in vocabulary, learning stories, reading books. You may think that reading books is simply what you do in school, but the book is a rare thing in American public schools. Kids are not asked to read a book because it’s too taxing for their now very short attention spans. These schools assume that a kid can read a book, likes to read a book, wants to read a book. By fifth or sixth grade, kids are reading great novels—The Count of Monte Cristo—and talking about them. They learn Latin from an early age. They read the Iliad and the Odyssey—first in adapted versions, and then in grown-up versions later on.

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

These schools tend to have a lot of rules. Kids wear uniforms. They’re schools that a lot of parents probably wouldn’t like—too strict, they would say. One of the things I say in my book is: let’s think about what general principles we can learn from these schools, even if you don’t like them. One of those principles is that kids can do a lot more if from an early age you start them with higher expectations and more rigorous forms of learning. I went to another school like that with only Hispanic kids, many of them from impoverished backgrounds. Their conversations weren’t as searching, but they were reading Crime and Punishment in 10th grade—which is pretty remarkable if it’s your second language and you didn’t grow up in an educated household.

Mounk: Do you think this is something that could be rolled out more broadly, and it’s really just a question of the wrong approach having gained too much ground in a lot of public schools? Or do you think there is a selection effect? You’re saying rightly that these schools have to admit everybody, but one of the big indicators of success in life is how educated your parents are. If you have educated parents and you have to flee your country penniless, come to a new country where your parents don’t have any social capital, don’t speak the language particularly well, don’t have any connections—statistically, the kids of those parents are going to do very well. So the self-selection by the parents may still play a crucial role here. How much do you think the education of our kids would improve if a lot more schools adopted either this classical curriculum or a modern version of it—not necessarily learning Latin and reading the Aeneid, but reading the great novels in the English language and great contemporary novels?

Traub: My little dream is to rename this form of education “liberal education,” which is what it would have been called in the Renaissance. If you took that word away, maybe you could have more of these schools in the blue world—and if you had them in the blue world, they would have different ideas about the way you want to shape a young person, which would be a good thing.

As for your first question, there are no studies of this yet. There is something called Core Knowledge—E.D. Hirsch, a great literary scholar at the University of Virginia, created a pedagogy which assumes that you need to have high familiarity with a lot of words and expressions in order to read your way through a text. It’s a focus in curriculum across almost all the major subjects on learning particular things as a means towards becoming a more thoughtful and critical actor. There are a lot of studies that show that it works. If you were to ask what can be generalized from these schools, that pedagogy is the big thing.

The people who say their daughter is not going to be happy in such a strict school, that she’s a free spirit—they shouldn’t go there. But there is nothing conservative about learning in the Core Knowledge way. There is nothing liberal—speaking as a liberal, a left liberal—about the other thing that calls itself progressive learning, though it has a liberal pedigree. When I talk to people, I get the strong sense that America is full of people whose politics are liberal but whose intuitive pedagogy is traditional, like me.

Mounk: What about the selection of the teachers? I assume that a lot of these classical schools also have teachers who may be more educated, may have a deeper love of learning, who are excited to be part of one of these model experimental schools because perhaps they are disenchanted with how normal schooling works. I imagine that the quality of the teachers in these schools may also be higher than in regular public schools.

More generally, what did you think about the quality of teachers in the schools you were in? I’m really struck by the fact that if you go back to the 1950s and 1960s, a lot of teachers were women who had graduated in the top 10% of their class. There obviously weren’t other job opportunities, or there was an expectation that they would do that job for a number of years and then get married and leave the workforce—so it was for bad reasons, produced by bad background conditions. But it was a great thing for the students of that age, because they were taught by very, very smart people. My understanding is that today the average, or median, teacher graduated in the bottom half of their class. If that’s the case, how can you get teachers who were not among the best and brightest when they were in school to inspire students who, even though they may not be getting a great education at 15 or 16, are in some ways smarter than their teachers? That is a hard setup.

Traub: If I were king, this is what I would do: increase teacher salaries—by a lot. Who knows what number, 50%, I don’t know. Then you not only would have a better pool to select from, but you could make different demands than the ones you make now. The schools that I was so admiring rarely take teachers whose graduate degree is in education. They take teachers whose graduate degree is in a subject matter. The school that I write about the most—a school in a middle-class, working-class suburb outside of Dallas called Louisville—the principal there said to me: I can help a teacher who doesn’t have the pedagogical background be a better teacher. I cannot take a teacher who doesn’t have the love for his or her subject and infuse that. He only hired teachers who had a master’s degree in a subject. Most of the other schools I went to, almost all the teachers—except the best ones—had an education degree and not a subject matter degree.

My book is not about teachers’ colleges, so I don’t hold myself out as an authority on the subject. But the overwhelming impression I got was that if a teacher isn’t fired with enthusiasm—to go back to that 1890s language—for the subject matter, then it’s very hard to fire the students with enthusiasm. The only way to get teachers who are fired with enthusiasm is to get people who are intellectually alive, and those people need to get paid better to go into teaching.

Mounk: Let me play devil’s advocate here, and this is perhaps my least popular position. When you look at the OECD studies of teacher pay, the United States is well above the average of OECD countries. Part of that is that the United States is rather richer than a number of other OECD countries. The way in which the United States stands out is not that teachers make comparatively low pay. Having lived in both the United States and Europe, I think the average American teacher has a better standard of life than most teachers in other countries. The problem is that lawyers and doctors in the United States earn so much more than their peers in other countries that the economic gap is very difficult to bridge. But I don’t know that the way to solve that is to get a teacher to earn 75% or 50% of what a doctor earns, when the doctor’s pay is already so high that it’s basically unsustainable for patients to seek medical care.

Subscribe now

The other thing to note about teacher pay is that when you look at the LA School District, for example, teachers actually make a lot of money. One of the problems is the structure of it—in the first few years you earn very little, but once you get a certain kind of tenure, you actually make a great deal. Part of this is the teachers unions, which structure pay in such a way that senior members of the profession live very well for years, with extreme job protections. But that’s not enticing younger talented people to go into the profession. How do you solve some of those structural issues?

Traub: I think you’re exaggerating the steepness of that curve. Salary structures in the teaching profession tend to follow a steady, stepwise progression, with a bigger step if you’ve gotten another diploma or achieved certain milestones. It’s not as radical as that. I don’t think you have a whole lot of teachers earning $160,000—that’s quite rare. But when your starting salary is $48,000 or something like that, that’s tough.

Interestingly, at these classical charter schools, teachers are paid less, for reasons having to do with how they’re compensated by the state. How do they get decent teachers? The answer, in some of these schools—not all of them, some had fairly normal teachers and nevertheless the kids were still doing well—but at the best one I went to, when I talked to the teachers and asked why they were there, they would say: “Because I get to teach beautiful things. I get to have elevated conversations. I get to talk about great books with the kids.” I admit there’s a limited number of people who will do that—it takes real dedication. But it does tell you that the atmosphere of the school matters a lot in terms of the kinds of teachers it’s going to attract.

Mounk: Let me push back on that a little bit, because I think it’s important to understand the incentives here. The first thing you said is that a lot of the teachers you were disappointed with had gone to teachers colleges. It’s interesting to talk about what teachers colleges teach and how they shape education. From my understanding, these are extremely ideological places, not just in the content they convey but also in what they teach about the right approaches to pedagogy.

We’ve been talking about kids that are older, but for a long time a lot of American school kids weren’t even taught how letters work—how letters correspond to particular sounds and how that allows you to form a word. It took me years of ambient awareness of this debate to understand what was even at issue, because it seemed so incomprehensible that you wouldn’t learn that.

Traub: That’s the most notorious aspect of what I’m calling progressive pedagogy. The premise behind what’s called “whole language” is that children acquire reading ability naturally—that if you place them in an immersive bath of language, they will learn to read. That’s wrong. You don’t acquire reading the same way you acquire talking. But that was based on a kind of ideological predisposition, a Rousseauian view of the child. It’s very hard to break that view. But that view has been broken.

Mounk: Just to go back to this question about teachers, because I think it’s helpful to have a sense of it: a new teacher with a regular credential in Los Angeles has a typical starting salary of about $69,000. That’s not a ton of money, but for a young person—presumably most of these people are in their 20s—that is about average household income in the United States. If you’re a couple and you’re both teachers, you start off at double the typical household income. You live in a high-expense area, and Los Angeles is an expensive place—that’s important to acknowledge. But that is hardly catastrophic. By the time you’ve taught for about 10 years, a typical veteran teacher can expect to make around $110,000. On your own, you’re making about twice US household income. If you were a couple where both people are teachers, you make about four times the average US household income. It’s just not clear to me that the solution is for teachers to be making significantly more than that. That doesn’t seem to me to be the crux of the problem.

Traub: Let’s ask ourselves a question: as a society, do we feel that teachers or lawyers play a more fundamental role in shaping good collective outcomes? These are market-driven things, but if we do feel that we have an educational crisis—and my general view is that most spending in schools doesn’t produce any different outcome at all, whether it’s fancier buildings, swimming pools, or even smaller classes—then the only place where I would be in favor of spending more money is in some agreed transaction between paying more and demanding more. Your point, I believe, is that we’re already paying enough to make bigger demands than we’re currently making. That may be true. All I would say is it isn’t working now, and so I’d be in favor of trying to find some other way of making it work better. That said, it’s not going to happen anyway—these are local decisions.

Mounk: Let me make myself even more unpopular with two other segments of my listeners. If the problem isn’t that, as an individual earning $110,000, you can’t have a decent life in LA, or as a couple earning $220,000 you can’t have a decent life in LA—and I don’t think that is the problem—then a lot of the problem is relative prestige. A young person in Germany who thinks they’d like to become a teacher might worry that classmates choosing to become doctors or lawyers are going to have higher social status and more money. But the gap isn’t that large, so they pursue their love of teaching. Whereas in the United States, that same person at a good college who’s excited about teaching might in the end decide they don’t want to be in such a different earnings category than their friends, don’t want to have such a different level of social status, don’t want their friends always feeling sorry for them or to wonder if they can go out to dinner with them. Perhaps they’ll go and apply for that job in consulting or tech or investment banking instead. I’m sure that happens a lot—I know people like that.

Traub: Except that you can go to Teach for America—that’s one prestigious thing you can do. You can spend two years as an enrollee in one of these programs. But then the question becomes: how do we keep those people? How do we make them feel like this should be a lifetime vocation?

Mounk: Part of the answer is that we can’t have the taxpayer try to compete with the salaries currently paid to doctors and lawyers, because they’re completely disproportionate. The best-paid doctors and lawyers are always going to earn more, and there’s nothing we can do about that in a free society. But there is one thing we can do, which is to bring down the salaries of lawyers and particularly doctors. The way to do that is to recognize that those salaries are kept artificially high by licensing. Particularly in the case of doctors, the American Medical Association refuses to credential new medical schools, which would lead to many more doctors being trained. America has one of the lowest rates of doctors per capita in the developed world, and that is one of the reasons why the salaries of doctors are so high. So rather than taking the politically more left-leaning solution of increasing teacher pay, perhaps what we should do is push against some of those artificial regulations that create that disproportion in the first place.

Traub: The lawyer situation is really driven by their clients. Law firms are always seeking corporate clients because that’s where the money lies, so having more law schools may not have much of an effect. In any case, we are now talking about what I offered as an idle fantasy—I don’t have an overwhelming belief in it.

Mounk: What is the role of teachers unions in all of this? I had Randi Weingarten on the podcast for an interesting and at times somewhat contentious conversation a few months ago. To what extent do these classical schools that you admire have less influence from teachers unions, or perhaps the teachers aren’t generally part of unions?

Traub: Probably not, because these tend to be red states—though I don’t know that for a fact, since municipal teachers unions are pretty widespread and individual schools can choose to have a union or not. I was on the board of a charter school in New York City for many years and the teachers—I think through our own fault—were unhappy. They said they needed the protection of a union, they joined one, and I thought this was going to be the end of the school. It didn’t make that much of a difference.

If you ask me whether the world would be better with or without teachers unions, it’s a hard question because there are positives and negatives on both sides. Teachers unions view charter schools as competition and oppose them—that’s the worst aspect of unionization, the closed-shop mentality where you’re not thinking about what’s good for the kids but about what’s good for you as a professional. That’s bad. On the other hand, raising salaries is good. If you said to me there should not be public sector unions—which is an interesting claim made by some on the right who believe in unionization but not public sector unionization—I might say, perhaps in an ideal world you’re right. There should be more unionization in the private sector and less in the public sector.

Mounk: I want to make sure that we circle back a little bit to the content. What did you see people being taught about America? The cliché is that in progressive schools, students are taught that America is bad and evil, and in conservative schools, they’re taught that slavery never happened. I’m sure the reality is much more subtle on both counts.

I happened to have had a conversation with some of my undergrads about this recently. There was a big range in what they were taught depending on where they were from. Students from blue suburbs of cities in Texas, for example, reported a fairly middle-of-the-road education. Students from somewhat more conservative-leaning parts of Kansas reported much the same. But the students who had gone to schools in the suburbs of New York, LA, and San Francisco reported two things. First, that they had a very identity-focused education, to put it as neutrally as possible. Some of these kids agreed with that education and were shaped by it; some had rebelled against it in various ways. But they agreed that it had really been the prevailing focus of everything in any social science related subject they studied. Second, a good number reported that some of their teachers were highly ideological in a way where it was clear you had to agree with them to get a good grade.

That’s the thing that really troubled me. I think it’s fine for teachers to betray their worldviews in some ways. I try in the classroom to always play devil’s advocate and present the strongest version of different views. But I’m also a public figure—my students can easily figure out what my views are and I don’t hide them in the classroom. What is really important to me is to empower each student to think for themselves. One way I put this is in how I give advice on writing an essay: if I already agree with your thesis, I want to make sure it’s the best version of that thesis. A lot of the time I’m not going to agree with your thesis at the beginning or at the end. What I’m going to judge the essay on is whether it moves me towards your position. Do I think, at the end of reading this essay, that there are arguments here I should grapple with, even if I’m unlikely to change my mind? I’d much rather you put forward a thesis I disagree with, smartly and in a logical, coherent way, than that you echo back something you think I might believe.

The students I spoke with were not particularly exercised by what they had been taught one way or the other. But you could see the seething anger in those who had felt that teachers were highly ideological—where you knew you could write really badly and get an A if you reflected back what the teacher believed, and you knew that no matter how well you wrote, you would be in trouble if you wrote something else. Sometimes these were students who actually agreed with their teacher and didn’t have a problem with the substance of their views, but they still chafed against that dynamic.

How did you experience in the classroom both what students were taught, and whether teachers empowered students to argue, to disagree, to put forward their own point of view—or whether they abused the power of the classroom?

Traub: In general, I found classrooms less ideological than people think they are. I remember a conversation I had with a teacher who said that when he was a college student, he wrote an essay defending the surveillance act that was passed after 9/11 and his teacher gave him a D simply because the teacher was a liberal and thought it was a bad thing. He said that stuck with him, and he would never do that to a student. His goal was to help students arrive at their own views. I found that to be the default position of teachers.

When I read a lot of the conservative critique of schools, I thought I needed to find out whether it was really true that there are whole parts of the country where schools are deeply ideological. So I went to some schools in Minneapolis. As it happened, the high school I went to was in the catchment area of George Floyd’s neighborhood, so it was profoundly affected by what had happened there. The principal—a person I admired in many ways, ambitious, active, and restless—was deeply committed to the anti-racist notion that all white people harbor innumerable unconscious racist impulses and that one of the purposes of education is to bring them out into the open. That was relentlessly part of the educational program. I said to her: do you really want to create a situation where a student who doesn’t hold that view is told that that’s probably a sign that he or she is racist? Her response was: what are you saying, that there’s no such thing as structural racism? So yes, that exists, and I have a whole chapter about it in the book.

One of the most fascinating things I saw was a meeting of this principal with her council of teachers, where she was saying they had to make sure they were not punishing students of color by giving them bad grades—that they should have the same fraction of good and bad grades among white students and students of color.

Mounk: That is a remarkable demand—a demand to not grade people on the basis of their individual work. The fact that a principal of a school is demanding this should be a national scandal. The fact that students on the basis of race are not going to be able to get the same grade in a public school is a national scandal.

Traub: It’s a terrible misfortune. The teachers pushed back and said: the white kids who go to this school come from a middle-class background, while the children of color come from a pretty impoverished background—they don’t have the same academic preparation and they’re not going to do as well. She backed off a little bit. At least it was a matter of debate and discussion. But it is true that she had these strong priors that made her insist on things that I thought were just wrong, and that showed me that the right-wing critique is not totally off base.

Mounk: Interesting. Was your impression that this is a subset of schools that are like that?

Traub: I didn’t go to that many of the kinds of schools you’re describing—well-to-do, very progressive suburban areas. That’s probably true of those. But here’s a counter example. I went to a school in New York City, in Long Island City, that was like the school you’d see in a TV show—kids of every color, race, and ethnicity. I never heard a word about identity. It was called the Academy of American Studies, and the premise of the school was to teach much more about American history, government, and economics than you would get in a normal school. Interestingly, it wasn’t the kind of well-to-do, progressive, largely white school that maybe your students went to—the kids who went there were working-class immigrant kids. Even though they lived in the very progressive culture of New York City, they were perhaps less inclined to think in identitarian terms than left-wing white kids. I didn’t encounter that kind of ideological atmosphere nearly the way I did in Minneapolis.

Mounk: For the record, the students I talked about were mostly non-white themselves. My impression is that they probably went to schools that were predominantly Asian, white, and high-achieving Latino, in quite multi-ethnic and multicultural parts of the country, but probably quite affluent districts. That is an interesting point of context.

In the rest of this conversation, Yascha and James discuss practical steps to reform education and whether boys and girls have different learning styles. This part of the conversation is reserved for paying subscribers…

Read more

Marc Lipsitch is the Berberian Professor at Stanford University, with appointments in Infectious Diseases, Biology, and the Center for International Security and Cooperation.

In this week’s conversation, Yascha Mounk and Marc Lipsitch discuss the good, the bad, and the ugly of gain-of-function research, how scientific incentives may be encouraging risky experimentation—and the recent outbreaks of Hantavirus and Ebola.

This transcript has been condensed and lightly edited for clarity.

Yascha Mounk: There are many topics I’m excited to talk to you about, and we’ll talk a little more broadly about the landscape of public health in the United States at the moment, which is interesting. But I asked you to be on the podcast because I’ve been trying to think through a subject that I find fascinating but don’t have much expertise on, which is gain-of-function research. To start off with, what is gain-of-function research?

Marc Lipsitch: “Gain of function” is a general term that means a biological experiment that gives an organism a new function—a new ability to do something, usually by a genetic change. The type of gain-of-function research that has been controversial and raised concerns among many people, including me, is what’s sometimes called enhancement of pandemic pathogens, or dangerous gain of function, or risky gain of function—there are various other names. It is a small category, but a particularly concerning one, in which the function that an organism gains is the ability to spread more easily or cause more disease in organisms we care about: people, crops, or other organisms.

The most notorious example of it was when researchers took a strain of influenza—a flu virus that had been causing sporadic disease in people and killing a large fraction of those it was infecting—and achieved the ability of that virus, through genetic and evolutionary manipulations, to transmit readily in the air. That strain was so-called bird flu, or H5N1 flu. The reason why that’s concerning is that if you do that to a strain of flu, there’s a risk that it actually becomes capable of causing a pandemic in people—a new outbreak. We’ve all experienced a pandemic recently, of coronavirus, and a flu pandemic would be equally, if not more, damaging. The idea of these studies is that the investigators are trying to understand what makes it possible for a virus like flu to become pandemic.

Mounk: At first, that just seems like a crazy thing to do. You’re taking a virus that’s somewhat dangerous and deliberately making it more dangerous by making it more easily transmitted or by making its properties worse. That sounds like a Bond villain, crazy scientist kind of thing to do. But the purported logic is that understanding these viruses better—understanding what kind of pathogens we might face—might allow us to prepare better for pandemics, to prepare vaccines. Talk me through in detail what it is that scientists are hoping to achieve with this research, which in their mind is very well-intentioned.

Lipsitch: The goal of the research, as described by the investigators, is to find ways to characterize viruses and determine which ones are dangerous, which ones might cause a pandemic if found in nature, and which ones probably aren’t capable of that. The idea is that by doing dry runs in the lab, researchers are able to compile a list of the characteristics or genetic properties of a virus that allow it to do that.

The critique from those of us who are skeptical of that is really about both the risks and the purported benefits. On the risk side, there’s good reason to be worried about the possibility that a strain of virus created that way could infect one of the lab workers and eventually transmit beyond the lab, that it could be stolen and used for deliberate mischief, or that it could enable people to figure out how to make a dangerous virus themselves. So safety, security, and the potential for misuse are all of concern.

In terms of the purported benefits, one of the arguments that I and others have made is that you learn potentially what’s true about that particular strain of flu. But there’s good reason to believe—and even experimental evidence to show—that if you do the same thing to a different strain of flu, you might not get the same result. You might even get the opposite result: it becomes less transmissible or less able to cause widespread harm. The generalizability of the information is very limited, and that has been at the heart of the debate.

Mounk: The way to think about this seems to be as a trade-off. There’s potentially information we can get out of doing this research—which is scientifically interesting in itself, but more importantly could be helpful in trying to prevent or treat a pandemic. That’s one side of the ledger. On the other side are all the risks: the risk that somebody gets infected and the virus we’ve created ends up infecting a lot of people, the risk that somebody might steal it, the risk that the knowledge we gain in the process could reach somebody trying to create a virus for purposes of bioterrorism or something else.

Let’s go through each side of this in a little more detail. You’ve started to talk about how understanding the particular ways in which the behavior of a virus might change may not generalize. What are the other purported benefits that people claim for this? What is the strongest case that defenders of this practice would make for why we absolutely need to be engaging in this work in order to gain scientifically valuable research that might actually prevent bad pandemics in the future?

Lipsitch: The case is that if we can do it in the lab, then we understand what nature will do. Nature is constantly trying things out with mutations happening and viruses spreading in animals. If we can somehow predict what will happen, then we can prevent it.

Mounk: Perhaps you can explain to laypeople why you are skeptical about whether doing experiments on one kind of virus tells us much about different kinds of viruses. But there are also a lot of other steps in this argument that you need to assume. If we figure out that the most dangerous potential viruses have a certain shape or set of characteristics, then our whole machinery of pandemic prevention needs to be able to take precautionary steps to prepare for that kind of pandemic—to pre-produce the relevant vaccines, or to plan for certain kinds of social interventions that would stop the spread.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

The part where I’m quite skeptical is that part. When you look at the experience of COVID, there was certainly a pandemic that roughly speaking was in line with what we might have expected. We’d gone through H1N1 and other kinds of viruses that behaved in roughly similar ways. It’s not like we were totally surprised by the characteristics of this virus, but all of that prior knowledge doesn’t seem to have helped us respond to that pandemic in a very rational way. Even if we know that the most dangerous kinds of pandemics are going to be roughly of a certain form and shape—and perhaps we have some degree of knowledge about that which we wouldn’t have without gain-of-function research—is that realistically going to put us in a position to prevent that pandemic, or to lessen its impact, in a way that goes way beyond what we might have been able to do without that specific knowledge?

Lipsitch: I largely agree with your skepticism, though I’ll push back on one piece of it at the end. Recently, with an undergraduate student from Harvard, where I was working until this year, we published a paper looking at the claim that finding viruses in nature is an important way to create vaccines as countermeasures. What we found was that over and over again, viruses have been found in nature and little has been done about it. Then, when there is a large-scale human outbreak—usually among people in the West, and especially among wealthy people in the West, and occasionally in other settings—vaccine development begins in earnest. Or when there’s concern about a virus as a bioterrorist agent, vaccine development begins in earnest. Simply finding a virus in nature, even with signs that it could be very dangerous, has not been enough to prompt significant development of countermeasures.

Mounk: Even if gain-of-function research shows that a certain kind of pathogen could be super dangerous—that it could be a huge problem for the world—and even if that prompts calls to develop a vaccine, that development is unlikely to actually happen. So the purported benefits from the scientific practice aren’t going to arrive.

Lipsitch: That’s an important distinction—what we studied was when people find new viruses in wild animals rather than in the lab. But the broader point is that we are pretty good at mobilizing resources when there are significant numbers of people at risk, especially in the developed world, and historically we are not very good at mobilizing resources otherwise.

Mounk: Vaccine development is an incredibly expensive, complicated, and under usual circumstances, slow process, so we can only develop vaccines against a relatively limited number of diseases. There are new technologies coming online—perhaps most importantly, mRNA vaccines, but also other kinds of approaches—that have the potential, if the regulatory environment is reformed in the right ways, to speed up vaccine development hugely.

Which way does that cut in this debate? On one hand, if we are becoming much faster at developing vaccines, then we need to know less about potential pathogens in advance, because what we should do is simply wait for them to emerge and produce actual human outbreaks. Rather than needing a five-year head start to have a chance of responding with a vaccine, we can develop those vaccines very quickly—and what we should be investing in is just the capacity to adapt and roll out vaccines at scale when that happens. On the other hand, perhaps if vaccine development has become much easier, it is becoming more feasible to pre-produce defensive vaccines against a whole range of pathogens that haven’t yet emerged in real-world outbreaks. So perhaps having knowledge about what kinds of pathogens could be more dangerous is now becoming more valuable than it was in the past, because we’re going to be able to do more about it. Which way do you think our progress in rapidly producing and adapting vaccines cuts in this debate?

Lipsitch: One of the other things we found in our study of existing vaccines and virus prospecting in nature was that there are numerous whole viral families—large groups of viruses that infect and cause widespread disease in humans—for which we have no existing vaccines. So on the broad scale, if we go beyond pandemics and look at anything that infects humans and spreads, there is a whole menu of things we could be doing to develop vaccines with the knowledge we already have.

That is not to say there’s no point in getting new knowledge. The question, to me, is whether the best way to get that new knowledge is by doing very risky experiments that tell us something narrow about a particular group of viruses, or whether there are alternative approaches to studying these viruses that would tell us something about which viruses to be concerned about, but without creating viruses that are themselves dangerous.

Mounk: The claim by people who defend gain-of-function research is that we need to understand all the potential kinds of pathogens there might be in order to develop vaccines against them. What you’re saying is that we already know there are a huge number of existing pathogens that could potentially cause pandemics, and we haven’t even developed vaccines against them. So how about we start developing vaccines against these known dangers before engaging in a really risky practice to discover other kinds of candidate pathogens against which we could pre-develop vaccines in a speculative manner?

Lipsitch: That’s right. Beyond that, there are many ways to study which viruses are potentially dangerous without creating new ones. If you want to understand what makes a flu virus dangerous, we already understand many of the components of that. We know what the major determinants of transmissibility in people are in a flu virus. So we can study individual proteins of the flu virus, ask how they work, how well different variants of those proteins work, and then look for those variants in viruses that we find in nature. It is not the same experiment and it doesn’t produce exactly the same result. But if you think about getting some large proportion of the information with none of the risk, it’s a pretty good trade-off.

One of the important points in thinking about which research to do is not to frame it as should we do experiment X or not, because there are hundreds of thousands of experiments that never get done in science. Every scientist is deciding to spend their money and time on this project and not that one, on this approach and not that one. Most experiments that would be interesting never get done. It’s just a question of where you allocate your resources.

If you think about allocating the resources that would otherwise be used for a gain of function experiment to the next most useful experiment—one that would give you similar information or a similar ability to develop countermeasures—it becomes very hard to say that the marginal benefit from the gain of function study is really enough to make it worth putting the world at risk. Not everyone agrees with that view, obviously.

Mounk: We’ve given a pretty fair hearing to some of the potential benefits from gain-of-function research, and I at least am convinced that those benefits are much smaller than some of the defenders of the practice claim. Talk us through in a little more detail some of the risks of this kind of research. The claim that defenders would make is presumably that this tends to happen in labs with high levels of security, and that therefore, even though the pathogens they create may be dangerous, the risk of them getting out into the world is quite low. Starting with that claim—how confident can we be that these pathogens really would remain contained in safe research environments?

Lipsitch: Many of the studies done in the West, at least, have been done in very carefully designed, highly skilled labs. There is a lot of global variability in laboratory safety standards—and that’s setting aside the whole issue of malicious use entirely.

On the safety aspects, much of the controversy that followed the discussion of Wuhan and the origin of the COVID virus highlighted—whatever you think of that controversy—that there have been much lower levels of containment for the same experiments in China than would be expected in the United States. In particular, viruses where you really don’t know the pandemic potential are not very well regulated.

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

There is also the fact that gain of function studies, which tend to appear in the most prestigious journals because they’re flashy, creates an incentive for researchers around the world to do these studies in order to advance their careers and get into the top journals—which is what all scientists try to do, and appropriately so. But we really don’t want a dynamic in which the more dangerous your experiment, the more prestigious the publication. The amount of biology we learned from those gain of function studies that appeared in Science and Nature—the top journals in the field—was comparable to many other papers that appear in much less prestigious places. I would argue they appeared in those places because of how shocking what had been done was.

Mounk: Give us a little historical context on lab leaks. We can come back to the question about the origins of COVID and the research undertaken in Wuhan. My understanding is that there’s actually a long history of significant accidental lab leaks. In the 1970s, H1N1 escaped, probably from a lab in the Soviet Union. There was an outbreak of smallpox in the United Kingdom. More recently, there were escapes of SARS in a number of labs across East Asia—in Singapore, Taiwan, Beijing, and other places. There were also serious biosafety failures at the CDC in the United States involving anthrax and other pathogens.

Historically, have we been really good at avoiding these kinds of lab leaks? Or is there a pattern of lab leaks repeatedly happening—not just in one country, not just in one culture, not just in one political system? The examples just reeled off span from the United States to the Soviet Union, from North America to East Asia.

Lipsitch: The list of prominent ones you’ve given is accurate, and there have been some bacterial pathogen leaks as well—there has been a bacterial pathogen leak in China in more recent years, and one of the foot and mouth epidemics in animals in the UK resulted from a lab leak from their highest security lab.

Humans are imperfect, and almost all of these leaks are not the result of a physical system failing. Very often the issue is that a person fails to inactivate a vial of pathogen but thinks they have, or switches the inactivated vial for one that is still live. It’s just human error, and that’s the hardest kind of thing to prevent. The skill of the lab is, of course, important. But the record is, as you say, one of repeated failures—and underreporting is acknowledged even by the people who run these labs. The incentives to report an incident or accident are very low. The Netherlands, for example, has published a paper assessing the risk of underreporting and finding it to be quite high.

Fortunately, setting aside COVID and the controversy over that, most of these incidents have not gone very far—the 1977 release of H1N1 flu being the most clear exception. But they do illustrate that infections happen and transmission can occur, as it did with SARS in one of the accidental releases in the mid-2000s. It is a real risk. Because the consequences are so high, we are just not good at thinking about that risk. Biosafety is a field mostly dedicated to protecting individual lab workers and the people immediately in the area. Biosafety on the population scale is a relatively new phenomenon, as we deal with more dangerous pathogens.

Mounk: From a philosophical perspective, that seems to make a real difference. We want to protect all scientists who work in labs and make sure they are not exposed to serious danger. But at the limit, you can say that scientists agree to work in an environment with highly infectious pathogens and, even though we should obviously do what we can to protect them, they understand there is some risk involved. They can make that decision for themselves. Here, however, we are talking not about exposing individuals to a risk they have knowingly taken on, but about thousands, or tens or hundreds of thousands, perhaps millions of people dying as a downstream consequence of choices they didn’t make and had no say in. That seems very different.

When you look at this trade-off—and that’s one of the hard things to do, one of the things that political scientists and commentators often fail to do—the potential dangers are very serious and very clear, while the potential benefits are minor and questionable. So we probably should be discouraging, or perhaps banning, this practice. Why is it that this practice is continuing despite how stark this trade-off is? Why haven’t there been serious efforts to discourage or ban gain-of-function research?

Lipsitch: That’s not entirely a fair characterization. Toward the end of the Biden administration, there was a large effort led by the National Science Advisory Board on Biosecurity, or the NSABB, which became an executive order from the Biden administration in May of 2024 that laid out a framework for very careful review of this kind of research. It was comprehensive and very much applauded by people like me and colleagues who thought it was a major step forward. It was not perfect—one of the major issues is that it’s very hard for the federal government to regulate work that it does not fund. It’s not impossible and it should be done, but it’s much easier to put guardrails around work that the federal government is itself funding.

It was a very good policy. Then, under the Trump administration, some of the same people were involved in developing an executive order that came out in May of 2025 that, had it come from any other administration—Democratic or Republican—I would have given a grade of around 95%. It put quite strong restrictions on what it calls dangerous gain of function and attempted to extend some of that regulation beyond federally funded work.

The reason I add an asterisk about which administration it came from is that even at the time, it was clear that regulation and discussions of research funding were much more complicated and destructive in many ways under this administration. When the actual implementation of that executive order happened, a number of studies were stopped that were being funded by the federal government but posed no reasonable risk—no reasonable person would say these were at risk of causing a pandemic. Some of them were types of studies that have been done for decades and were simply not what any of us have been concerned about. Some that were stopped might raise legitimate concerns, but I think it would be very hard at the moment to get funding to do dangerous gain-of-function research. The pendulum has now swung in the direction of collateral damage to other kinds of research that have no plausible risk. The Washington Post reported that the administration simply wanted more things banned because they hadn’t banned enough.

Mounk: There is obviously a much broader attack on research funding from this administration that raises serious concerns. But give me a sense of the kind of risk we are currently running from gain-of-function research worldwide. You’re saying it has become much harder to get funding for that in the United States—is there still some dangerous gain-of-function research going on in labs in the United States? Has there been an international understanding to scale back this kind of research, or is it ongoing in labs in China, Russia, Europe, and other parts of the world? To what extent have we actually been able to build a consensus that we should abstain from the most dangerous forms of this research?

Lipsitch: It’s really hard to say what’s going on because there’s no central tracking and because the definitions are different in different places. There was a very good article by Tony Mills at the American Enterprise Institute, about a year or two ago, called “How the Virologists Lost the Gain-of-Function Debate,” which made the point that the leave us alone and let us do our science attitude that some virologists were taking was really self-defeating and not sustainable, particularly after people had experienced a pandemic in their own lives and seen what it really meant.

Subscribe now

There has been some significant shift of opinion and more attention paid to the concerns about it, at least in this country. China has a relatively new biosecurity law that may represent some move in that direction, but there are a lot of things on people’s minds and this is probably not at the top of most people’s list. The sentiment is shifting to some degree towards seeing dangerous gain of function as a problem and something not to be done, but I don’t think we’ve quite reached the point of everyone agreeing to that.

Mounk: One of the elephants in the room is COVID. Early on in the pandemic, there was a pretty concerted attempt by public health officials and parts of the scientific community to rule out the hypothesis that it could have been caused by an accidental lab leak—which would likely be related to gain-of-function research going on in particular at the Wuhan Institute of Virology. Later, it seemed as though the consensus was shifting towards thinking that it had indeed been a lab leak. A number of assessments by American and European intelligence agencies strongly suggested this was the most likely cause of the pandemic. My understanding is that perhaps the consensus has shifted back somewhat, and some people are now more skeptical of the lab leak hypothesis than they were a few years ago.

What is your assessment of how likely it is that the most consequential pandemic in 100 years stemmed from an accidental lab leak? How should people who are not scientists think through this question?

Lipsitch: The best way to understand this discussion is that there is no definitive evidence either way. People who hold strong prior beliefs really have nothing more than that, because the evidence is so lacking in both directions. There is some circumstantial evidence in each direction. I am totally happy to have an honest discussion with somebody who thinks it’s 99% either way. It’s the 100% that is the problem—and opinion is roughly split between the lab leak hypothesis and zoonotic or animal origin. Certainty in either direction is just an implausible reading of the evidence. People can interpret the available signals in whatever way their priors allow, and that’s understandable. But anybody who is certain is simply dismissing all the flaws in their own evidence.

Mounk: Stepping back to a broader question: what should someone think about public health in the United States at this juncture? Going back to the pandemic, there was an obvious failure of public health. There are important things that public health did during the pandemic to keep the situation manageable, but on the whole, the Centers for Disease Control were built to prevent pandemics and coordinate a rational response to them. They had an enormous budget devoted to exactly this. In the first major pandemic in 100 years, there was all kinds of malfunction—from the difficulty of developing a reliable diagnostic test for COVID early in the pandemic, to the refusal to allow private labs to step into the breach during a crucial period, to often conflicting guidance about masks that changed rapidly from one moment to the next and then changed back.

Something seems to have gone wrong, and there is probably a serious need for reform. On the other side, the way in which the Trump administration is trying to blow up public health—changing guidance around vaccines for diseases like measles in ways that have already led to serious outbreaks of this highly infectious disease, and mistrusting all of science—seems to be sometimes aiming at the right problems but doing it in a way that is only likely to make those problems much worse.

How should somebody who sees the failings of some of our institutions, who takes the criticisms of them seriously, but who doesn’t want to throw the baby out with the bathwater, approach this debate? What kind of reform do these institutions need? Why are the reforms currently being imposed on them by the Trump administration going, by and large, in the wrong direction?

Lipsitch: As a first cut, there are significant problems with how we did public health during the pandemic, and almost everything that’s been done by the Trump administration would make it worse the next time. The biggest threat to our biosecurity right now is undermining trust in vaccines, which most Americans very strongly approve of. If you wanted to make it easier for an adversary to cause trouble in our country with biology, one of the best ways to do it is to undermine trust in the most beneficial public health innovation of the 20th century, which is vaccines. There are real problems, but shooting holes in the best parts of public health is a catastrophic way to address them.

In my darker moments, my sense is that this is simply about shrinking government down to the point where we can drown it in the bathtub, as Grover Norquist once said, and that it has nothing to do with reforming public health. I don’t think that’s the motivation of everybody involved, and I think there are some things that genuinely need to be done. The diagnosis that we made major mistakes in the pandemic is correct. Many of those early mistakes were under the first Trump administration and were partly due to denying that this was going to be a problem when all the experts were saying very loudly that it was and that we needed to prepare. We had a warning in December of 2019 that this could be a problem, and certainly by January it was a three-alarm fire. We didn’t get into gear nationally fast enough because there was too much denial going on.

Mounk: The story of who was taking it seriously and when was a little more complicated. There was a period in January and February where a portion of the Republican party was taking it quite seriously—Tom Cotton or some other senator was really warning about it. Meanwhile, Nancy Pelosi was giving press conferences in Chinatown in San Francisco saying that anybody who was concerned about this was essentially being racist against Chinese people, and that people should continue going out to restaurants. That was in part a response to some genuine craziness where Asian Americans were being discriminated against in a totally abhorrent way. But in January and February of 2020, the partisan politics of this were somewhat unclear, and large parts of the Democratic party were really downplaying it. It was around March and April that the very dangerous dynamic got set into place—where Trump wanted to keep the economy going and downplay the danger of the virus, and Democrats became more in favor of social distancing measures and closing things down.

Lipsitch: The details of what everyone said are hard to remember. There was a current of anti-Asian racism at that time, and I attended an event in Boston’s Chinatown in January or February that made some of the same points. But saying that this is not an occasion to express racism is a different claim from saying this is not a threat to public health that we need to be taking seriously. Some of us were able to make that distinction, and if certain politicians couldn’t, that’s a failing on their part.

Mounk: In general, we would all profit from politicians who are better able to make those kinds of obvious distinctions, which somehow they don’t seem to be.

Lipsitch: The CDC is an institution that can benefit from reform. I was working there part-time from 2021 until 2025 and saw a lot of great things, but also saw that the ability to explain ambiguity to the public has been lost to some degree. I remember from the 2009 flu pandemic there were really excellent examples of that from some of the top officials. There’s now a sense that conveying ambiguity is too hard and that overstating certainty is preferable. On the communication side, there is a lot to be done.

The same kind of conservatism is part of why the CDC didn’t allow other labs to do testing for COVID for a period. It’s an understandable impulse in public health—to try to discourage people from drawing their own conclusions—but not a good one. There is a lot that could change at the CDC. But undermining vaccination, which is really the one thing that doesn’t need to change except to become even more widely used, is a sign of bad faith and destructive impulses rather than reconstructive ones.

Mounk: The point about communication is really interesting. I never had the impression that there was bad faith involved in any of these failings. I don’t think that at any point important public health officials were lying or misleading the public out of commercial interest or because they were in the pocket of Pfizer or anything like that. I did often feel that they treated the public like children—that they thought, for the good of the country, we need to induce certain behaviors, like staying at home or not buying up all the masks that might otherwise be used by medical personnel when personal protective equipment was incredibly scarce in the early stages of the pandemic. The impulse was to communicate in such a way that better outcomes for everybody were more likely to come about. But there was neither a reckoning with the fact that people can smell dishonesty and that this approach was not necessarily likely to induce the behavior they were hoping for, nor a reckoning with the second-order effects—that if people feel misled about one thing, they are going to be much less likely to listen on the next thing, and trust erodes over time.

One of the things in the debate about masks and the early guidance that masks were not useful—that I think public health officials didn’t think through—was other disciplines of social science, like economics. The earlier the market gets the signal that there is going to be large and ongoing demand for masks, the faster factories producing other kinds of things could retrofit to produce more masks. Sitting in Atlanta, somebody at the CDC probably wasn’t thinking about that.

Is that a fair account of some of those communication failures? How can these institutions do better? How can they warn citizens about real dangers and exhort them to engage in the behaviors most likely to save the largest number of lives, but do so in a way that takes citizens seriously as reasoning adults who can make decisions for themselves?

Lipsitch: That’s exactly right, and I agree fully that this was not about financial corruption or anything like that. I would add to the list of concerns that it is fundamentally undemocratic—or anti-democratic—to deceive the citizenry, and that’s a problem in itself.

We actually need more research on how to effectively communicate in public health. If there is good research on this, it’s not being used, because we continue to see the failures. My hypothesis for such research would be that in any given moment, you are operating in an environment shaped by your previous actions. If you have been communicating with the public under normal circumstances in a way that suggests you always know exactly what’s going on, that you won’t revise your views, that the science is all settled—then people will be genuinely surprised when you say you don’t know. In fact, on many scientific and public health topics, there is real uncertainty. Very few things change as fast as a pandemic, to be fair. But there is uncertainty about what the best diet is, uncertainty about the timing of flu in a given year, uncertainty about a whole range of public health topics.

People deal with uncertainty in their lives all the time and don’t think that the person who acknowledges uncertainty is an idiot or doing their job badly. Even when a doctor says a treatment has a risk of helping and a risk of not helping, people can deal with that. The key is to accustom people to expect that public health, like all human endeavors, involves some science and some uncertainty—and that especially in fast-moving situations, guidance is going to change. Not because anyone was lying, but because the best understanding changes and there will be disagreement. All of that is something people could get used to, but it would be very hard to absorb in the middle of a crisis, because people don’t react well to crises. It’s a long-term project.

Mounk: The slogan that most galled me during the pandemic was “believe the science,” because I do believe in science and put great trust in it. But precisely what distinguishes science from systems of blind faith is that in science, no point of view is sacrosanct. There is genuine debate about most propositions, and when somebody comes in with a paper that challenges a long-held assumption in the field, the right way to respond is neither to dismiss it nor to blindly believe it, but to probe how strong the evidence is and then come to a conclusion about whether to revise our views. The slogan “believe the science” was so often invoked in a way that was actually against the spirit of falsifiability and rational inquiry—which is the very reason we should provisionally trust a scientific consensus in the first place.

The second thought comes from my own teaching practice. There are many insecure teachers who hate to be asked challenging questions by students because they fear embarrassment or loss of authority if they can’t answer. I love it when a student asks me a question to which I don’t have the answer, because when I can genuinely and authentically say, “I don’t know—do you want to look it up? Or I’ll look it up and come back next week”—I can feel the trust in everything else I say go up. I’ve just signaled that when I give an answer, it’s because I’m pretty sure I know it, and when I don’t know, I’ll say so. That actually increases the trust people have in a professor, at least in the classroom.

Lipsitch: Some of the smartest journalists about infectious disease have the same instinct. Helen Branswell, who is perhaps the smartest one that I know well and writes about infectious disease, once said that she doesn’t trust anyone who never says “I don’t know.” That’s exactly right.

Going back to the “believe the science” point—the phrase I heard more often was “follow the science,” and I think that’s even more problematic, in the sense that science tells you what is, not what you should do. There are consequences for what you should do that follow from what is, but the science doesn’t say you should stay home. The science says that if you don’t stay home and other people don’t stay home, there might be significant consequences for disease spread.

My wife and I and a colleague wrote an article in the New England Journal of Medicine in July of 2020 calling for primary schools to be reopened. My wife is an educational ethicist, this colleague is an infectious disease pediatrician, and I am an epidemiologist. That call was based on science—on an emerging and imperfect understanding that children were not major transmitters—and on the view that education is an important good not worth sacrificing on that scale for some benefit in reducing disease transmission, particularly given the other choices we were making at the time. People could debate that, and in fact they did. Most schools did not reopen in the fall of 2020. It’s not that the science says you must do that. The science says here are the consequences as best we understand them—and that understanding was changing rapidly.

Mounk: I went through a version of this myself. I wrote a viral article early in the pandemic encouraging the cancellation of large-scale events—this was, I think, March 6, 2020. At that point, we were still having mass concerts and mass political rallies. The debate wasn’t yet about whether there should be government restrictions on movement; it was just about whether we should be cancelling a lot of these privately held events. I got a very angry email from an old friend who had also been a boss of mine for a while, saying I wasn’t an epidemiologist and should stay out of it. My answer was that I’m a political theorist, and I was basing the factual assertions I was making on the best statements by people with expertise in virology. I wasn’t pretending to understand something about the properties of the virus that they didn’t understand. But the question of what consequences to draw from that is an ethical question and a political question, and as a citizen I should be able to speak to that—and as somebody trained in reflecting about normative issues, I actually have some professional expertise in doing so.

Later on, I was among the earlier people to say that now that a lot of people were vaccinated and we knew more about the properties of this virus, we should start reopening a bunch of social institutions. The people who liked me the first time disliked me the second time around. That is one of the hazards of writing.

There’s a lot of ethical reflection in the field of public health, and it often comes from one particular disciplinary angle. A lot of attention, for example, is given to the precautionary principle—ideas that I think would be very controversial in philosophy departments, where many moral philosophers would be far less convinced this is a black and white issue than the consensus in public health seems to imply. These ideas then guide public policy in ways that can backfire.

One example is that we might have been able to get a vaccine even earlier than we did—and it is an achievement how quickly we did get it—through human challenge trials. Those were ruled out as unethical. Yet if some people had been willing to expose themselves to a pathogen that was dangerous but unlikely to kill any one individual, we might have been able to save hundreds of thousands of lives. It is not at all obvious to me that that would have been unethical, but there seemed to be a relatively broad consensus within public health that it was.

More broadly, there was a very interesting story about somebody designing a custom-made vaccine for their dog’s tumor in Australia with the help of a number of AI systems. They said that over 50% of the work they put into this effort was filling out the enormous amounts of paperwork required to get permission to inject a novel drug into a dog that already had a likely lethal form of cancer. Do you think that in the field of public health, while in some areas like gain-of-function research we are continuing with really dangerous practices, in other areas—like human challenge trials, or the bureaucratic burdens for getting drugs to patients who are very high risk of harm in any case—the rules are too strong to actually maximize the potential benefits of medicine?

Lipsitch: It’s interesting you bring up human challenge trials, because with a moral philosopher, Nir Eyal, and another epidemiologist, Peter Smith, the three of us wrote one of the first articles advocating for human challenge trials early in COVID for exactly the reasons you describe. In retrospect, the benefit would have been pretty small, because the United States, Brazil, South Africa, and the UK—where vaccine trials turned out to be run—provided the world with a massive global public good, which was a large enough epidemic to test vaccines in. We would rather have not provided that good, but we did.

Mounk: Because in this particular case so many people were infected in the first place, there was less need for the human challenge trial. One can imagine lots of other circumstances where human challenge trials could save a great many lives.

Lipsitch: There was some movement on human challenge trials—at least one member of the U.S. Congress was quite interested, and the UK did establish a human challenge model, though somewhat later. It’s a complicated example. But the field of research ethics as a whole is very precautionary and much more concerned about avoiding harm than about doing good. Some people have begun to push back against that. Alex London wrote a book about two or three years ago that pushes back in the right direction. It has been recognized as a problem, but the level of reflection about basic principles of ethics is not that high in an applied field—it’s more about how to apply what has been taught, and what has been taught is largely non-maleficence, beneficence, and justice. Those principles are interpreted, in the standard way, to mean it is better to miss the opportunity to do something great than to cause harm. The conservatism is a real thing.

In the rest of this conversation, Yascha and Marc discuss the recent outbreaks of Hantavirus and Ebola—how concerned we should be, how the general public may respond to pandemics in future, and when it’s time to worry about the current outbreaks. This part of the conversation is reserved for paying subscribers…

Read more

In this week’s episode of The Good Fight Club, Yascha Mounk, Amanda Ripley, Jesse Singal, and Thomas Chatterton Williams explore whether Trump has succeeded in remaking American culture in his image, the rise of white identity politics and its psychological drivers, and how America might break free from cycles of political revenge and backlash.

Amanda Ripley is the founder of Good Conflict. Her latest book is High Conflict: Why We Get Trapped and How We Get Out.

Jesse Singal is the co-host of Blocked and Reported and the author of The Quick Fix: Why Fad Psychology Can’t Cure Our Social Ills.

Thomas Chatterton Williams is a staff writer at The Atlantic. His latest book is Summer of Our Discontent.

Will you be in London on Sunday, September 6? I’ll be interviewing Francis Fukuyama about his life and thought to mark the publication of his memoir In the Realm of the Last Man at the Sekforde at 5pm. Find out more and book tickets here. Paying subscribers can access a code for free tickets here. —Yascha

This transcript has been condensed and lightly edited for clarity.

NEW: You can now watch the episode below! All episodes of The Good Fight are now on the Persuasion YouTube channel.

Yascha Mounk: It strikes me that we’re in a strange, frenetic moment in American culture that also is a kind of weird interlude. Around 2024 and 2025, there was the much-discussed vibe shift, which is part of what got Donald Trump elected. When I look at other countries in which figures who resemble Trump in certain ways came to office, some of them were really able to enjoy a kind of cultural victory—imposing their values on a large swath of the public and becoming the default mainstream way of thinking about the country. I would argue—and I’ll see whether you disagree with me—that in the second year of his presidency, it is already quite clear that Trump has failed at doing that. Whatever the vibe shift felt like 14 or 15 months ago, with a lot of young people, a lot of Latinos, et cetera, voting for Trump, it does not feel like Trump is remaking all of America in his own image. Obviously his administration is using executive orders and his power very aggressively to change institutions and do all kinds of things, but it doesn’t feel like Trump is managing to remake the country in his image. So where is American culture going? Thomas, how should we think through this?

Thomas Chatterton Williams: Politically, I agree with you, but on a cultural level, I think one of the lasting impacts of this new Trump era—Trump 2.0—is that white Americans have gone all in on claiming a standpoint position for themselves. They are getting involved in the kind of battle royale of identity and declaring that they’re not neutral, that white culture and white identity is not the atmosphere that everybody moves in. White identity is racialized, and they want to advocate for themselves as such. Jeremy Carl’s testimony before the Senate, when he was asked by Senator Murphy of Connecticut what he meant by white culture and white identity, is instructive here. He said he was advocating for heritage Americans and that they had been really harmed by immigration. I think this is going to be a lasting change: white Americans—a diminishing majority, but still a large numerical part of the population—are now going to be participating in the game of anti-racism that was really installed in the past 10 to 15 years during the Great Awokening. I think that is going to be, lastingly, to the detriment of society.

Mounk: That’s really interesting. Eric Kaufmann has this point in Whiteshift where he talks about asymmetrical multiculturalism. The idea is that there’s this kind of encouragement of taking pride in your group for all of the minority groups, and Kaufman’s point was that that’s unsustainable—eventually, the white majority group is going to say that’s the way the game is played, and they’re also going to take pride of place. I think Eric felt that was a kind of natural downstream effect and we should just embrace that. I was always much more worried about it. But perhaps it turns out that he’s right about the natural tendency of society. That’s kind of what we’ve been saying, Thomas—that we’ve now gotten to the point where multiculturalism is no longer asymmetrical, and whites as an identity group are playing the same game. Amanda, that would be bad for a modern society, would it not?

Amanda Ripley: What we’re saying is that in some ways Trump has imposed potentially lasting cultural shifts on the country, and in other ways he has not. I think one of the biggest challenges of dealing with someone like Trump is that in trying to oppose him, you can end up playing the very same game and thereby perpetuate it. It’s very tricky not to do that.

More than anything else this cycle, the thing that I’m frustrated by in the way that we’re covering culture and politics is that we’re not talking about the psychology of it. The psychology of Trump is about the addiction to revenge-seeking. It’s not just Trump. This is why Fox News is successful. It’s why many media outlets are successful. It’s less about outrage and more about revenge. Revenge operates in a really interesting way in the brain, and we’re just starting to understand this. It is addictive. There is a way in which you can see everything as a grievance and then, feeling aggrieved, seek revenge. A lot of our media outlets and politicians on both the right and the left are now not just giving us grievances—which they’ve done for a very long time—but also serving up a revenge fantasy.

That’s what you see with the gerrymandering race to the bottom. That’s what you see with the coarsening of the rhetoric: if you did it, then I’m going to do it—this tit-for-tat cycle that leads to endless revenge. The research is also very clear that to get out of that, you have to play a very different game. You can’t just do the same thing, or you perpetuate it. The ways in which Trumpism is likely to endure have to do with that psychological piece as much as the bigger, more obvious changes he’s imposing on our norms and institutions.

Jesse Singal: On this whitelash question, I may be skeptical of the extent of it. Persuadable voters who voted for Trump voted for him for a lot of reasons, and I think a lot of those reasons are a little bit obscured to people like us. This won’t be news to the panel or to anyone listening, but voters just make these decisions in a very different way from the way we do. It’s less ideological and more vibes-driven. Looking at the graph of Trump’s disapproval rate on Nate Silver’s website, he’s having an extremely bad presidency from a public opinion standpoint. So I think if there were any real appetite for asserting a white identity—maybe I’m just skeptical—I think there are some bad actors in and around his administration who want that, and there are certainly some online influencers who want that, not least the vice president. The memes alone posted by DHS and others are very creepy and white nationalist-coded, and I say that as someone who is often skeptical of claims of racism. They’re bizarre and they harken back to an America that never really existed. But I think there’s this problem that a lot of politicians have between telling the difference between what the people want and what their creepiest online fans want. I don’t think there’s a huge market for this stuff among the general public.

What we’re seeing is what feels like an exhausted democracy that’s going to keep barely electing people, then quickly getting mad at them, then moving on to the next person. It’s all pretty dark, but I think it’s maybe more complex than white identity reasserting itself.

Mounk: Let’s separate out two different things. One point is that the vibe shift, when it was happening, was remarkable in part because it went beyond white voters. The number of black voters who voted for Trump in 2024 was double the number who had voted for him in 2016. The number of Asian-American voters who voted for him was going up a lot. Numerically, the most important group that increased the vote for him was Latino voters. There was a feeling that the whole society was sufficiently disappointed in the Biden presidency and perhaps saw something in the second promise of Trump that made them willing to make that jump.

The second point is that perhaps we have to distinguish between the kind of white nationalist, creepy DHS images and so on—which I think is a very real theme of this administration—and a broader kind of social logic. I’d agree with Jesse that that’s certainly not the driving force of most of the people who voted for Trump. It’s just become such a part of American life—this natural reference to identity all the time—that perhaps it becomes natural for white people to identify themselves in that way. It doesn’t have to be as creepy as the DHS memes. The form that white self-identification takes doesn’t need to be as extreme as that. But I would still be quite worried if that becomes the default form of expression in the country in general, and particularly for the white majority group, even if it doesn’t take the form of the craziest meme on the DHS Twitter feed.

Williams: That’s right. I think that’s what JD Vance is trying to slip into the common parlance when he talks about heritage Americans—this desire to be able to speak as a heritage American and make claims to grievance that then get you special privileges or consideration in the zero-sum competition for status and prestige in the society. I think it operates along the lines of the logic of mimetic desire that René Girard talks about. After all of this time that groups have been advocating for themselves as a group, a lot of white Americans want that too now. That desire has become mimetically replicated among a lot of whites. I think Trump really did act as a kind of tribune for these people, even as he excited some working-class and downwardly mobile minorities and less educated minorities as well.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

Whiteness is really much more complicated in the 21st century than it was in previous eras of American history. One of the most prominent spokesmen for white nationalism is Nick Fuentes, who admits to his followers that he’s not only part Mexican but may very well have African ancestry. This stuff is complex.

Mounk: Jordan Bardella, who may well be the next president of France, is mostly an Italian immigrant to France. It’s already interesting that the leader of the far right is an immigrant from a neighboring European country. A little known fact in the United States is that he’s also partially of North African ancestry.

Williams: This stuff is going to get weirder and weirder. I think that Trump—and some of the voices of white identity would agree with Ta-Nehisi Coates on this—is the first white president.

Singal: What concerns me, in terms of trying to figure out how this happened again, is that some of this talk may distract from one of the biggest drivers of Trump’s second election, which was the border crisis. There was a genuine border crisis, and wanting a secure border is not just a white point of view. Most Americans want a secure border and are perhaps comfortable with some deportations, but are not comfortable with what Trump is doing. There is a lot of talk about the connection between Trump and white nationalism, and I get that, especially in this second and creepier term. But perhaps one of the lessons is that we simply need a normal, moderate immigration policy—which people in Biden’s orbit, by the end of his presidency, realized they had gotten catastrophically wrong. That framing brings the conversation down to something more tractable: what can Democrats do, and how should they message? The white identity framing, which I agree is menacing and weird, is perhaps a little tricky to pin onto specific political outcomes.

Ripley: What Thomas said earlier—about zero-sum thinking and the search for status—is important. That can take a lot of different forms. In our country, it is often about race; in other countries, it’s more about religion, class, or other things.

To bring us back to reality: Jesse, I think you’re right. Lee Drutman, a political scientist whose Substack I really love—he wrote Breaking the Two-Party Doom Loop a couple of years ago—was recently arguing that the biggest mistake Democrats are about to make is thinking that a backlash against Trump and the GOP means endorsement of the Democratic Party. This is what we keep seeing over and over: this nauseating swing back and forth. Public opinion always moves against the party in power. Every president since FDR who has had control of Congress has also lost seats in the midterms.

So then what? What does anyone who cares about this country and wants to get out of this never-ending cycle do? We know a couple of things. We know that about 80% of Americans believe we’re in a political crisis. Almost half the country now identifies as independent. Americans don’t just dislike both parties—they want more of them. This is what Lee has been arguing for a very long time: that one way to exit this roller coaster is to promise—whether you’re a Republican or a Democrat—that if elected, you will give Americans more parties and move toward proportional representation, which is actually achievable. It doesn’t require a constitutional amendment.

Mounk: Lee is an old friend and we were colleagues at New America for many years. I think the idea of proportional representation in America is deeply foolish for two reasons. The first is that the causes of political fragmentation are much deeper than the electoral system. You see it just as strongly in Germany, in Spain, obviously in Israel, obviously in Italy—countries that have systems of proportional representation, or, in the case of Italy, ran away from a system of proportional representation because of the chaos it was causing. More fundamentally, it is a complete pipe dream to think you can have proportional representation in the United States, because the only chamber in which you could have it is the House of Representatives. To have proportional representation in one chamber—the least powerful chamber—would probably mean five, six, or ten different political parties in the House, while the Senate would still be controlled by two parties, since by definition you’re electing one senator per state per election. The presidency would have a very bimodal outcome. I think that would make America far less governable than it already is.

The more profound question is this. I agree on the thermostatic voting—which is the political science term for this, though I think it’s a dumb term; I’m not sure what a thermostat has to do with it. The idea is that public opinion always moves against the president. Because of gerrymandering, the primary system, social media, and the donor class, each party always goes way overboard, putting in place what the activists want rather than what average voters want. Voters say they hate the people in power, give the other side a try, the other side goes too extreme, and then voters say they hate that even more and put the original people back. That is a huge problem and a huge danger for Democrats in 2028. At this point, I think Democrats will likely win the midterms and may well win 2028. I’m very worried about what is going to happen in 2032 if they get it wrong once again.

I have a more fundamental question: has the basis for any kind of social cohesion just gone away? In my optimistic moments, I look at the opinion polls on immigration, on America’s relationship to its history, on what should be taught in schools, on trans rights, and on all of these issues, and the average American has a very reasonable opinion. The average American is pretty tolerant and pretty sensible. Perhaps we can finally find somebody who actually governs in the name of that majority and escape this spiral of polarization. In my less optimistic moments, I think our society is now so fragmented—with so many different ideological tribes, so many people getting information from such different sources—that the basic mechanisms of our public discourse are driving people toward polarization to an extent that we’re never going to have a mainstream big enough to keep our parties together again. What do you think?

Ripley: I totally agree, Yascha, that our problems run way deeper than a winner-take-all system. In many countries that have proportional representation, you still see very serious polarization. On average, countries that have multiple parties and proportional representation are less polarized than countries that don’t—but there are so many factors there, to your point. It’s not the solution. It’s one of many things we need to do to get outside of the game we’re in. We can’t keep doing this. We have to do something very different that Americans can see is very different, because average Americans are not represented in Congress. That’s not how primaries work. One of the reasons primaries are the way they are is because of gerrymandered districts.

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

The phrase gerrymandering comes from a Massachusetts Democrat whose last name was actually Gerry, who carved up the districts into something that a political cartoonist thought looked like a salamander—combining those words to make “gerrymandering.” This has been going on for a long time. But there are three big differences now. One is the sophistication of the mapping and the tools used to carve up districts. The second is our profound contempt for one another—these revenge cycles that make us very vulnerable to thinking the only option is scorched earth and extreme measures like gerrymandering. The third is the Supreme Court, which has abdicated any role in keeping campaign fundraising and gerrymandering under control. A lot of these things need to get fixed. One way to prevent the splitting we are doing—dividing the world cleanly into good and evil—is to mix it up and have more than two choices.

Mounk: Jesse, Thomas—you can choose to speak to this more narrow institutional debate or to the broader point. Is there still a kind of American mainstream, a reasonable voter that some political force could successfully speak to and mobilize? Does the center still hold in American society more broadly?

Williams: Whether the center holds is really dependent on having the right kind of politician who can articulate ideas in a way that reaches that broad and sensible center. There is a center-left and center-right that’s not entirely comfortable with the extreme rhetoric and constant warfare of contemporary political culture. But it would require somebody who has a lot more talent than what we’ve been seeing.

Barack Obama was clearly able to reach Americans on a rhetorical level that inspired a certain amount of hatred and animosity, but also really inspired a lot of Americans to exercise their better tendencies, especially in the first term. He’s still quite popular. That’s asking a lot of a candidate, and I don’t see that kind of talent on the Democratic side at all. A lot of Americans could also be relieved to vote for somebody with the sensibility of Mitt Romney, if that were on offer from the right. I think a lot of people would like to turn the temperature down. What do you think, Jesse?

Singal: It’s interesting that a lot of the most charismatic politicians—and I’d put folks like AOC and Mamdani in that camp—have a lot of baggage from having said radical things, and they don’t really do the Obama style. They’re probably capable of that sort of rhetoric, and I think we’re going to see them shift toward it as their ambitions get higher. Obama is a generational political figure, but to answer the question: we’re not so broken that we couldn’t see another Obama type who talks in sweeping, unifying terms from a center-left perspective. In other ways, though, we’re experiencing runaway fragmentation that is not going to get better. The idea of a truly shared culture—beyond the Super Bowl and Bad Bunny—seems increasingly remote. There will be a few super-culture attractors that most people know, but I don’t think we’re going back to any sort of shared culture in terms of people’s day-to-day consumption of music, art, political figures, and so on. To the extent there are superstar political streamers, they tend to be hyper-partisan, which is a problem. So mostly I’m pessimistic about this whole America thing.

Mounk: There are surveys which show that in the 1980s, when you asked people who the most famous actor was, the answer was someone in their 20s. The same was true in the 1990s. Then the age started going up. The reason is that when Matt Damon, George Clooney, and those figures were first in the big movies, everybody was talking about those movies—they were central to American culture, and so those actors reached a degree of celebrity that goes beyond what Timothée Chalamet now enjoys. The common basis for that shared culture has eroded over the last few years because of technological developments.

I want to segue into the second topic I loosely thought we would talk about today, which is how the Democrats are reacting to all of this. We have the failed vibe shift of a few years ago and the recognition—which I think is valid—that Trump is actually starting to fade. He’s perhaps managing to remake a part of Americanism, but he’s not winning an all-out victory where it feels like everybody is now on board with Trump. There were a few months when it felt like that might happen. I think that danger is now quite clearly banished.

Subscribe now

How have the Democrats reacted? During the first Trump administration, part of the reaction was to fully embrace a new identity ideology on the left that felt very new and exciting and was able to sweep through institutions quite quickly. This time around, it doesn’t feel like Democrats have explicitly gone back to fully fighting for woke. In some ways they’ve superficially surrendered on some of those issues. It’s surprising that not many Democrats are going around saying the first thing they will do is nominate Supreme Court justices who are going to put affirmative action back in place. Perhaps that is in fact what they would do, but it’s not a prominent calling card.

At the same time, it feels like they’ve inverted Michelle Obama’s line—”when they go low, we go high.” One of the lessons the Democrats seem to have taken from the Trump era is that he won because he went low, and this time they’re going to go low as well. We see them campaigning with people like Hasan Piker, who is, in the old-fashioned description of the left, someone who actually defends the dictators of tanks rolling over other countries. We see Democrats making excuses and getting very angry at people who criticize them, like Graham Platner, who had a Totenkopf tattoo on his chest for over ten years before it became a campaign issue. You even see it in some of the institutional fights—on gerrymandering, the attitude seems to be: if the Republicans are gerrymandering as hard as they can, we’re going to gerrymander as hard as we can as well.

How should we think about the evolution of the Democrats? Is it fair to say that’s the lesson they’ve taken, and how do we feel about that?

Williams: It certainly is the lesson that Gavin Newsom has taken, at least stylistically, and I find that extraordinarily off-putting. The strength would be to model the kind of political behavior that many of us would recognize from the past and miss, as opposed to mimetically replicating Trump’s online behavior—the trolling and the extraordinary vulgarity. I find that deeply depressing.

I’m a little more agnostic about whether you have to accept that people are flawed and find ways to reconcile with previously flawed candidates if they have a chance of winning and have rejected the previous behavior you find problematic. Someone like Graham Plattner doesn’t stand by the Totenkopf. He says—whether we believe it or not—that it was a mistake, that he got rid of it, and that it’s not what he stands for now. Realistically, there has to be a path towards reconciliation, especially when somebody has momentum, because the alternative is that you don’t support him and you empower Donald Trump to have one more advocate in the Senate. There’s not one response to that question—it depends on the situation. Gerrymandering also presents another set of calculations.

Ripley: It’s hard to generalize about Republicans or Democrats, but if you look at some of the loudest voices—House Minority Leader Hakeem Jeffries keeps referring to maximum warfare everywhere all the time when it comes to gerrymandering—this is literally, as Thomas said, mimicking the behavior of your opponents. It doesn’t get you out of the trap you’re in.

Plattner is a really interesting example. We know from the research on human behavior in these revenge addiction cycles that one way out—and I know this is going to sound incredibly squishy, and it sounds squishy to me too, but it is backed up by the research—is that you have to start forgiving people. That is the only way out. You have to do it even internally. Even if you never say it out loud, it actually has an effect on that kind of addiction cycle. Plattner has spoken enough to the whole issue of the tattoo. We just can’t keep doing this.

Mounk: Has he, though? When Barack Obama had the most serious crisis of his primary campaign in early 2008, tapes of Reverend Wright were released in which he was saying “God damn the USA,” et cetera. Obama gave a major speech in which he laid out his relationship to race. It wasn’t Obama who was on tape saying these things—it was the pastor of the church he attended. He said: let me take this really seriously. I understand why a lot of people are going to be upset about that, and they will be right to be upset. Even so, he laid out what the pastor’s perspective was and where he came from. He threaded the needle in both a beautiful speech—which I often assign to my students—and in a serious way: not hiding who he is, not debasing himself, but explaining it and recognizing why it was so upsetting to people.

I have seen nothing like that from Graham Plattner. All I’ve seen from him is the claim that this is trolls on the right weaponizing something he didn’t know about. His story does not add up. His chief of staff has said that he referred to it as “my Totenkopf” years ago, and yet he later claims he never knew what it was. He hasn’t explained that in a satisfactory way.

I’ve said many times that I’m against cancel culture. But being concerned about a candidate who literally had a tattoo of a symbol associated with the SS on his chest for ten years, covered it up once it became a national political issue, and never addressed it seriously—that’s a different matter. Ninety percent of what he has said about it is that this is just right-wingers trying to exploit it against him. I’m open to him perhaps one day being a great senator, and I’m open to forgiveness. But has he actually, in a serious way, explained what it is, where it comes from, why it’s upsetting to people, and what it means for the Democratic Party? Aren’t we being deeply hypocritical? If somebody on the right had a Totenkopf tattoo for ten years, would we really be saying it’s perfectly natural to forgive that?

Singal: It does seem disingenuous. David French complained about the slide in decency and standards, writing that “the slide begins when you tell yourself that the stakes are just too high for normal politics”—that of course you wouldn’t support this candidate in better times, but now American democracy is at stake. He’s saying: don’t go down that road. I understand the argument given how tight the Senate is and what one seat means. But I find it demoralizing that people act like it’s a crazy cancel culture campaign to be concerned about a Nazi tattoo. In many cases, it’s the same people who, before the vibe shift, were calling for people’s careers to be ruined over much lighter offenses. There’s a degree of disingenuousness on display that I just don’t like.

Mounk: The same thing applies to Hasan Piker, where people completely mix up two different things. When somebody has a big media platform, should you go and make your case in front of that platform? Sure. Should Democrats go on Fox News? Should they go on Joe Rogan? Should they go on those shows and be authentically themselves and try to reach those audiences? Yes. If somebody wants to go on Hasan Piker’s stream and argue for their political views, I don’t have a problem with that—even though I am deeply concerned about Piker’s views and have some strategic questions about whether that is actually going to be very useful, given that his audience is very different from the mostly apolitical audience of Joe Rogan. But let the consultants make that call. That’s different from campaigning with him. That’s different from saying he’s the future of the Democratic Party.

Ripley: Litigating the tattoo back and forth at this point is not super helpful. What I find revealing about Plattner is that while he has apologized for that—and I think in a more believable way apologized for some of the things he said about sexual assault in the military—he still uses the tool of revenge-seeking. It’s just different subjects, different scapegoats, different targets of blame, but it’s still the same mindset of splitting the world into good and evil, with the conviction that you are on the side of good and that anything is justified because the other side is so evil. That’s the piece of his behavior today that gives me pause. I know people who have gotten idiotic tattoos in their 20s that they didn’t know the meaning of. We perhaps can’t expect an Obama-level reflection from this guy, but we could expect people to start thinking a little more critically about this constant cycle of oversimplifying and blame.

Williams: I understand the arguments that Yascha and Jesse just made and the point Amanda is making. But at the end of the day, if you are actually concerned about things like anti-Semitism—and if you have any familiarity with German culture—look at the memes coming out under the banner of the White House and the Department of Homeland Security. These are not things being rejected or covered over. They are deliberate emulations of actual Nazi propaganda and rhetoric. If anti-Semitism is your concern, Graham Plattner is not where you should be channeling all of your outrage. Unfortunately, in some situations, we do have a binary choice.

Mounk: That’s where Amanda is saying: why do we debase our standards in that way? Why can’t we both be outraged at the DHS Twitter feed and the symbolism of Trump and all of that, and also say—excuse me—the party that was, as Jesse was saying, canceling people, is now telling us how dare we be upset about a man who had a Totenkopf tattoo for ten years? I wrote about an electrician in San Diego who was actually of Latino heritage, who had his hand dangling out of a truck and somebody thought it looked like the OK sign, which they took to be a white supremacist gesture. He lost the best job he ever had. Those same people are now telling me I have no right to be upset about Plattner. Excuse me if I find that a little galling.

Williams: There’s hypocrisy to go all the way around and excuse me if I got Amanda’s point wrong. I just solely disagree with Jesse.

Ripley: People will justify anything because they want to feel right and because the other side feels threatening. It never ends. I do it too. When I think about one of the things that has most made me ashamed of this country—a country I was born in, am a citizen of, and love—it was the way we withdrew from Afghanistan. At the time, I was angry at the Biden administration about that, but I didn’t necessarily blame Biden personally in the way I now very quickly catch myself blaming Trump for most things personally. You can catch yourself doing this, and the trick is to at least notice it and try—as Yascha is pleading with us to do—to reflect on the ways in which we are debasing ourselves, extending a huge benefit of the doubt to our own side while denying it to the other.

Mounk: Two more thoughts on Maine, and it’s not about the tattoo. The first is that it shows the really terrible field in so many of these Democratic races. On the one hand, you have a candidate who doesn’t have any very clear achievements, has a history of erratic political opinions, and looks like the image that people drinking coffee in Brooklyn have of what appeals to the country—in a way that I’m not sure it actually does. On the other hand, you have a governor who by all accounts is perfectly competent and decent, but who is 78 years old and has no charisma and no real ability to articulate what the campaign is about. Given those two choices, I understand where some primary voters ended up going for Plattner. I felt something similar on the day of the midterms when a couple of pretty reasonable governors were elected in New Jersey and Virginia. I tried to watch the victory speeches and they were really quite boring. Meanwhile, Zohran Mamdani has a bunch of policies I disagree with, but he’s clearly a talented and charismatic politician. That’s a real concern for where the Democratic Party is.

The other question I want to raise is whether “when they go low, we go low” is actually a sensible electoral calculus. It relies on the idea that what drives Trump’s appeal is that he goes low, and it’s not clear to me that that’s what people like about him. Perhaps it is, but it deserves thinking through. If you embrace the same style without the same charisma, the same political cause, or the same political coalition, it’s not obvious that’s going to benefit you. Gavin Newsom going around signaling his social media dominance clearly appeals to the base, but is that going to win over independents in the way that Trump, for whatever reason, was able to in 2016 and particularly in 2024? I have my doubts.

Williams: For the people that style does appeal to, they can get it from Trump, and he does it better than anybody who’s trying to mimic that behavior—he certainly does it better than Gavin Newsom. So if that’s your thing, replicating it will not convert anybody to your side. What it can do is repel a lot of people who don’t like that quality in Trump but put up with it because they like certain other things he’s delivering.

Ripley: Is the lowness what people are actually attracted to? Some people are genuinely attracted to it. It feels authentic. It feels like this is a fearless person, because people feel—and they are not wrong—that the whole system is rigged. They’re finding the wrong solution, in my view, but they’re not wrong that there’s something deeply wrong, that there’s distortion and corruption and dishonesty in the dealings between the political class and the American people. So Trump’s lowness functions as a proxy for honesty and courage. It’s a false idol, and a dangerous one, but it looks courageous even when it’s deeply cowardly.

All sides can play that game and have tried, but as Thomas said, Trump is much better at it for a number of reasons, including his addiction to vengeance. This is not a healthy person, and it’s easy to forget that. Most of us were not involved in 4,000 lawsuits before 2016. Before he even ran for president, Trump was using the justice system as a tool of revenge. If you talk to any seasoned lawyer, they know people like this. In modern history, we do not have another example of a president routinely suing news organizations he doesn’t like while in office. He has sued Trevor Noah, the BBC, the Washington Post, and on and on. This is not normal behavior, but it is what Trump has always done, and his behavior is not going to change. He is a very fragile person, and lawsuits function like self-medication—a way to temporarily feel better.

That is the behavior we need to be looking for. The party affiliation matters less to me than the pattern itself: this conflict entrepreneurship behavior, and our collective failure to recognize it and become less vulnerable to it.

In the rest of this conversation, Yascha, Amanda, Thomas, and Jesse discuss the state of free speech in the United States, whether Trumpism will continue beyond the presidency of Donald Trump, and what America will look like in 2030. This part of the conversation is reserved for paying subscribers…

Read more

Alvin E. Roth is the Craig and Susan McCaw Professor of Economics at Stanford University and the George Gund Professor of Economics and Business Administration Emeritus at Harvard University. He was awarded the Nobel Prize in Economics in 2012. His latest book is Moral Economics: From Prostitution to Organ Sales, What Controversial Transactions Reveal About How Markets Work.

In this week’s conversation, Yascha Mounk and Al Roth discuss the impact of moral disgust on solving economic problems, whether we should allow financial payments for organ donation, and what the rise of OnlyFans tells us about changing attitudes towards the self and economic transactions.

This transcript has been condensed and lightly edited for clarity.

Yascha Mounk: Economists have a lot of influence on our lives, but a lot of the time that influence is less direct. As we’re recording this, there was a decision by the Federal Reserve to keep interest rates constant—and that’s going to influence the lives of people listening to this podcast in all kinds of complicated ways. But you have had influence on the world in a much more direct way. If some of the people listening to this are doctors in the United States, where they went to residency was probably decided by a procedure that you helped to put in place. If you have children who go to public school, there’s a good chance that which public school your children ended up in was decided by a mechanism that you helped to put in place. Tell us about the class of problems that you started to tackle in your academic work, and that then had this huge influence on the world—how to allocate scarce goods to people under particular kinds of circumstances.

Al Roth: Economists study marketplaces, broadly defined, and game theory gave us a way to study the rules by which marketplaces are organized. That opened up the possibility that economists could start studying the design of markets—and I’m using the word “design” there as a noun. Markets have designs, and you can study them. After I’d studied some market designs, I started to get asked to help design and redesign some markets. Now “design” is a verb. That’s part of a pretty new movement in economics called market design.

Of course, market design itself is an ancient human activity. People have been creating markets of various sorts, with or without money, for a long time. Stone tools distributed in the archaeological record, far from where they were quarried, tell us that our prehistoric ancestors knew something about trade. They knew something about markets. They could move goods far across the world. Markets are a little bit like languages—they’re tools that human beings build to cooperate and coordinate and communicate and compete with each other. Often we treat markets as if we just received them, but of course markets are built, and they can be fixed when they’re broken.

Mounk: Tell us about some of these examples. How should we think about what good or bad design looks like? Even when we’re not thinking about designing markets, you’re saying that laws, rules, cultural expectations and habits effectively design a market. But when you’re going about trying to design them in a more conscious, more explicit way, what kind of criteria should we apply? How should we assess whether they are doing a good job or a bad job?

Roth: Sometimes markets are broken in a very obvious way. That is, the people trying to transact with each other are having trouble making connections. That’s often where market designers get called in.

Mounk: Give me an example of that.

Roth: A lot of labor markets—like the market for new doctors—once had all sorts of problems with the timing of offers. People would get offers that they had to answer very quickly, exploding offers, before they knew what other opportunities they had. That market, in the first half of the last century, had unraveled to the point where doctors were getting offers of their post-graduation jobs very early in their education—too early for them to really know what kind of jobs they wanted, and too early for the institutions hiring them to be able to tell who was going to be a good doctor and who was not. That was a problem that the medical establishment tried to solve in various ways.

Today, there’s a centralized clearinghouse for how new doctors get their first jobs. I was involved in redesigning that when it ran into some problems. Sometimes there are operational difficulties in a market that make it clear it needs some help and redesign. Also—and this is part of the subject of my current book—markets need social support to work well. When you design a marketplace to operate in some big economic environment, people have other options. They could make their transactions outside the marketplace. So you have to attract people to participate and come to the marketplace. That starts to involve questions of equity and transparency—making the market work well so that people want to support it and participate in it.

Mounk: This speaks to a broader confusion about the relationship between economics and ethics, economics and morality. Since its inception, economics has sometimes been known as the dismal science—the science that is amoral, perhaps even immoral. There was a time in which economists claimed that the principles they followed were in some sense free of moral choices. I don’t think that is how most sophisticated economists think about these questions today. But your disciplinary hat is that of an economist, not that of a political theorist or a moral philosopher. To what extent are you making normative choices? Where do those normative choices come in? How explicit are they?

Perhaps you can tell us about a concrete system you were involved in designing or advising on, in which there was a moral fork in the road—where you had to explicitly choose one kind of market design that would serve one set of moral goods, or another kind that might serve a different set of moral goods, and it wasn’t obvious which was preferable on purely economic grounds.

Roth: One way to answer that is to say that this book on moral economics that’s just about to come out is my second book aimed at a general audience. The first one was called Who Gets What and Why?—a somewhat cheerful book about designing marketplaces where the questions were relatively straightforward, where they were technical, where it was about helping people make transactions that they wanted to make. This book is about markets that are also broken, but that are harder to fix because we can’t agree on what we want. These are morally contested markets where people have very different visions of how the market should work.

One thread that connects them is my work on kidney transplantation. Right now there are about 130,000 cases of kidney failure a year in the United States, but we only do fewer than 30,000 transplants a year. Transplantation is the treatment of choice for kidney failure. So most people who could benefit from a transplant—whose lives could be saved and whose healthy lives could be made much longer—are going to die without getting one. One of the questions was how to help people get more transplants.

A topic I talk about a lot in the current book, Moral Economics, is the widespread feeling around the world—codified in many laws—that it is illegal to pay a living donor to donate a kidney. Kidneys are a little bit special because healthy people have two and can remain healthy with one. So it’s possible, if you love someone who has kidney failure, that you might be able to save their life by donating a kidney to them. That happens about 7,000 times a year in the United States. But if someone wanted to pay you for a kidney, the two of you would be involved in a felony. It’s against the law to be paid for a kidney, although you’re allowed to give one.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

What I’ve done with my economist colleagues—and how we’ve helped our surgical colleagues—is make more efficient the idea of kidney exchange. Perhaps you want to give a kidney to someone you love and you’re healthy enough to do so, but you can’t give it to the person you love because kidneys have to be well matched and yours is incompatible with your intended recipient. Perhaps I’m in the same situation. It used to be that the two of us, as potential donors, would be sent home—sorry, but no thank you—and that would deprive our loved ones of living donor transplants. But now it might be that I can give a kidney to your patient and you can give a kidney to my patient. That’s a way in which each patient can get a compatible kidney from another patient’s intended donor. That’s called kidney exchange, and it has blossomed in many ways. A lot of transplants are now done that way that couldn’t be done before. An essential part of this, when you’re talking about morally contested markets, is that the donors are not paid. That avoids the repugnance that much of the world feels towards the idea of paying for kidneys.

Mounk: This is obviously an amazing win-win solution. It means that a lot more kidney transplants are facilitated every year through this process. By and large, people who have moral objections about the buying and selling of kidneys are perfectly fine with it because it still operates as part of a kind of gift economy—people saying they want to donate a kidney without a financial reward. There’s just a more complicated chain of transactions, which means that a donor can thereby assure that their loved one survives even if they are not directly giving them the kidney.

What about the underlying question? Should we in fact allow the buying and selling of kidneys? This mechanism that you were involved with is saving thousands of lives in the United States every year—a fantastic thing. I don’t often have a distinguished podcast guest of whom I can say with some confidence that they helped save thousands of lives in the United States and around the world every year. But as you’re saying, a lot of people still die from kidney disease without having received a transplant. Many of these people would be able to pay for a kidney. Perhaps if we collectively decided that buying and selling kidneys should be allowed, and health insurance covered that, even people without significant means who now die would be able to survive. How should we think about that moral question?

Roth: That’s a question I have a chapter devoted to in my current book, because although we’ve had some good achievements in increasing the number of kidney transplants, these are victories in a war that we’re losing. There are more people who need kidney transplants today, and the gap between how many transplants we can produce and how many people need them is growing.

A natural thing to think about is whether we can increase the supply of kidneys so that there won’t be almost 100,000 people on the waiting list for a deceased donor kidney, as there are today. When economists see a long queue of people waiting for a scarce resource, they wonder if prices could increase the supply so it wouldn’t be so scarce anymore. Kidney exchange increased the supply, but not enough to solve the problem. We still have this terrible shortage.

One of the things that people worry about is that we might not want, or be able to get support for, a market in which only rich people could get transplants by buying kidneys from poor people. A lot of the designs that have been proposed get around that by, for instance, proposing that only the federal government could pay for kidneys, and that we’d continue to allocate kidneys based on need—the way we now allocate deceased donor kidneys, not that that’s a perfect process either.

There is legislation being proposed in Congress that would allow at least a small experiment in that direction. The particular legislation currently seeking a hearing is called the End Kidney Deaths Act. It proposes simply to give a tax credit, over a period of time, to non-directed donors—people who donate a kidney anonymously to someone they don’t know, rather than having a particular patient in mind. Those donors help facilitate a lot of kidney exchange transplants because, as you alluded to, they allow long chains of transplants to be effectuated. There are about 500 such donors a year in the United States at the moment.

Mounk: To be clear, the much more common case is that a parent, child, or spouse needs a kidney, and a loved one happens to be a match—or, as part of a kidney exchange, is willing to give somebody else a kidney if it causally results in their loved one receiving one. The rarer case is that of a healthy person with two functioning kidneys who decides to save the life of a stranger by going to the hospital and having one removed. That is obviously a very altruistic act, and the number of people willing to do that is about 500 a year.

Roth: There is a proposal to experiment with whether that number might be increased if we were more generous to those donors through the tax system—a more grateful nation, so to speak. It’s an incremental proposal. On the other hand, if it substantially increased the number of non-directed donors, it would not only increase transplants, but would serve as an indicator that we might start to think more generally about being more generous to donors than we are now.

Mounk: Should we do that? I think this kind of proposal is interesting, but it’s quite limited. The broader question remains very urgent. Even if this proposal means that 700 rather than 500 people a year give a non-directed kidney, that doesn’t seem to actually grapple with the most fundamental problem—the tens of thousands of people who don’t get a kidney every year. Should we abolish any limit? Obviously it needs to be regulated, but should we in principle allow the buying and selling of kidneys?

Roth: If an experiment like the one proposed by the current legislation works well, then we’ll be able to garner support for broader programs of generosity to donors. As a practical matter, even small legislative steps like this have failed in the past. So I think it’s probably necessary, as a matter of practical public policy, to start small—with these very generous non-directed donors—to see if we can express our gratitude to them for all they do. Opponents of paying for kidneys worry that it will lead to the exploitation of poor people in ways that we would regret. But if that doesn’t happen, then I would certainly be in favor of pursuing a much broader program.

Mounk: Let me go through some of the obvious objections to this scheme. One you just mentioned is the exploitation of the poor. Presumably you could also imagine terrible cases of exploitation in a system where you can’t buy and sell kidneys. One of the things a free market often does is alleviate scarcity. At the moment we have an artificial scarcity in kidneys because it is impossible to pay people for them. Presumably, if we had a liquid market in kidneys, it would also be much less dangerous to donate one. One concern is: what if I donate a kidney and my remaining kidney later becomes diseased? In a world where buying and selling kidneys is allowed—and where, if you’re poor, the state steps in to ensure you can get access to a kidney—that danger would be much lower, because I could in that circumstance receive a kidney myself. You could also imagine cases of exploitation arising precisely because of the current scarcity. If you have a loved one who says they will die unless you donate your kidney, there can be enormous pressure exerted on people—in a life-or-death situation—over whom they might have some power or influence. That danger too would be mitigated if you had a liquid market in those organs.

Roth: The whole point of making kidney transplants more available is to make kidney disease less deadly and therefore reduce the danger of this life-or-death decision. Given that paying a donor for a kidney is illegal almost everywhere in the world—with the single exception of the Islamic Republic of Iran—there are black markets. Because it’s illegal to be paid for a kidney, these black markets often operate outside of the traditional medical establishment, with surgeries being done outside of hospitals. That means very low quality medicine and very serious dangers both to donors and recipients.

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

One of the ways to fight black markets is with legal markets. One of the examples I use in Moral Economics is prohibition. When we had prohibition of selling alcohol in the United States, there were lots of black markets and organized crime grew up to mediate them. When we ended prohibition in the early 1930s, alcohol remained problematic—there’s still alcoholism and drunk driving—but you can no longer buy moonshine whiskey from gangsters. We eliminated the organized crime component that prohibition had produced.

If we want to fight black markets and make kidneys more available, those two goals go together. Many opponents mistakenly, I think, believe otherwise. They look at the black markets and say these markets are very bad because donors are being paid for their kidneys. But really they are very bad because they deliver very low quality medicine in very inequitable ways. If we could design an acceptable market in which we could pay donors for kidneys, kidney disease would be much less a cause of death and criminality.

Mounk: What about another objection that people might make, which comes downstream from a book that I think is fun, interesting, and insightful, but that has perhaps overly influenced people’s understanding of economics—Freakonomics? A Freakonomics-style argument that you could make about this, and I’ve heard moral philosophers make similar arguments, is that sometimes when you pay people for something, they stop giving it for free. If there are 500 people who think of themselves as moral actors and want to give a kidney to save somebody’s life out of altruism, giving them minor tax incentives—or perhaps even paying them a significant amount—might cause some of them to say: I’m not the kind of person who sells a part of my own body for money. That seems like a very different kind of moral act than donating a part of my body so that somebody else may live. How sure can we be that compensation would actually help alleviate this shortage?

Roth: That’s an empirical question. There is vast evidence from the world’s economies over centuries that while paying for things might stop people giving them away for free, that effect is more than compensated for by the fact that if you pay enough, people are willing to be paid providers. There’s a famous passage in Adam Smith that says it’s not through the generosity of the butcher and the baker that you get your food—that’s how they earn their living. They sell food. By and large, we are not short of food.

Blood plasma is a good example. I was an early adopter of COVID, and when I recovered I went to Stanford Blood Center, where you can donate blood plasma without being paid, because there was a demand for convalescent plasma. Once my antibody levels fell below a certain threshold, they let me go. I still give whole blood, because we do that for free in the United States and that’s how we get whole blood. But I no longer donate plasma, because there’s plenty of plasma in the United States—we pay plasma donors, and there’s no shortage. In fact, the United States exports about 70% of the world’s plasma, because it’s in short supply in the many countries that consider it immoral to pay plasma donors. Those countries can always buy plasma from the United States. So yes, I’ve been crowded out of the plasma market. It’s awkward and not particularly enjoyable to donate plasma, and I don’t need the $70. I also don’t want to take that option away from people who are supplementing their income by donating plasma, which is plentiful here.

Mounk: There is another kind of argument that a number of moral philosophers have made: that there is a way we should value our bodies and our organs, and that making them the kinds of things that can be bought and sold is to dishonor what our bodies are. It is simply valuing them in the wrong kind of way. My doctoral advisor made this argument, among others. How would you respond to that?

Roth: Millions of people are dying of kidney disease. I would want to at least weigh that idea against the cost of all these deaths. You would have to be very sure that this aesthetic judgment was more important than millions of deaths a year around the world. In most places, kidney disease is among the top ten causes of death, and it’s growing—as we beat back infectious disease as a cause of death, kidney disease becomes relatively more prominent. You have to be very sure that you find the idea of people selling their kidneys objectionable before you condemn so many people to death.

Mounk: To show my hand: I want to make sure I run through the devil’s advocate arguments in a fair way, and I think those are some of the main ones people make against these kinds of transactions. I agree that there is something queasy about the buying and selling of organs, and I absolutely understand why a lot of people—probably many listeners—have a deeper objection to it. But one of the fundamental things in politics is that you always have to think in trade-offs and understand what’s on the other side. In cases like this, the trade-off—not in some theoretical or distant way, but in a very immediate, everyday sense—is people needlessly dying. You need to be really sure that the moral arguments you make against a practice that could alleviate untold suffering are sufficiently weighty to justify persisting in that position.

To make a meta-ethical point, which I try to refrain from on this podcast: I think this is actually an instructive way of thinking about utilitarianism. In the broader debate in moral philosophy between consequentialist outlooks—which evaluate actions and rules by their consequences, of which utilitarianism is the most prominent, measuring policy in terms of whether it increases the balance of happiness over pain—and more deontological notions, which hold that certain things are in themselves wrong and that rights obtain irrespective of consequences, I am not a straightforward consequentialist. I don’t think it is obviously true that in every realm of our lives we should simply look at what increases the balance of happiness over pain. But I do think that when the consequences of a rule or an action are very obviously very bad—as they obviously are in the case of prohibiting the buying and selling of organs, leading to tens of thousands of people needlessly dying in the United States and many more around the world—you had better have really weighty arguments on the other side. Consequentialism is not necessarily the right philosophy in every context, but it makes a sensible default in policy situations when the consequences of a rule are as disastrous as I believe the consequences of not allowing the buying and selling of kidneys to be.

Roth: I agree. The market design question also requires us to be very careful to make sure that bad things don’t happen to paid donors. That is a matter of high importance in the design of any eventual market that might relieve the shortage of kidneys.

Mounk: That leads to my next question. Let’s imagine that Congress finally changes its mind and decides to allow the buying and selling of kidneys—and asks you to come in and help design this market. How would you go about doing that?

Roth: Any market you design would have to be constantly monitored, and if bad things were happening you would want to change the rules to try to avoid them. A good way to start would be to say: we’re not going to have a laissez-faire market for kidneys. Rather, we’re simply going to amend the law so that the federal government can pay for kidneys. The kidneys obtained in this way would be regarded as the same kind of national resource that we now regard deceased donor kidneys, and would be allocated according to need. Medicare would pay for the costs so that the rich and the poor alike would be able to get the suddenly more abundant kidneys.

Some of the payment to donors might involve coming back for annual checkups for the next ten years, because we would want to be really sure that donors are doing well, and if there’s something that can be fixed about the way we’re dealing with them, we should fix it. One of the problems now is that it’s a little bit hard to keep track of donors because they’re healthy—they’re off skiing somewhere. The goal would be to incentivize donors not just as an expression of gratitude for providing a kidney, but to make sure that they continue to allow their health to be monitored, so that donating a kidney becomes a very well-understood and desirable thing to do.

Mounk: Kidneys are a fascinating case that has allowed us to make concrete a lot of these broader questions. Let me ask a different one. What should we do about kidneys today, while they remain scarce? And setting kidneys aside, even if we solved that problem, there will still likely be a scarcity of livers, hearts, and other organs that might be transplanted. How should we think about the best algorithms and decision-making procedures for determining which patients get those organs and in which order? Obviously, we likely don’t want the richest person to get priority, or for allocation to be based on discriminatory criteria such as race. But how should we think, beyond those baseline assumptions, about designing systems to distribute these very scarce goods?

Roth: That gets squarely into the area of technical difficulties in allocating deceased donor organs, because you can’t do living donation of hearts. Livers are different—you can donate a lobe of a liver—but hearts will always involve deceased donation.

With deceased donation, you first have to decide when the donor is dead. That’s a little controversial because we have two criteria that in ancient times were the same: brain death and circulatory death. The reason those used to coincide is that once your heart stops beating, your brain very quickly dies. The special thing about transplantation is that death isn’t exactly binary. A donor can be dead while some of their organs are still viable enough to be transplanted—but you have to work very quickly, and an organ doesn’t stay viable for long outside the body. So organs have to be allocated quite quickly.

Subscribe now

There are many more people who need a heart transplant than there are hearts, and similarly for other organs. Some of the questions involve efficiency and how it relates to equity. Consider a waiting list for hearts that depends either on how long you’ve been waiting or on how sick you are—those are different ways to prioritize patients. In the United States, the idea is that a deceased donor organ should be offered first to the highest priority person on the list, then to the next highest, and so forth, with priorities depending on many things including location. But the problem is that it takes a little while for an offer to be considered and accepted or rejected. If it’s rejected, the heart has a little less chance of reaching someone who will accept it while it’s still transplantable.

Sometimes you have to reckon with the fact that a particular organ is hard to place. It’s a marginal organ—it will save somebody’s life, but many high-priority patients may choose to wait for a better one. Can you skip over people with higher priority and offer it to someone you think will accept it now? Some organs are offered too many times, rejected repeatedly, and then are no longer usable. Among the things we have to think about is how to balance efficiency and equity: how strictly must you follow a predetermined priority list, or can you expedite an organ through a different process? That is a matter of current debate.

Mounk: There is obviously a black market for organs, and it is significant. But transporting an organ requires a lot of attention—many people involved, a great deal of infrastructure. The prohibition on buying and selling organs has not been perfectly enforced, but it has had a very real effect in suppressing those transactions.

There are other areas where the state’s ability to suppress transactions is much more in doubt. Drugs and other illegal substances are probably the prime example. There must be many, many transactions every year involving the buying and selling of methamphetamine, cocaine, or heroin in the United States—many more than there are of people buying and selling organs. The harms from drug prohibition are straightforward and obvious: people going to jail, enormous resources spent on enforcement, users being criminalized and potentially unable to access help. But the dangers of completely opening up those markets—of potentially having very sophisticated corporations advertise and sell drugs in ways that could have hugely damaging consequences for individuals—are also very real.

From my philosophical perspective as a philosophical liberal, this case pushes harder against my usual framework than organs do. With organs, you can invoke the harm principle and argue that nobody is clearly harmed by allowing buying and selling, as long as certain side conditions are met—that people can’t be coerced, and that no one is so destitute that selling an organ is their only alternative to starvation. In the case of drugs, a strict harm principle reading might say: if an adult chooses to use heroin, the person being harmed is themselves—why should we intervene? But drug use can lead to significant rises in crime in neighborhoods and imposes real costs on society at large.

Yesterday, walking back from dinner with conference participants here in Boston, we came past a gentleman who was crying for help and appeared to be in the grip of addiction. We stayed and called 911. When the EMTs arrived, they recognized him immediately by his first name—he was clearly a frequent user of that service. I’m very glad he got help, and he seemed to be a decent person in a very desperate situation. But it is also obvious that that imposes costs on the broader community. You can try to argue that even on the harm principle, drug prohibition is justified. But it’s not clear to me how clean that argument is, because the primary harm is to the individual. This is a topic on which I am genuinely torn. How do we puzzle through it with more clarity?

Roth: One thing I say in the conclusion to my book is that when we think about the things we are morally obligated to do, they have to be things that we can actually do. Even if we feel morally obligated to prevent heroin use, we have to recognize that there is still a great deal of heroin, fentanyl, and crystal meth in circulation. Our prisons are full of drug offenders, and yet the drugs remain. One of the thought experiments I discuss in the book is why it is so easy to buy drugs and so hard to hire a hitman. I think it has to do with the social attitude towards drugs on the one hand, and towards killing for hire on the other.

Mounk: A more obvious explanation may be this: every time someone uses a drug, they may be breaking the law, but there is no immediate external victim—which is the point I was making about the harm principle—and the state has no immediate way of knowing it has happened. Every time a hitman kills somebody, they are harming another person, and it is likely that someone will notice the victim is missing and the state will begin an investigation. Part of why drug crime is so prevalent is that it is quite easy to get away with, at least at a low level or as a user. You can violate a drug law and the state has little way of knowing. It is quite difficult to violate the prohibition on murder without the perpetrator eventually going to jail.

Roth: That is largely true. But there have been years in which there were more than 100,000 opioid overdose deaths, which vastly exceeds the number of homicides from any cause in the United States—I believe fewer than 20,000. Drugs kill far more people than murder does. And yet murder for hire is so rare in the United States that it doesn’t even appear in the national crime statistics.

Mounk: If you think that you’ve successfully hired a hitman, 99 out of a 100 times, it’s an FBI agent who’s about to arrest you.

Roth: If someone told me they wanted to hire a hitman, I might call the police, and they would say, why don’t you go back to Professor Roth and tell him there’s a bar where he can find one—and that’s where I’d end up trying to hire an FBI agent. Whereas if someone called the police and said Professor Roth is thinking about trying heroin and wants advice on where to get it, the response would be: why are you calling us? We’re a busy police department.

Part of the conclusion I draw in the book is that not only do markets need social support to work well, but so do bans on markets. The ban on drugs doesn’t have enough social support to work well—we have a great deal of drugs even though they’re against the law. We are losing the war on drugs. But it also turns out it’s not so easy to accept our surrender. In places like Portland, where they’ve tried decriminalizing use, the result has been open-air drug use that is also very distressing and makes cities difficult to live in. That is a real cost to citizens, and Portland has walked that back.

Drugs are the market about which I have the least ability to make helpful suggestions, except that we should continue to experiment. In many places we allow clean needle exchange, and at least we’re not seeing as much HIV transmission alongside drug use as we used to. That seems to help a little. We may find that incarceration remains among the options available for dealing with addiction, but it will certainly have to be combined with more medical approaches to help people deal with it. We don’t know the answer yet, but that calls for more experimentation and more search for solutions, not less.

Mounk: What does success actually look like in this case? Laws against jaywalking, which I think are largely silly, presumably succeed if fewer pedestrians die in car accidents—that’s the actual ultimate goal. You wouldn’t say the goal of the prohibition on jaywalking is that nobody ever jaywalks. Perhaps the prohibition actually does lead to fewer pedestrian deaths, in which case it’s a sensible law. I suspect that it doesn’t, though I haven’t seen empirical studies on it, and if it doesn’t, I think we should abolish it.

When you think about the success of a prohibition on theft, part of that is allowing a thriving economy with economic exchange and sparing the costs of extremely elaborate safety procedures. Part of it is reducing the number of thefts, because it is upsetting to have something stolen from you. We can consider the law against theft in the United States broadly successful even though things are obviously stolen every day. We don’t need perfect compliance for a law to be successful.

When we think about the prohibition on heroin or methamphetamine, what is the criterion of success, and how are the laws doing relative to that criterion?

Roth: Jaywalking isn’t a bad example, because I believe that in New York City, jaywalking has been decriminalized. One reason it was decriminalized is that it gave too much discretion on enforcement and was enforced inequitably against people of lower socioeconomic status. Part of having laws is asking whether we support the way they are enforced.

To come back to drugs: part of the reason it’s hard to move forward is that we have deep disagreement about what our goals are. There are people who think it is simply wrong to allow sales of drugs like heroin or fentanyl for private use—though fentanyl, of course, is an essential drug in medicine. I’ve been given fentanyl during surgeries. It’s a quick-acting anesthetic.

Mounk: I was astonished when I had a wisdom tooth out—which turned out to involve relatively minor pain, perhaps I was lucky—and my dentist prescribed me an opioid without telling me it was one. When she handed me the prescription, I Googled it and asked whether it was an opioid. She said yes. I asked if I needed it, and she said I could just try some Advil instead. So I never filled the opioid prescription, partly out of caution. If I had been in terrible pain, I would have. But the ease with which opioids were being prescribed in those circumstances—and this was only a couple of years ago—is astonishing.

Roth: Consider alcohol. We ended prohibition because it turned the United States into a nation of lawbreakers—whenever you had a drink, you were aiding and abetting whoever sold it to you. Prohibition didn’t have a big effect on alcohol consumption, but it did have a big effect on creating organized crime. Ending prohibition didn’t solve the problem of alcohol, though. There are still alcoholics, organizations like Alcoholics Anonymous that try to help people with addiction, and people who drive drunk and kill others. The problems that caused people to support prohibition didn’t go away—prohibition just didn’t solve them.

That’s essentially where we are with drugs. The problem of addictive, lethal drugs is very large. It’s not even entirely clear, from the libertarian point of view, that people are choosing to take these addictive drugs freely—you may choose to take one initially and then find yourself enslaved by it. So there’s also a distinction worth drawing between what I call repugnance and what’s called paternalism. Sometimes we’re trying to protect people from mistakes that will harm themselves. We have many laws that do that, like prescription drug laws. Before you got that opioid prescription, it would have been more difficult to obtain opioids without one, but your dentist judged that you might need them, and on that advice you were legally allowed to buy them. We require prescriptions for a vast array of drugs because we believe you need expert advice before consuming them safely. Going forward on drug use, we may move in the direction of treating addicts more like patients than like criminals—but it’s not simple, because addiction is complicated and enslaving.

Nicotine is a useful drug to think about in this context, because tobacco causes a lot more deaths than opioids do, and even more than legal alcohol does—though some of those deaths are delayed. Cirrhosis of the liver is not quite as dramatic as someone collapsing on the sidewalk. Tobacco is legal, but heavily and increasingly regulated. Americans smoke much less than they used to. There were large advertising campaigns—you mentioned not wanting heroin advertised by some corporate equivalent, but Big Tobacco ran extensive ads. The Virginia Slims campaign told women “you’ve come a long way, baby,” and tried to convince more women to smoke. It was, I think, largely successful at the time, at least among those women who didn’t mind being referred to as “baby.”

Mounk: We tend to think in terms of fully open markets or fully prohibited markets, but there are many regulatory arrangements in between. One strange irony is that in many contexts now, smoking a cigarette is very poorly viewed—at the very least a breach of a taboo, and often against the rules. Meanwhile, in many places, smoking a joint has become far more morally acceptable. In Brooklyn, you probably see more people smoking joints on the street than smoking cigarettes, which is rather astonishing. There is very little open advertising of cigarettes because of extensive regulation, but in my neighborhood there are three or four shops very aggressively advertising marijuana.

Marijuana is an important case. It is the drug where the war on drugs clearly went too far and was most damaging, because it is less immediately dangerous and deadly than many other drugs, and yet many people went to prison for selling it—and sometimes for consuming it, though that was rarer. The case for overturning prohibition on marijuana was very strong.

At the same time, we are seeing some of the damaging mechanisms you might have feared. Marijuana is much stronger now than it was even twenty years ago—the nature of the drug has actually changed. There are some suggestive findings that heavy use is very bad for users and may be bad for those around them. One recent study suggests that if you are a significant user of marijuana in the months before a pregnancy—even as a man—the likelihood that your offspring has autism or another serious developmental challenge is significantly increased. The mechanism by which organized commercial interests are now spreading this product, marketing it, perfecting it, and making it more powerful—in ways that also make it more addictive and more dangerous—is fully in motion. We are in the middle of an experiment whose results aren’t yet clear. How do you feel about how marijuana legalization has gone? What are the positive elements, and where are you concerned?

Roth: Marijuana is becoming more like tobacco—legal, with companies perhaps perfecting its addictive qualities, as tobacco companies apparently did quite consciously. They understood the addictive properties of nicotine. But the laws against consuming marijuana were not effective. If the point of those laws was to make marijuana unavailable, they failed. They also had the consequence of sending people to jail, forcing them to deal with criminals, and turning them into criminals themselves—producing widespread law breaking. Recent surveys suggest that the number of Americans who use marijuana every day now exceeds the number who have a drink of an alcoholic beverage every day. What you’re observing in New York is easy to verify with your nose when you’re near a cannabis user.

Cannabis is now much more potent. There was an organization called the California Growers Association—and California has many organizations with similar names, though normally they have a vegetable in the title, like the California Artichoke Growers Association. The California Growers Association was ambivalent about legalizing marijuana. People who grew marijuana illegally talked about how many plants they had—hundreds or thousands of plants. But farmers talk about yield per acre. The artisanal growers who were hiding their crops in national forests to avoid raids by the Drug Enforcement Agency have been largely put out of business.

We haven’t reached equilibrium yet. It’s not a crime to smoke marijuana or consume edibles, but we don’t yet know how cannabis will compare to alcohol, tobacco, sugar, or other things we consume. It remains to be seen. We shouldn’t regret that marijuana is now more available, because it was also available when it was a crime. The question is how to manage it—and leaving the management of these difficult questions in the hands of criminals is not the best way to do so.

In the rest of this conversation, Yascha and Al discuss the ethics of surrogacy, if sex work should be legal, and whether the rise of OnlyFans is concerning or empowering. This part of the conversation is reserved for paying subscribers…

Read more

Timothy Garton Ash is the author of Homelands: A Personal History of Europe and writes the newsletter History of the Present. His upcoming book, Europe in 7½ Chapters, will be published in October 2026.

In this week’s conversation, Yascha Mounk and Timothy Garton Ash discuss the crisis of Labour and rise of Reform, why Europeans are struggling to adapt to a new political, cultural, and technological age, and the future of the war in Ukraine.

Will you be in London on Sunday, September 6? I’ll be interviewing Francis Fukuyama about his life and thought to mark the publication of his memoir In the Realm of the Last Man at the Sekforde at 5pm. Find out more and book tickets here. Paying subscribers can access a code for free tickets here. —Yascha

This transcript has been condensed and lightly edited for clarity.

Yascha Mounk: I thought we would do a little European tour with you. I didn’t tell you that we had hired you as a tour guide, but you are our historically and politically informed tour guide. We are going to do a little trip from the northwest of the continent to the southeast in rough geographical order. We will start with local elections in Britain. That does not sound particularly exciting.

The results are coming in as we speak, and they seem to herald a very significant political shift in British politics. What is your read on what is happening?

Timothy Garton Ash: The motto for my intellectual tour company is, of course, “pessimism of the intellect, optimism of the will.” That goes on the brochure. These elections, which are both local but also Scottish and Welsh national elections, are absolutely fascinating. What they show you is that this country, which voted to leave Europe just about 10 years ago on the 23rd of June 2016, is now becoming ever more European in its politics.

Number one, by far the biggest winner in England is Reform UK, which is a classic hard-right populist nationalist party of the kind we basically did not have in British politics for decades, if not centuries. It is very like Fratelli d’Italia, or Rassemblement National in France, or Alternative for Deutschland. I sometimes like to call it Fratelli d’Ingleterra, the Brothers of England. They are sweeping the board, taking votes both from Labour and the Conservatives.

Finding number two, it is also becoming very European: tremendous fragmentation. This used to be the country of the two-party system: Labour and Conservatives, His or Her Majesty’s Government, and His or Her Majesty’s Opposition. Now we have a five-party system in England; and if you take in the nationalists in Scotland and Wales, both of whom are doing very well, you have a seven-party system.

Thirdly, as in Catalonia and the Basque Country, the discontents that flow into a populist vote also flow into votes for separatists, nationalists, or regional parties. The Scottish National Party does spectacularly well in Scotland and, for the first time ever, Plaid Cymru seems to be doing pretty well in Wales. Welcome to Britain, a very typically European country.

Mounk: We are in this strange situation where the Conservatives have been discredited by a long and very chaotic stretch in government. A whole bunch of different prime ministers struggled to deliver on Brexit in the way they promised. They now have a rather impressive new leader in the form of Kemi Badenoch, who has been improving in the polls a little bit in the last months. However, that last decade-plus of chaotic government is really hanging around her neck like a millstone.

We have a Labour Party, the other big traditional party in that two-party system you invoked, that was elected on a huge parliamentary landslide in the general elections about two years ago, but not with a huge share of the vote. Already at that point, part of the reason for its landslide victory was the fragmentation of the political system, but they managed to concentrate enough of that vote on themselves to have this huge parliamentary majority.

They came in with a leader, Keir Starmer, who is a little bit of a chameleon. He was a loyal adjutant to Jeremy Corbyn when the Labour Party was extremely far left. He managed to win the leadership of the Labour Party by being somewhat acceptable to different wings of the party. Then he became very moderate as the leader of the Labour Party and clearly expelled the Corbyn wing from the party, but he never really seemed to have a positive program. The British public has soured on him very quickly once he got into office.

So perhaps the first question is, why is it that Labour has fallen from grace quite so quickly? And why is it that the Conservatives have not been able to pick up the slack of an unpopular incumbent political party, as might be expected in a top two-party system that still has a hold over the country?

Garton Ash: Remember that Labour’s “landslide” victory was, to a significant degree, because the right was split between the Conservatives and Reform UK. Actually, the Labour share of the vote was slightly down from the previous election, so it was not a landslide in that sense.

Secondly, it turns out that Keir Starmer is not a very good politician. He has made a series of often foolish mistakes and an endless series of u-turns. These involve apparently trivial matters, like appointing Peter Mandelson as UK ambassador to Washington. Then, more material from the Epstein files is released, and it turns out that Mandelson was deeply compromised and even sharing inside information from the Cabinet table. That in itself seems trivial, but there has been a whole series of those incidents.

Above all, what are they doing with power? What are they doing with this enormous parliamentary majority? It is totally unclear to everyone where they are going. That is partly because it is difficult to work out where a post-Brexit Britain goes when it is being abandoned by its “best friend,” the United States, and has very little money left in the kitty. With soaring public debt, deficits, and an overburdened welfare state, it is difficult to work out what you would do in those circumstances; but whatever it is you might do, they are not doing it.

Mounk: What about the Conservatives? Why is it that they have not been able to pick up the slack? Is it just the fact that they have been in government for ten-plus years and people are sick of them, and not enough time has passed for them to be able to represent themselves? Is it that Kemi Badenoch is not effective as a leader, or that she has not figured out a way forward for a center-right party?

Is it part of a broader trend? We used to talk about the slow death of social democracy in the 2000s and early 2010s, but it turned out that the slow death of social democracy was just a precursor to the slow death of the old catch-all parties. Christian Democrats and Conservatives across Europe and other countries are now declining in the way that social democrats were, and the Tories are just one instance of that.

Garton Ash: I incline toward the second explanation because that is what we are seeing across Europe. One could even go more broadly than Europe. What was peculiar, not just to Britain but to Europe post-1945, was liberal conservatism, called Christian Democracy on the continent. Now, as we see everywhere, the barrier between that and hard-right nationalist populism is breaking down. If anything, the voters are going off to the hard-right populists, so I would favor that structural explanation.

On top of that comes 14 years of Tory misrule. People have not forgotten that from 2010 to 2024, we had a Conservative government which implemented stark austerity. For many people, even Conservative remain voters, they took the country out of the EU, and everyone can now see that was a terrible mistake.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

Then there is a slightly sensitive issue, which is that on the right, and particularly for populists, immigration is the key issue. Immigration is not just immigration; it is about deeper cultural changes in the country. People feel they do not recognize their country anymore. If you have voters like that who are, say, elderly, white, and middle class, you have a leader who is extremely impressive in her own way, Kemi Badenoch, but who actually grew up in Nigeria and did not know she was a British citizen until the age of 14. She is up against Nigel Farage, the bluff, pint-of-beer-quaffing man from the 19th hole in the golf club. That is an uncomfortable subject to point to, but I do think that for an elderly electorate—and the Tories have an elderly electorate—that is a significant part of the explanation.

Mounk: If I am remembering correctly, Kemi Badenoch was born in the United Kingdom, then grew up mostly in Nigeria until she was a teenager and came to Britain. Unlike Kamala Harris, who claimed to have worked in McDonald’s, but I think there was never any very strong evidence that she had, she did indeed work in McDonald’s and work her way up in a very impressive way. I met her once when I was giving a presentation in Parliament when she was a backbench MP and I had never heard of her. She came and asked a number of somewhat aggressive but very perceptive questions. I remember being very impressed with her at the time and have been following her rise with interest since. There are many things I disagree with her on, but from that first meeting, I thought that she was a very impressive person.

I am a little skeptical about the role that race plays in this for the following reason. I had a debate with a good friend who knows British politics well when there was the leadership election for the Conservative Party. The way it works is that the Members of Parliament narrow the field down to two candidates, and then there is a choice among the membership of the Conservative Party. The membership of the Conservative Party skews old, very conservative, and somewhat away from London. I thought that Kemi Badenoch would win that election quite clearly. My friend was making the same argument that you just made, which is to say, those old Tory Party members are very conservative and quite wedded to an old vision of England. Are they really going to pick somebody like Kemi Badenoch over Robert Jenrick, a white guy who went to university at the same time as me in Cambridge?

Kemi Badenoch won that election very clearly. In that case, it seems we have quite strong evidence that this was not so strong an obstacle to her. It seems to me more broadly that when you look at even right-wing populist movements, people are very open to voting for ethnic minority candidates that they feel represent their views and that actually reassure them that immigrants to the country are able to stand up for the values of their homeland as they see them. Kemi Badenoch, of course, is not averse to a little bit of culture warring. She does very loudly represent what people on the right of politics would think of as traditional English values.

To me, it is not obvious. Surely there are going to be some people who are not going to vote for her because she is Black, because they are out-and-out racist. There are also going to be a lot of people who actually find that very appealing, who say that she is able to make a very full-throated, persuasive case for those values precisely because she is not the stereotype of the person who might make those claims. Why do you think that she was able to win the membership vote in the Conservative Party so strongly? Is that a very different electorate? What is the difference?

Garton Ash: First of all, I really do not want to put too much weight on this particular factor. I think the other two factors are significantly more important. Secondly, I was very impressed when they went for Kemi Badenoch. A party which had the first Jewish prime minister in British history, Benjamin Disraeli, the first female prime minister in Margaret Thatcher, and now they go for a woman of color. Very impressive.

The fact is that an awful lot of Conservative voters have gone to Reform UK. The man you mentioned, Robert Jenrick, who was effectively the runner-up and a leading figure in the Conservative Party, is now a leading figure in Reform UK, along with several other former Conservative ministers.

Let us put to one side the question of race, which in Britain means something slightly different than in the United States and for Kemi Badenoch. The fact is that at the drop of a hat, not just voters but also ministers and senior politicians from the Conservative Party are going off to Reform UK.

Mounk: As a side note, Robert Jenrick was at university with me studying history as well and nobody can remember him, which is a very strange thing.

Garton Ash: It is said—although I am not sure if this is a reliable authority, but according to the columnist Matthew Parris—that he is the only person to have lost an election in which he was the only candidate.

Mounk: I can disconfirm that because the same thing happened when I was an undergraduate in my college. There was a women’s officer for the Students Union for which the electorate was exclusively women. The candidate who ran had electoral posters which were slightly ill-conceived in general, and perhaps particularly ill-conceived given that the electorate was exclusively female, which read: “Vote for me because I’m gorgeous.”

She lost to the option RON, which is short for Reopen Nominations. There is at least one other election I can remember in which that happened.

Garton Ash : Robert Generic is his nickname, and the point is that he is so transparently opportunistic and chameleon-like that it helps explain why he lost to Kemi Badenoch. That is by the by. I think these are secondary issues.

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

Now, just one further word on England specifically: let us remember that the Conservative Party is the most successful party in modern political history bar none. I would still not want to write it off. When the memory of 14 years of Tory misrule is a bit more remote and Reform UK starts getting some scandals, which it is bound to do, it may be a different story by the time of the general election, which has to happen before 2029.

Mounk: Speaking of Reform UK, tell us a little bit more about Nigel Farage and Reform UK. Earlier, you compared the party to Fratelli d’Italia, to the Rassemblement National, and to the Alternative for Deutschland. I increasingly think that we may need to make finer distinctions between different right-populist movements. I certainly think that Reform UK is a right-populist movement.

I see the resemblance in certain respects to Fratelli d’Italia, though Italy and Britain are very different countries. I also see the resemblance to someone like Marine Le Pen or even Jordan Bardella, who are clearly in the right-populist camp but have also distanced themselves from the post-fascist roots of what used to be the Front National and the figure of Marine’s father, Jean-Marie.

It does also seem to me as though there is a different wing of European right-wing populism which retains a deeper flirtation with the past and which is more deeply uncomfortable with any form of ethnic and religious diversity. I want to see to what extent you stand by the idea that Reform and the AfD in Germany are comparable, or to what extent it is helpful to think of those as really part of the same families of political parties.

Garton Ash: I think it is. It is certainly more like those classic continental populist hard-right parties than anything we have seen in mainstream British politics for a very long time. Such parties have been absolutely marginal in British politics. In that sense, it is more like them. Its major themes, like immigration, are very much the same, as is the mix of cultural conservatism and nationalism with more left-wing social and economic policy ideas, such as generous provision for welfare state pensions.

In its basic morphology, you are right that it is a more moderate or civilized version of some on that spectrum. It is very much more to the Meloni end of the spectrum than the Orbán or the East German AfD. That is undoubtedly the case.

An important point is that the effect of all this is not necessarily that we see a Nigel Farage Prime Minister, but that either we have a coalition of Conservatives and Reform UK—which is exactly what we are seeing all over Europe, the temptation for the center-right to go with the hard-right because there is no alternative—or we get a reform of our electoral system. If the only party that is doing well out of the electoral system seems to be Reform UK, and the former big parties—Labour and Conservatives, along with Lib Dems, Greens, and others—all go down to under 20%, then suddenly there might actually be a majority for the reform of the electoral system.

Mounk: That would be remarkable. Historically, in the United Kingdom, we had a two-party system between Liberals and Conservatives. The Liberals were supplanted by Labour. When Labour was weakened and far to the left in the late 1970s and 1980s, there was a rejuvenation of the liberal movement through the Liberal Democrats. The Labour Party and the Conservatives, as the two major political parties, were traditionally against electoral reform because “first-past-the-post” was most likely to give them periods of consolidated rule, as it continued to do until recently. It was the lone voice of the Liberal Democrats, along with scattered smaller parties, that wanted electoral reform.

It would be quite a turnabout for Labour, and possibly the Conservatives, to now vote for electoral reform. In the British system, which has very few checks and balances, Labour could just decide to do it; they have a parliamentary majority to put it in place. But it would be an extraordinary anticipatory capitulation, would it not? It would be a recognition that we have no chance of being re-elected or even of having a period in opposition for five or ten years before roaring back to be the governing party. We are giving up on the historic place that we had in the electoral system. Whether you can get the strategic initiative within the party and the willingness to bear that public humiliation is a very interesting question.

I want to cover the other part of the electoral transformation. We have talked about the weakness of Labour, the struggles of the Conservatives, and the rise of Reform. We are also seeing in many opinion polls, including those for the next general election, Labour running more or less head-to-head with the Green Party, which is led by Zack Polanski. In these local elections, another political force has gained significantly: Muslim independent candidates. These are basically sectarian candidates in largely Muslim parts of the United Kingdom running, in part, on issues regarding the Middle East and Israel, but also on deeply conservative social policies.

Subscribe now

You have two trends. The first is a form of sectarianism of which we have started to see inklings in different European countries. In the Netherlands, which has a system of proportional representation and a low electoral threshold, you had the rise of the Denk Party, which is a kind of Erdoğan-esque, mostly Turkish-Dutch party. Secondly, the Greens look very similar to the La France Insoumise of Jean-Luc Mélenchon in France. What the French sometimes call islamo-gauchisme is a strange mixture between a bourgeois, highly educated, urban environmentalist left and a deeply conservative, if not reactionary, Muslim identitarian movement under the flag of one political party.

How do you assess the threat that the Green Party poses to the Labour Party, its prospects of establishing itself as one of the major forces on the left, and what do you think about these Muslim independent candidates?

Garton Ash: I do not think the comparison with La France Insoumise really holds up. I would say that Zack Polanski’s Greens are what the French call bobo, or bourgeois-bohemian. His appeal is that he is radical—significantly to the left of Labour on many issues—and, of course, very strong on Gaza.

The classic green issues are also a factor. I am talking to you from Oxford, where the Greens have just had a tremendous success in the local elections, and those are not the kinds of voters you are talking about.

What I would say is that Gaza, and then Trump’s war—or the U.S.-Israeli war against Iran—are not only making us in Europe, including Britain, seem utterly hypocritical through the question of double standards—Ukraine versus Gaza and the Iran war, respect for international law, and so on—but are really damaging what have been relatively good inter-community relations.

In a country where the figure for London is well over 40% foreign-born and nationwide is over 20% foreign-born, with large Muslim communities but also significant Jewish communities, we have had horrifying, repeated antisemitic incidents and stories of antisemitic violence. The Middle Eastern element is a very important part of the story; that is what is helping to tease our politics apart. But that said, Yascha, I think the situation here is nowhere near as bad in that respect as it is in France.

Mounk: You know both countries well, and I have spent a significant amount of time in both as well. I am more pessimistic than you are. When I look at what Zack Polanski has said as leader of the Green Party and how he has positioned himself—he is himself Jewish, but there are members of La France Insoumise who are Jewish as well.

After the terrible terrorist attack in Golders Green recently, in which someone stabbed two visibly Jewish men in a very Jewish neighborhood of London, deliberately targeting Jews, the response of the leader of the Green Party was remarkable, even after other attacks on British Jews in the preceding weeks and months. He had doubted whether a feeling of unsafety among British Jews was rooted in reality or merely perception. His first response after this terrorist attack was to retweet somebody criticizing the police response for subduing the attacker too harshly. Again, we’re talking here about somebody who is in the process of stabbing people, grievously injuring them in a heavily Jewish neighborhood in a terrorist attack and Polanski’s first response was to criticize the police for doing too much to stop this attack.

This is, to me, a level of denial of a very clear and present threat to Jewish life in Britain which is every bit as bad as what Jean-Luc Mélenchon has done in France.

Garton Ash: It is indefensible what he said; I could not agree more. For me, it is utterly shocking that ordinary British Jewish families are talking about needing to emigrate because they do not feel safe in this country.

Where I want to push back a little bit is that La France Insoumise is a major electoral and political force in France. Indeed, there are scenarios in which the second-round runoff might be between someone from La France Insoumise and the Rassemblement National. The Greens are nowhere near that, so I think their larger political significance is much less. However, we are absolutely in danger of attracting the kind of politics in this country from which we have been relatively free.

Mounk: In Oxford, the appeal of the Green Party is to students and left-leaning academics. It is a bobo party—the party of the bourgeoisie bohémien, a term David Brooks invented. I have an episode with him on the podcast about that. In other parts of Britain, the appeal of the Green Party is very different. You see a political coalition in the mix of candidates for local elections: socially progressive students in Oxford and socially conservative Muslim candidates who care about the Middle East. The Green Party is trying to keep this together, but it is not clear how long it can last.

Perhaps that is the natural transition to France. In polls for the next general election—though there are questions about sustainability in a first-past-the-post system—the Greens in Britain currently get between 15% and 20% of the vote. That puts them on par with Labour in some polls. In France, Jean-Luc Mélenchon is around 10% in preferences for the first round of the presidential election. It is not obvious that the Greens are that much weaker than La France Insoumise is in France.

Regarding France, we have important presidential elections coming up in the spring of 2027. Emmanuel Macron cannot run again. Interestingly, there is no lifetime limit on presidential ambitions in France, so Macron can run again in the future, but he cannot run this year. It is not clear there is an obvious centrist candidate to stand in for his movement. The likely candidate for the center is Édouard Philippe, or perhaps the young former prime minister.

The frontrunner is on the right: either Marine Le Pen, if she is allowed to run—which will be determined in a court appeal this summer—or her 30-year-old stand-in, Jordan Bardella. Bardella is an interesting figure who grew up in the suburbs of Paris. Then there is a chaotic left with candidates ranging from the “red-green” Jean-Luc Mélenchon to more moderate candidates like Raphaël Glucksmann in the center-left. How would you describe the political situation in France today?

Garton Ash: As it looks at the moment, the likely next president of France is either Marine Le Pen or Jordan Bardella. If Marine Le Pen is not allowed to stand, the chances may be even better for Jordan Bardella for the simple reason that his name is not Le Pen. In the last quarter century, the French have voted three times in the second round of a presidential election to keep out a candidate called Le Pen: twice to keep out Marine Le Pen, and once before that to keep out her father. There is something about that name which raises a certain allergy. Bardella is presenting himself as the very model of a modern populist.

As every presidential candidate must do in France, he has published a book. There was reporting of business leaders meeting discreetly with him for a nice dinner. I think that is what it is looking like, unless a candidate like Édouard Philippe or another unites everything from Macron’s liberal center to the center-left. It does not look likely at the moment.

The question becomes: how does a Europe with a President Bardella look? Is he more like Meloni or Orbán? If he were really Viktor Orbán, trying to put a spanner in the works at every step in Brussels and simply pursuing the French national interest coûte que coûte, that would be disastrous for the EU. This comes at a moment when it faces an unprecedented triple challenge: under attack militarily from Russia, politically from the United States, and economically from China.

If it seems more likely that it is a Meloni option, maybe that could work. We might actually work out some way of strengthening European defense rather rapidly, because we know we can no longer rely on Donald Trump, and Vladimir Putin might have a go at Europe in the next two to three years. We will then have a rather consolidated transition of the European right. The European right will be something different from what we thought it was 15 years ago, when it was center-right, Christian Democrat, and liberal conservative.

Mounk: That is very interesting. I absolutely agree with you about the electoral prospects of Jordan Bardella. Interestingly, nobody I talked to in France agrees with that. In the Macronist camp, they all think they can beat Marine Le Pen because they have done it before. For some of the reasons you mentioned, they are convinced that Bardella is too young and unproven, and that he would fall apart in an electoral campaign. They think it is all hype and TikTok, and that he would fall apart in the debate between the two remaining candidates between the first and the second round. I have heard from quite senior people that they are reassured about the prospects of Bardella running. To me, that gives me echoes of what a lot of Democrats were saying in 2016: “Give us Donald Trump; this is going to be great, and we are going to beat him easily.” It did not turn out to be that way.

Tell me more broadly about how you see the shape of France and perhaps Germany at this point. These remain—despite the enlargement of the European Union and the fact that this Franco-German couple is much less at the center of the EU in political and economic terms—the two most important countries in the European Union, even more so now that Britain has left the bloc. They both seem to be in a deep malaise in different ways.

Germany, it seems to me, had a postwar model that worked very well for decades. In certain respects, it has squandered that model; in others, it has failed to update it in a world where it no longer applies. France, in certain respects, has never quite had a model that worked, at least in a number of decades, and is struggling to find one now. It feels as though there is just this significant weakness at the heart of Europe in economic terms and in terms of a self-understanding of what role the countries of the continent can play in the world.

Perhaps we should raise our eyes a bit above the political battle and think about why it is that, from Germany to France to the United Kingdom, citizens have this deep feeling of economic stagnation, of fear of the future, of growing irrelevance, and of a social contract no longer really holding up.

Garton Ash: Germany is a different matter from France. France is a matter of slowly accumulating problems over a long period of time—an ancien régime which Emmanuel Macron attempted to change and failed to change. There is massive social spending, and the French Revolution is repeated because they want to raise the retirement age to 63 or 64, whereas in other countries people are working to 68 or 69.

Mounk: The striking fact according to the Financial Times is that the average pensioner in France has a higher income than the average working person.

Garton Ash: Yes, and by the way, it creates a beautiful, wonderful way of life which many Brexit-voting Brits love to enjoy in their retirement.

The German case, as you know very well, is one of an incredibly successful national business model which has been blown out of the water by the triple challenge I just mentioned: the famous triad of cheap energy from Russia, cheap security from the United States, and easy exports to China. Crash, crash, crash. All three have gone. Actually, China is doing to Germany what Germany did for much of the world—the mercantilist model of export, export, export—and now it is turning the other way around. China is actually catalyzing a rapid deindustrialization of Germany. Tens of thousands of jobs are being lost in the car industry. The situation is dramatic in the German case.

Unlike France and Britain, what Germany has is money in the public purse. Friedrich Merz has roughly a trillion euros to spend over the next few years on defense and infrastructure. The question becomes: why is Merz not doing so well? I do not think it is so much about the individual. It is true that he does not have much government experience, so he does not always know exactly what levers to press, and sometimes he “shoots his mouth off,” but I do not think it is primarily that.

I think it is what Jarosław Kaczyński, the Polish populist leader, used to call “impossibilism”—the impossibility of making things happen. It seems to me that there is a real structural problem in Germany. This system, which was designed to prevent the emergence of another Adolf Hitler—a federal country, decentralized, with lots of checks and balances—has in the meantime acquired so many more bureaucratic and party-political checks and balances, including a complicated coalition, that it is actually very difficult to change things in a big way. Even a Maggie Thatcher, arguably, would be frustrated in the German system.

The answers are to hand. The big European problem is quite simply the gulf between our huge potential and our actual. Mario Draghi in his report and Enrico Letta in his report have told us what to do: make it a proper single market; have a unified capital market; have a single digital space; have a Europeanized defense industry. There are ways in which we could get much more dynamism back into Europe economically as well as politically, but at the moment the politics, which are still national, are simply preventing us from doing so.

Mounk: I have a few thoughts on this. First, it strikes me that if you talk to any leading French politician or civil servant, they are very impressive people. They are often somewhat provincial—many of them do not have much experience outside of France and do not necessarily speak good English—though that is changing in the younger generation. But they are intellectually brilliant, highly educated, erudite, and hardworking; they are elites in a self-conscious way, with the education to show for it. Yet, France has, by and large, been quite badly governed for the last 60 years.

You speak to German elites and, while there are exceptions, they are mostly provincials. Of the three candidates Germans could choose from for Chancellor in 2021, Olaf Scholz, who became Chancellor, had what I think Churchill said about Baldwin: a “municipal mind.” He had been a somewhat successful mayor of Hamburg, but was far from an impressive visionary or leader, with no charisma. Armin Laschet, whom I spent an evening with at a conference recently, is a very pleasant guy—he drinks five shots of schnapps, smokes three cigarettes, and is very jovial—but he really does not have a great understanding of world politics. He simply does not know or understand basic things about the United States, for example. Annalena Baerbock was pushed as an exciting figure—she is now the Secretary-General of the UN General Assembly—but I think she has very deep political and other weaknesses as well.

It is striking that you have a country like Germany, an impressive place with so many smart, hardworking people, and yet the top political personnel is just far less impressive than that of most peer countries. There are structural reasons for that, such as the federal system, but it is striking because Germany has been quite well-governed for most of the postwar period. However, I think that is because they stumbled upon a model that worked. As long as that model did not need changing, things were okay.

The German political intellectual class has learned the slogan with which Konrad Adenauer won a number of elections in the 1950s and 1960s: Keine Experimente—no experiments. Let us just stick with what we have because the alternative is “who knows what.” Now, they have in part demolished that model. Angela Merkel’s decision to switch off nuclear reactors and instead import coal from Poland—which is effectively what happened—has led to very high energy costs that are strangling German industry.

In part, the model just does not work anymore. As Constanze Stelzenmüller has said, Germany used to outsource its energy needs to Russia, its market to China, and its military needs to the United States. None of those three parts of the model work anymore. There seems to be a lack of serious conversation about this in Germany. People are talking about it, but not in a serious way, and there is a lack of imagination for what the new role of Germany in the world could be. There is a complete failure to recognize that these changes need to happen.

That is where the second point comes in: the coalitional math. Because the populists are now strong on the right and, to some extent, on the left, you never have an ideologically coherent government anymore. It is always some form of “grand coalition.” At the moment, it is technically a Große Koalition—a coalition between the Social Democrats and the Christian Democrats, the two traditionally dominant Volksparteien in Germany. Before that, it was an effective cross-ideological coalition between the Social Democrats, the Greens, and the Free Democrats. You always have someone blocking change in any direction. In the current government, a big part of the reason for Merz’s failure is the complete immobility of the Social Democrats, who have just not understood anything about this moment and are blocking any attempted reform.

Garton Ash: I have to say, I think you’re slightly over-egging the pudding in your description of the German versus the French elites and political elites. I’ve met some very impressive people in Germany; traditionally post-1945, the most impressive people have been in business or science. That was traditionally the way it worked.

Actually, Berlin—I was just there a couple of weeks ago—is a very interesting, lively place with a very interesting think tank landscape. If you look at the rethinking on Russia and on security and defense policy, I would say it’s more impressive than that in London and Paris, partly because more rethinking needed to be done. But above all, what I want to say is this: I don’t think the perennial slogan of the politics of the Federal Republic of Germany was “no experiments.” Schröder’s labor market reforms in the early 2000s were quite expensive. I think the slogan or the motto was “change through consensus.” That was the key to the success of the German system—not just consensus within Parliament or consensus between the federal government and the Länder, but also, of course, between capital and labor.

It’s a phrase that I owe to my good friend Michael Mattis: “change through consensus.” The problem is, as you rightly point out and as I was saying a moment ago, that the system has become so complicated with the fragmentation of the party landscape—which is absolutely characteristic of Europe in our time—plus this accretion, sort of barnacles on a ship’s hull, of not just constitutional checks and balances, which are already very large, but regulatory and bureaucratic and procedural checks and balances. “Change through consensus” is just proving extremely difficult to achieve, even if you have a trillion euros to spend.

The larger point I just want to make—I think I’ve made this to you before, but I want to make it because it’s so important to understand: in most European languages, there is no separate word for “policy.” The key structural problem of Europe—I mean the European Union—is that the policies we need are European, but the politics are still national. We need European-scale defense industry, defense policy, capital markets, etc. But the politics keep holding us back because they’re always national. For me, the future of Europe hangs to a large degree on how we manage that tension. Obviously, the logical answer would be to say, well, let’s make the politics European. Let’s have European political parties and direct elections to European Parliament. Been there, tried that; it hasn’t got us very far. So my view is just get on with it and try and make it work.

Mounk: That is a huge structural challenge where a lot of important decisions are now taken at the European level, but it is very hard to muster the will for some kind of coherent change at any of the national levels, and even harder to then bundle that at the European level.

Garton Ash: One consequence of that is that you only need a single veto player in Brussels in the decision-making of the European Union. This brings us on our little tour to Hungary because the prime minister of one small European country, because it was a member state of the EU, could hold up packages of sanctions that everyone else wanted against Russia. He could hold up 90 billion euros, which are crucial to the future of Ukraine for the next two years, until he exacted his price.

The unique structural nature of this voluntary empire, which is the European Union—this empire by consent which empowers small and medium-sized countries—enables them to block larger countries because of the requirement for unanimity. It is a beautiful thing and something new in history, but it does mean that national politics can put an enormous spanner in the works of a very big machine.

Mounk: I see on American lists of “great German words” many phrases that would not be instantly recognizable to most Germans. They might understand them in the sense that they can scan the components and grasp the meaning, but they are not existing idioms. However, there is a lovely German phrase that really does exist: Besitzstandswahrung.

It is the idea that people are motivated by preserving what they have—the protection of vested rights or acquired status. It applies to the German political scene and society, and it is one reason for the current lack of imagination. It applies in France as well. I took part in an exercise called France 2050, which was forward-planning by the French government. An interesting poll found that most French people think things will be a little worse in the future than today; they are pessimistic, but they think things will not change that much.

It strikes me that this is the attitude of many Europeans today outside of France as well. I do not think it is realistic. Either Europe reforms itself properly—in which case things might be better in 20 or 50 years—or it fails to reform, and things could really get a lot worse.

To what extent is Besitzstandswahrung the real condition of Europe? We see it in Britain with the triple lock on the pension system. We see it in France with the rage at Macron for attempting reforms. In Germany, perhaps some think-tank people in Berlin are starting to understand the need for change, but certainly it is not translating into policy, and it is not really translating into a demand for real change from voters either, other than perhaps relative to immigration with people migrating over to the Alternative for Germany. To what extent is Besitzstandswahrung the real condition of Europe, and can the continent overcome it?

Garton Ash: I think that is the question of questions. Because we are still, in spite of all the multiple crises, on the whole—in most, certainly, Western European countries—too comfortable. And look, the largest war in Europe since 1945 has been going on just next door for more than four years, longer than the Great Patriotic War, Russia in the Second World War. I think as of June 11th, it will be longer than the First World War. And most of life just goes on as normal in Western and Central Europe.

What I have just written is a new book, a very short introduction to Europe. One of the points of the book is precisely to say to Europeans: wake up; learn from history. If you do not change, then things can collapse quite suddenly. Tancredi in Giuseppe di Lampedusa’s The Leopard said, “Things must change so they can remain the same”—the famous, oft-repeated line. I think that is exactly where we are.

I think there is an intellectual understanding of that among a lot of the European elites. If you go to a policy intellectual conference, people will absolutely understand that. But does it translate into our politics, which are all about defending our pensions, defending the welfare state, and defending this and that? I think not. And therefore, the quite possible, perhaps even probable scenario is that Europeans in their majority, particularly Western and Central Europeans, will choose gentle decline and suddenly one day find out that, like the old joke about going bankrupt, it is not gentle; it is quite sudden.

Mounk: That is my fear as well: that there is this option of gradual, genteel decline in the minds of a lot of Europeans that may turn out not to exist.

We briefly alluded to Hungary. I feel like a good conversation makes people depressed and gives them some hope, so let us try and give people some hope about Hungary and perhaps about Ukraine for the rest of this conversation. Viktor Orbán ruled for 16 years. He clearly undermined democratic institutions in a very serious way. There were huge amounts of personal corruption with a great enriching of the entire environment of Viktor Orbán, including old friends and relatives. You had a genuinely unfree media landscape. I remember when I was doing some reporting in Hungary, seeing every newspaper reporting on a speech by Orbán with the same flattering photograph and the same positive write-ups—something quite striking to see. A lot of people thought that it would no longer be possible to displace Orbán from elected office at the ballot box.

He finally overstayed his welcome, in part because of some straight-up scandals of governance—pardons for people who were running institutions in which child abuse had taken place and things like that. There was, ironically in part because of electoral rules that Orbán introduced to boost his majority, a particularly strong majority against him in the new parliament.

I have two questions about this. First, is democracy more resilient than we thought? Is the fact that the opposition was able to displace Orbán at the ballot box—that he did not call out the military or somehow try and sustain himself in power, and that the opposition now has a two-thirds majority—a sign that democracy is more resilient than we have given it credit for in the last 10 or so years?

Secondly, what are the dilemmas of post-populist rule? The new government will have a choice: either tolerate a lot of Orbán loyalists who are not committed to democratic rules and norms and who often are not qualified for their positions—keeping them in place, which has obvious downsides—or try to throw them out of office. But that, of course, just normalizes the rule that every new government comes in and throws out the old guys. It could even, at the extreme, lead to an over-consolidation of power in the hands of a new prime minister.

In general, should we be more optimistic about the resilience of democracy? And can the new government deal with this post-populist dilemma?

Garton Ash: I booked my ticket to Budapest many months ago because I have known Viktor Orbán since 1988. I first met him when he was a seemingly idealistic student leader, just three months after they had founded this wonderful new party called the Young Democrats, Fidesz. I still actually have the notes in my notebook where they were telling me about the rule of law, multi-party democracy, and all these wonderful things that he proceeded to destroy. I wanted to be there to see his fall.

The day and the evening in Budapest were magical; they even exceeded my wildest dreams because it was not a level playing field. For all the reasons you have given—gerrymandering, media control, abuse of the state administration, and abuse of state funds—it was an uphill battle. But there was such a massive outpouring of popular will and a sense that it was time for a change that it just swept all those obstacles away.

What you had was this overwhelming sentiment, mainly about social and economic conditions, meeting a seemingly credible change. Péter Magyar had a very clever campaign. He only gave one interview to the international media. He did virtually nothing in Budapest. He did not touch the classic urban liberal themes; he just went from small town to small town and village to village talking about social and economic issues and corruption. The “time for a change” sentiment met what looked like a credible change, and it was a fantastic moment.

Now, to the question you posed: you and I have for some time been saying—or I, at least, have certainly been saying—that Hungary is no longer a democracy. We have a member of the European Union which is no longer a democracy; it is a competitive authoritarian or electoral authoritarian system. So how come he could win an election? I think there is an answer to that. Elections are dangerous moments, even for competitive authoritarian regimes, even with all those advantages. Remember Slobodan Milošević was toppled by an election in 2000. Remember the spark for the Orange Revolution in 2004 was an election.

Elections in competitive authoritarian systems are unpredictable. On the day, it was relatively clean—it wasn’t massively rigged like an election in North Korea. They thought they had the system under control, as they had for many years. The difference from other cases is that those often involved an element of violence. To our astonishment, at about 9:20 p.m. on election night, a message flashed on Péter Magyar’s Facebook page: “Viktor Orbán has just rung me and congratulated me on my victory.”

I think you have to say that being inside the EU was a constraint. Were you really going to send in the thugs and use the Russians to try and falsify what was clearly a landslide result through force and fraud? It remains true that it was a competitive authoritarian system, but it was overcome by the combination of a “time for a change” sentiment, a credible alternative, and the external constraint of the EU.

To your second question: it is absolutely fascinating because, as you know, in Poland, we had another wonderful electoral moment when Donald Tusk was re-elected in autumn 2023, and this has gone completely pear-shaped. We have what is now famously called the “post-populist trilemma.” There are three things you want to be: rapid, effective, and legal. The trilemma says you can only be any two of them at once. Poland is now completely stuck in a situation of legal chaos, something almost like what Trotsky called “dual power” between Donald Tusk’s government and the Law and Justice president.

Fortunately, there is very good news about Hungary. Because the incoming TISZA government has a constitutional majority, and because of the way Orbán built his system—wanting formally to comply with EU rules while violating the norms—it is in a much better position. It can, with a bit of luck, restore the legal order and a neutral civil service, and legally replace the people Orbán put in place. It will not be easy. There will be local power holders in place and massive issues with corruption. But for me, if you can undo such far-reaching state capture by essentially legal and constitutional means, that will be a very positive sign for the rest of Europe and perhaps even for the United States.

In the rest of this conversation, Yascha and Timothy discuss the Russia-Ukraine war and the future of the relationship between Europe and the United States. This part of the conversation is reserved for paying subscribers…

Read more

Laurenz Guenther is a political economist at the Toulouse School of Economics and a Fellow at the Institute for European Policymaking at Bocconi University. His research and Substack focus on representation, populism, and immigration in Western democracies.

In this week’s conversation, Yascha Mounk and Laurenz Guenther discuss why there’s a massive representation gap between political elites and voters on cultural issues, how this explains the rise of populist parties like the AfD in Germany, and whether new parties can successfully occupy the economically left but socially conservative political space.

This transcript has been condensed and lightly edited for clarity.

Yascha Mounk: You managed to write one of these papers that goes viral quite quickly because it really shows something interesting. Often in the social sciences, the best kind of papers show what we all already kind of knew, but actually demonstrate it in a thorough and methodological way. What you show is that there is a significant gap in representation, particularly of representation on cultural issues, between political elites and ordinary people. To me, the most striking graph in this paper came from Germany, and it looked at attitudes about immigration among members of the Bundestag, among elected politicians, and ordinary voters—I believe in 2013. Tell us a little bit about that specific data and what it shows us.

Laurenz Guenther: What we have is answers to surveys of representative samples of citizens and also of parliamentarians. These two groups answer the exact same question, which enables comparability. The example that you mentioned is about immigration—to what extent immigration should be facilitated or restricted. What one can see in this graph is that Germans, of course, have heterogeneous preferences, but most Germans want to restrict immigration to Germany. What I show is the average response of the members of all the parties that were relevant at this point in time in Germany, and all of these parties—measured by the average position of the member—wanted to facilitate immigration. There was a huge mismatch, even in the direction, in the sense that most people wanted to go to the right on immigration, but all the parties wanted to go to the left.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

Mounk: One of the striking things about this is that this is before the rise of the Alternative for Germany, which I think was founded right about then, but was not yet represented in the Bundestag. The most right-leaning political party in the Bundestag was the Christian Democratic Union, which was led at the time by Angela Merkel. The view of the average parliamentarian—not just in the Bundestag, but in the Christian Democratic Party—was way to the left of where the average view in the population was. Is that right?

Guenther: Yes, this is right. All of these people who are right-leaning on immigration—from their perspective, it must have looked like everyone, even the supposed right-wing politicians, were much more left-wing. They had absolutely no representation on this topic. This is one of the ideas of this paper: this provided fertile ground for the AfD, which then subsequently also rose.

Mounk: Over half of the population didn’t really have their views represented in the Bundestag. What happens? The AfD was founded as a political party by these slightly stodgy economics professors who were really worried about the euro and opposed the single currency. But the longer the party existed, the more it focused on issues like immigration. By the time it managed to get elected to the Bundestag in 2017, its main focus really was on restricting immigration. What this paper strikingly shows is that there was all of this fertile ground in which it could fish for voters, because so many voters weren’t represented by the pre-existing political parties.

How general is this? That is a really striking data point about Germany. To what extent do you have similar data about other countries? To what extent can we generalize from the existence of this kind of cultural representation gap in the German case to the existence of similar cultural representation gaps in other countries in Europe or beyond?

Guenther: The aim of this paper is to do that more systematically—to look at other countries and other issues. This can be generalized across European countries, and across cultural issues. Looking at 27 European countries, I find the same patterns on cultural issues for all of them. By cultural issues, I mean immigration, but also issues like gender relations, punishment for criminals, assimilation, and multiculturalism. On all of these, voters are much more right-wing than the parliamentarians of their countries. On economic issues it is much more mixed and the gaps are smaller. Notably, there is a great article in the Financial Times where this analysis was also extended to the United States, and there you find a similar pattern.

Mounk: Tell us a little bit more about each side of this. On the cultural representation gap, what other metrics do you look at where the views of ordinary citizens tend to be quite far apart from the views of parliamentarians?

Guenther: Immigration is certainly one of the issues where you have one of the largest differences. Another issue where the difference is similarly large is punishment for criminals—should sentences for criminals become harsher? Here again, majorities in basically all European countries say that this should be done, but parliamentarians disagree. This is a directional disagreement: majorities of parliamentarians disagree with that position, while around 70% of the population say that it should be done. There are also big differences on gender relations and European unification. European unification, however, is not so directional. Parliamentarians seem to be very strongly in favor, and voters are also somewhat in favor, but apparently want a much slower unification.

Mounk: There is an interesting contrast here. On some issues, political elites feel really strongly about something, but they are actually going in the same direction as ordinary voters. But when talking about things like immigration in particular, and to some extent assimilation and how strong criminal sentences should be, they are not just far apart—they are going in opposite directions. The average view of a parliamentarian is that immigration is a good thing for the country. The average view of a voter is that immigration is a bad thing for the country. That feels like a more significant gap. Even if on a 1 to 10 scale the gap is three points on each of them—and I don’t know exactly what the gap is on your scale, I’m making these numbers up—if that three means that overall preferences go in different directions, that seems to matter more than if three means one set of people is very enthusiastic and the other is somewhat enthusiastic.

Guenther: In the public discourse, people sometimes don’t use that term, but this idea often comes up that there may be differences between parliamentarians and voters. It is always framed in the sense that parliamentarians are somewhat ahead and voters want everything to go a little bit more slowly—which, of course, adds a normative dimension, and that is a whole different issue. But it also does not really address these findings. Because, as you say, this is not just a matter of everyone in principle wanting the same thing, with some wanting it a bit more quickly and others a bit more slowly. It is really that parliamentarians and the average voter disagree on the goal—on where we should be heading as a society. That is indeed a very different thing.

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

Mounk: Tell me a little bit about those economic metrics, because I think that especially on the left, a lot of people want to think the real disjuncture between voters and the people is on economic issues. People want redistribution, they want a robust welfare state, and then there are the evil political elites—probably from a much more affluent milieu, who have to find donations for political parties, especially in the United States, but also in Europe—who are moving in fancy circles and don’t want any redistribution. That is really where the gap lies. Your data seems to suggest that the story is much more complicated. There are obviously some gaps on various issues of economic policy as well, but they seem to be much smaller on average than on cultural issues. Tell us about the extent to which the views of ordinary people and of political elites match up on economic issues.

Guenther: On economics it depends a bit. On redistribution, there is an item that asks people about redistribution, and qualitatively this is indeed what people on the left would think: ordinary people are a bit more in favor of redistribution than parliamentarians. Notably, wealthier ordinary people—those above median income—have similar preferences to the average parliamentarian of their country, and the gap is driven by the poorer half of the population, who really want much more redistribution. That makes a lot of sense.

But this gap—and I would have to look up the exact numbers—is much smaller than on cultural issues, by a factor of roughly five. The gap on immigration is really about five times as large as the gap on redistribution. On other economic issues it depends a bit more. Generally speaking, it looks like people want a bit more redistribution than parliamentarians, and they want less state intervention.

One thing that is missing here is trade. In the dataset that I use, there are no questions on trade, so there may be a representation gap there—we don’t really know.

Mounk: We would expect political elites to be probably more pro-free trade and ordinary people probably less so. There is another way of thinking about this, which is a two-dimensional graph where you look on one dimension at where people are on social and cultural issues and on the other dimension where they are on economic issues. What tends to be well represented, particularly in two-party systems but to some extent even in systems of proportional representation, is the things that go on the axis from the bottom left to the top right. On the one hand, there are right-leaning political parties that want less redistribution and are reasonably restrictive on migration. On the other hand, there are left-leaning political parties that are quite open to immigration, or even want more immigration, and are also pro-redistribution.

On one side, there are libertarians who are very socially liberal but want less economic redistribution. Those tend to be overrepresented in elite political discourse, but they are actually a relatively small part of the population. There is a much larger part of the population that occupies what is often called the populist quadrant—people who actually want a reasonably high level of redistribution, who do not oppose the welfare state and are not libertarians on economic policy, but who are quite conservative on social issues.

How does your research intersect with that line of thinking? Do you think it is basically right that it is that last quadrant—the people who are socially quite conservative but economically and fiscally reasonably progressive—that are most underrepresented?

Guenther: If you think about these four quadrants, this is the group that is least represented. In this two-dimensional space, the parties in most countries fall along a diagonal. The more you condition on political knowledge or participation in politics, the more you get a strong correlation to that line.

Subscribe now

Mounk: People who are very engaged in politics, if they are economically liberal, are also likely to be socially liberal. But people who are not very interested in politics are going to have views that are more all over the place, which doesn’t necessarily mean they are less coherent. They are just not bundled in the way that our political system conditions us to bundle our views.

Guenther: For the general population, views are more dispersed—it looks more like a circle, like a two-dimensional normal distribution. What my paper shows is that this line of the politically active has actually shifted downward. If you think about that, you can see that the people who are socially conservative and economically left-wing are particularly far away from the closest party in many countries. There is no major party that bundles their views, which I would predict would be an opportunity for new parties.

The populists that are very successful in many countries usually supply this policy position for the upper-right quadrant—that is, for people who are socially conservative and also economically right-wing, even though they are often quite flexible on economics. If you plot them over time in this space, you can see how they move around somewhat to attract more voters from the left.

Mounk: The one thing that defines right-wing populism today in most contexts—perhaps excluding Argentina or certain other countries—is not economics. It is those social and cultural issues. The AfD is interesting in that sense because it was founded on an economically populist issue, but over time it really became defined by those social and cultural issues much more than economic ones.

The economic views of populist parties are a little bit all over the place in different countries. A party like the AfD, in part because of its roots, is probably mostly right-leaning on economic issues, though it certainly isn’t a radical libertarian party and doesn’t want to abolish the welfare state. But a party like the Rassemblement National in France is much more left-leaning on economic issues. They have promised in many ways to preserve the welfare state and opposed Emmanuel Macron’s pension reform, wanting to preserve those entitlements for people. Even Donald Trump plays a strange role in this. On the one hand, a lot of the economic policies he has passed have been quite right-leaning economically—huge tax benefits for the rich, much less for ordinary people. But when you look at how he distinguished himself from his Republican competitors in the 2016 primaries, it was in part by saying things like, perhaps the state does have a responsibility to make sure that everybody has access to healthcare, which was something that the other fifteen Republican candidates did not say.

Why is it so hard for populist parties, particularly in systems of proportional representation, to really appeal to that quadrant? There is something surprising about the fact that you don’t see more political movements move squarely into the space that is economically relatively left-wing and socially right-wing. We have seen an attempt at something like that in the Bündnis Sahra Wagenknecht in Germany—a party that for all kinds of reasons I personally don’t particularly like. They nearly got into the Bundestag in the last elections but fell just short by a few thousand votes and now seem to be falling apart. It seems to be hard for political parties to move into that space, even though that clearly is where a significant portion of the electorate is. Tell us a little bit about who Sahra Wagenknecht is and what this party is.

Guenther: Let me answer this in two steps, looking at two parties that may have moved into that space: the AfD and the Bündnis Sahra Wagenknecht.

The AfD was founded as an economically right-leaning party, largely driven by economics professors opposed to the European Union, and then became very anti-immigration during the refugee crisis. At that point it was probably already positioned, just because of its members, as an economically right-wing party, and it is difficult to move from there. This plays a big role: it is difficult to completely change your position over time, which I think is also one of the reasons why established parties don’t do that. To really capture this quadrant, a party would have to move so much that it risks a split. It would have to make much bigger moves than, for instance, Merkel did during the refugee crisis, which was already a big stress test for the party.

One thing worth mentioning briefly is that in Germany, the quadrant of economically right-wing and socially right-wing people is quite a bit larger than the quadrant of socially right-wing and economically left-wing people. The latter—the quadrant we are interested in—is still a big quadrant, larger than the libertarian quadrant. But if you want to be the socially conservative party and you are simply thinking about maximizing votes, it makes sense to first go after the people who are right-wing across the board, then perhaps move to the center, and then take the second quadrant. The AfD does seem to be trying to do something like that.

Now, turning to Sahra Wagenknecht. She was for a very long time one of the most prominent members of Die Linke—the left-wing party that is basically a descendant of the former Communist Party of East Germany. It is a minor party, polling around five to ten percent, that is very economically left-wing but also very socially progressive and liberal. Wagenknecht distanced herself from this party, largely driven by her views on cultural issues, in particular immigration, on which she was more conservative than much of the rest of her party. This ultimately led to a split in which she left the party together with a few other members of parliament and founded her own party, Bündnis Sahra Wagenknecht—Alliance Sahra Wagenknecht.

This party initially did very well in the polls and was seen as considerably more successful than the original left-wing party. In the election, however, they nearly failed to clear the 5% threshold required for parliamentary representation, while Die Linke did make it into parliament. In some sense, it was a respectable performance—it is very rare for new parties in Germany to clear that threshold, and it was the party’s first election. Now they seem to be falling apart.

My sense is that on cultural issues they were not right-wing enough to capture the quadrant of socially conservative, economically left-wing voters. I think in the beginning people expected Wagenknecht to be more conservative on social issues, and that position appeared to be significantly watered down over time, which I think was a problem for the party.

Mounk: That is interesting. She is also a very charismatic but polarizing figure, and there is always a tension between a political party really just being a personal vehicle for its leader. The party’s name, BSW, literally consists of her initials—Bündnis Sahra Wagenknecht, or Alliance Sahra Wagenknecht. She has now distanced herself a little from the party, and that probably means it is not going to survive. But it was an interesting attempt at trying to capture that quadrant of the electorate.

Let me present to you a reading of your paper which I know you deliberately don’t put in the paper, but which I think is how a lot of people have read it. Tell me to what extent, speaking as a private citizen rather than a scholar of social science, you agree with that interpretation.

Here is the first part. Here is the real explanation of populism. We have been having this debate for ten or fifteen years about why populism is rising around the world, and people point to this explanation and that one. The one thing nobody has really been talking about is whether the problem is simply that the established political parties aren’t listening to what voters want. You might agree with the views of the established parties, or you might agree with the views of voters. On many things, I’m probably closer to the views of members of a political elite, because having a PhD in political science, I’m virtually by definition a member of the political elite. But when we collectively become so distanced from what ordinary people think, they are going to get upset. What happened here is that over time, this cultural representation gap increased. People no longer listened to voters. Election after election, people said they wanted more controls on immigration. Elite political parties did not deliver on those preferences. Finally, voters rebelled by voting for the one party willing to give them what they want.

The second part is a more provocative statement. Perhaps one of the things that established political parties need to do in order to deal with the rise of populism is to actually listen to what people want and adopt some of those policy positions. Not every single established party needs to do this. In the German context, it probably makes sense for the Green Party to continue to be very pro-immigration, because its electorate is very pro-immigration—and that is the virtue of a system of proportional representation. But if the Christian Democrats or even the Social Democrats want to compete with the AfD for the many voters who have shifted towards it over the last few years—a party that now polls roughly equal with the Christian Democrats in first place, having grown from less than 5% of the vote in 2013, around the time of your paper, to around 23 to 25% in polls today—the straightforward thing to do is to get closer to what a lot of ordinary people want. Is that a plausible interpretation of your paper?

Guenther: I think that is a plausible interpretation. The first part stays relatively true to the paper. The previous literature has done a lot of very valuable work, and what I do is not a substitute for that—I’m not saying everyone else was wrong. I have the impression that the previous literature just didn’t really look at this specific part of the puzzle. It looked more at how the financial crisis contributed to making people vote for populist parties, what the loss of manufacturing jobs did, and so on. But there was less focus on the choices of mainstream parties and in particular how they would respond. That ties into the second question, which I also think is a fair interpretation.

This paper is also interesting because it looks at a variable that can be easily influenced. We have all of these variables in mind when it comes to populism—lack of trust, slow economic growth, certain cultural characteristics, and so on. But these things are more or less given and very difficult to change. If you take the estimates seriously, winning back voters by increasing economic growth would require growth rates that are just completely unrealistic. What this paper—and several related papers—does is look at the positioning of parties, which is a variable that can be adjusted relatively easily. In that sense, I do agree that if mainstream parties want to win back voters and weaken populist parties, they have to move their political positions.

Importantly—and this is often a point of confusion—they need to deliver on policy outputs, not just rhetoric. There are some papers that show it backfires if parties merely shift their rhetoric. There was a famous speech by Starmer—“island of strangers” and so on—and the analysis of that suggests it probably backfired.

Mounk: The idea here being that if politicians say they have heard the voters and are going to do something about immigration, and then there is no action to follow up on it, voters conclude that the politicians are simply being hypocritical and cannot be trusted.

Guenther: The facts on the ground must change. In some sense, this is a chance—and maybe the last chance—for the mainstream parties, because populists are often not that good at delivering real results or delivering on their promises, at least in their first attempt. During the first Trump administration, for instance, there was no major deportation effort. The second time, Trump is delivering on that: immigration has fallen dramatically and deportations are proceeding at scale. He learned from the first term and delivered on those promises in the second.

Something similar may be happening in Europe. Often, in their first time in power, right-wing populists don’t really deliver. We have seen this with right-wing populists in Austria, with Meloni, and others. Mainstream parties therefore have a chance to deliver now, or even if populists come to power, to use that period to develop a new strategy and new policy positions. If mainstream parties then win the election after the populists leave power, they have another chance. But I do think it is an increasingly difficult situation for mainstream parties across Europe. They have one or two election cycles to deliver. If they don’t, populist parties could, in the medium term, simply outperform them and come to dominate European policymaking.

Mounk: What do you say to people who claim that when mainstream parties try to emulate populists, it actually only reinforces the populists? This is a strand of research that I have seen repeatedly, and more than anything else it seems to be conventional wisdom among a lot of political scientists—that there are these studies, sometimes somewhat dubious studies I think, that demonstrate that when mainstream political parties start to use the verbal register of populists or emulate some of their policy positions, voters are simply going to vote for the original. This is an argument you hear a lot. How convincing do you find it, and why does it sound like you don’t agree with it?

Guenther: This depends a lot on whether you deliver results or whether you do something else—which is why I made the rhetoric versus results distinction. There are a lot of studies on this and they show mixed results. Some find that people vote for the original populist party if a mainstream party moves toward the populists. An example is the analysis of the Starmer “island of strangers” speech that I alluded to before. Other papers show the opposite: that if a party moves to the right on immigration, voters shift toward that party and away from the populists. There is a study in Denmark that shows this, and we also have a study done in Germany just before the most recent election that finds that if the CDU moves to the right and fills the representation gap, the AfD loses voters and the CDU gains voters.

I do think it depends, and it depends most strongly on whether actual policy positions change. If the policy positions really change and the output is delivered, people do not simply vote for the original.

One thing that is often forgotten in this context is that right-wing populism in Germany is only in some sense a third wave. There were early right-wing populists just after the foundation of the German republic who were strong in the 1950s. Then there were the Republikaner—the Republicans—in Germany in the 1980s, who were in some sense similar to the AfD. In both cases, the CDU—the centre-right party—accommodated them, moved toward their positions, and in the 1950s under Adenauer even gave them ministerial posts. Later, in the context of the asylum wave from Yugoslavia in the late 1980s and early 1990s, immigration had become a major issue and people wanted less of it, and the constitution was actually changed to limit asylum immigration significantly.

Mounk: Your point is that it worked at the time—the Republikaner really were a kind of insurgent populist force in the 1980s and 1990s. They had some significant successes, ended up being represented in a number of state parliaments, but never quite made it into the Bundestag. What you are saying, I take it, is that the Christian Democrats moved to the right in order to limit the oxygen for that political party. Rather than people saying that this somehow reinforced their impression that the Republikaner were onto something and they should vote for them, they said that some of their concerns were being taken care of by the political party they traditionally voted for and moved back toward voting for the Christian Democrats.

On this telling, it was Angela Merkel’s very deliberate strategy of capturing the political center—and no longer covering the right of politics—that allowed this insurgent movement to establish itself as a permanent political force. There are two ways of thinking about this. Strauss, a very influential Bavarian prime minister, always used to say that between the Christian Democrats and the right, nothing should be able to fit—meaning that he always wanted to make sure that the Christian Democrats, as a democratic political party, would cover the flank on the right far enough to prevent any party to their right from establishing itself. Merkel’s approach was very different: she would make the Christian Democrats into the party of the political center. But of course, that raises the question of who deeply conservative people—those who don’t think of themselves as being in the political center—should be voting for.

One way I have sometimes put this about German politics—and I am obviously a German citizen—is the following. I have been politically socialized on the left and continue to think of myself as being on the left, even though I have significant criticisms of the shape that the left is taking in many countries today. If I look at Angela Merkel and find that I share a lot of her basic value coordinates—even if I think she made a lot of bad decisions—we have a problem, because there are a lot of people in Germany who are considerably more conservative than I am. If the leader of the most conservative democratic political party in Germany holds views that someone on the left feels broadly comfortable with, that may be fine as long as that person is in charge. But it leaves a huge segment of the German electorate without political representation. What happens? They go and vote for the AfD.

There is another line of argument, which is that left-wing political parties in particular have not made sufficiently large redistributive demands. There was a period in which the left and the right had very similar positions on economic policy. If you look at Germany today—the Social Democrats and the Christian Democrats—or at Britain—Labour and the Conservatives—they are not that far apart on a lot of economic policy issues. People therefore no longer invest significance in economic policy, and that is why cultural issues become so salient. Why are we debating immigration? Because if there are no deep ideological differences in economic policy, that is the only thing left to debate. On this argument, what is really explaining the rise of right-wing political parties is the convergence on economic policy, and one way to fight that—particularly for left-wing political parties—is to move to the left on economics, making that a more charged political issue and shifting the public debate away from issues like immigration. This is an argument that has been made by a number of political scientists and that I often hear in the debate. I imagine that you are skeptical about that line of argument as well.

Guenther: I am indeed skeptical about that line of argument. It leads to an interesting deeper question, which I will address first. I do believe that in principle the theoretical argument is correct. All else being equal, if left-wing parties moved further to the left, we would probably talk more about economic topics and, because people’s attention is limited, they would probably think a bit less about cultural issues. However, I would guess—and I don’t think we have particularly strong studies on the relative importance of this effect—that it is a relatively minor thing. That is my main criticism. I’m sure the effect probably exists; all kinds of effects exist. But how big is it?

I don’t think people will stop caring about immigration because talk shows discuss economics more, or because the policy options on offer are somewhat more distinct from each other. This belief is driven by two things. The first is simply talking to Germans. I am German, most of my friends are German, and I do a lot of interviews and survey work where people tell their stories. Immigration and issues related to it is such a huge issue, and touches on such fundamental fears about identity and belonging, that I think it will always be salient for these people as long as immigration is high or looks like it could increase further.

The second thing is that if you look at the immigration that people say they care about—specifically immigration from the Middle East and Africa—the actual numbers of arrivals from these groups are extremely highly correlated with how much people care about the issue. During the refugee crisis, for instance, people were asked how important immigration was to them, and that line tracks extremely closely with the actual number of asylum applications in Germany. This suggests that concern about immigration is tied to things that really happen on the ground.

Notably—and this is an argument that is often made, but I think it is close to a straw man—how much people care about immigration is not really predicted by overall immigration numbers, that is, by how many people come to Germany in total. Most people don’t really care about, say, a French student coming to Germany to study. It is about very specific groups, and the correlation with real-world arrivals of those groups is clearly there.

Mounk: You said a little while ago that perhaps the last chance for established political parties to win the race against the populists is not just to shift their rhetoric, but to actually show results. The problem is that especially in systems of proportional representation, this becomes more and more difficult the higher the share of parliamentary seats populists hold. In Germany now, it is barely imaginable that there will be a right-wing governing majority in the Bundestag that excludes left-wing political parties while also maintaining the Cordon Sanitaire—the Brandmauer—the separation between the traditional democratic parties and the new right-wing populists.

The last German government was a left-leaning coalition in which the Social Democrats and the Greens governed alongside the right-of-center Liberal Party, the FDP. Now there is a conservative chancellor in power who is in various ways more conservative than Angela Merkel, but his coalition partner is the Social Democrats, the left-of-center political party. Even though Merz was talking a great deal about the need to curb immigration during the election campaign, it is actually very difficult for him to deliver on those policies because his coalition partners in the Social Democrats are opposed to many of them.

Do you think realistically that traditional political parties are going to be able to rein in immigration to a sufficient extent to make voters feel that their preferences are being represented? Or do you think we are now in a structural situation in which the inability of mainstream political parties to deal with that issue is simply going to lead to their continued decline in vote share and to parties like the AfD continuing to grow?

Guenther: This depends a lot on the country. In Germany, also for historical reasons, the resistance to accommodation is particularly high, even though as we discussed, it did happen in past episodes. I would guess—and this has to be taken with a grain of salt—that the current coalition has the last real opportunity to deliver the legislation needed to genuinely reduce asylum immigration and pursue deportation efforts.

After that, I would guess the AfD will be stronger in the next election than in this one, and it will then be very difficult to find a coalition that addresses immigration effectively while leaving the AfD aside. If efforts don’t increase significantly—and I would guess they won’t—the next government will probably also fail to deliver, the AfD will be even stronger, and it will probably eventually become part of the governing coalition. That is my guess. We can check in perhaps nine or ten years.

For other countries, the picture is a bit different. Populists are participating in or supporting governments in several countries. In Sweden, for instance, with the support of the Sweden Democrats, immigration policy has changed quite significantly—net asylum immigration is now reportedly negative. It is possible. Germany tends to look mostly at itself, but in other countries, such as Denmark, these things have happened.

Mounk: Let me ask you about how this theory applies in some other contexts. We haven’t talked much about the United States. America actually, on a lot of the generic questions about immigration that are asked cross-nationally, continues to have more pro-immigrant attitudes than most European countries, which is interesting. The failure of the Biden administration to deal with the southern border with Mexico is clearly one of the big reasons why Donald Trump won reelection in 2024. It was even one of the reasons why new voting groups like Latinos shifted to the Republican Party. At the same time, as you said earlier, the current administration is in some sense very successful in its policies—it has reduced the number of immigrants coming to the United States very significantly. But it has done that using cruel and indiscriminate tactics that turn out to be very unpopular. The immigration enforcement agency is very unpopular, and Donald Trump’s approval ratings on immigration are quite low.

Is that an American specificity, where American attitudes about immigration are simply somewhat more permissive than European ones? Or do you think that if, say, Marine Le Pen—or if she is not allowed to run, Jordan Bardella—became the next president of France and effectively cracked down on immigration, deporting large numbers of undocumented people, they might face a backlash as well? Not just from people with more progressive attitudes who become activated by those policies, but perhaps from some of the swing voters who helped them get into power in the first place—as, according to polls, seems to be happening in the United States.

Guenther: You can always overdo it. The vast majority of Europeans want lower immigration levels, but you can also have levels that are too low. The amount of cruelty—for lack of a better word—that voters are willing to tolerate is certainly limited. I do think that voters will support severe actions to reduce immigration, but there is some limit. The threshold in Europe might be a bit higher than in the United States, so Europeans might be willing to accept more severe measures than Americans, but even there some limit exists. It is possible that populists, when they come to power, overdo it to some extent—one could argue that Trump somewhat overdid it.

Relatedly—and this is a slightly different point—it also depends on how efficiently these policies are carried out. Some degree of severity may be necessary to achieve a goal, but there is also cruelty that is arguably indiscriminate and serves no purpose in achieving the desired outcome. I would guess that because populists are considerably less experienced, they will tend to be less efficient. There will be problems when they come to power and pursue immigration policies. This is something that mainstream parties can exploit by shifting their positions but then arguing that they are the professional alternative—that they will pursue these goals in a more targeted and discriminate way, and that they won’t overdo it. I think that is one way that mainstream parties can make their case.

In the rest of this conversation, Yascha and Laurenz discuss whether journalists and politicians have as much impact on public thought as they think they do, to what extent self-censorship has increased in recent years, and the impact AI will have on the media. This part of the conversation is reserved for paying subscribers…

Read more

Lant Pritchett is a development economist from Idaho. Having now thrice retired, he is currently a Visiting Professor at the London School of Economics in the School of Public Policy and the co-founder and Research Director of Labor Mobility Partnerships (LaMP).

In this week’s conversation, Yascha Mounk and Lant Pritchett discuss why the traditional foreign aid approach to development is fundamentally misguided, how countries actually achieve prosperity through organic national transformation, and whether the classic path to development remains viable in the 21st century.

Will you be in London on Sunday, September 6? I’ll be interviewing Francis Fukuyama about his life and thought to mark the publication of his memoir In the Realm of the Last Man at the Sekforde at 5pm. Find out more and book tickets here. Paying subscribers can access a code for free tickets here. —Yascha

This transcript has been condensed and lightly edited for clarity.

Yascha Mounk: When I talk to people who care a lot about economic development, they love to talk about inclusive development, about sustainable development. They tend to focus on how much more affluent countries should donate to less affluent countries, perhaps figuring out what the best intervention is—whether foreign aid should be spent on this kind of thing or on that kind of thing. Not to simplify, but you think that whole approach is basically wrong. Why is that?

Lant Pritchett: What I think of when I think of development is what we call “getting to Denmark.” There was a historical process whereby many countries—not just Western countries, but other countries—had a fourfold transformation. They had a transformation from a low productivity to a high productivity economy, and it was mostly broad-based. That is economic growth, and inclusive economic growth, if you want to add the adjective. They also went through a transformation of acquiring state capability—the ability of the public sector to do things that needed to be done, like regulation and providing certain services. They also went through a transformation from subject to citizen, to a polity that was based on responsiveness to the needs and wishes of the citizen rather than vice versa. They also went through this harder-to-describe transition of equality under the rule of law, whereby kith and kin and other identities became reduced in importance and everybody was treated equally.

That is what development meant in the post-decolonization era, after World War II, as countries became independent from their colonial overlords. It meant this big fourfold transformation. Foreign aid can be modestly helpful with that, but it is not very central to it. The more you think about donors and what donations should be and what the right interventions are, the more you lose the plot.

In my papers, I show that if you get to what I call national development—this fourfold process—that is a machinery for producing good things. We are worried about whether people have clean water, decent housing, and all of these material things that are really good for people to have. But if you get to national development, you get to that, and vice versa.

There is a big question of whether what we talk about as foreign aid is aid to the process of national development, or whether it is aid to mitigating the worst consequences for human well-being of that lack of national development. The latter is not a strategy for the former. By and large, a lot of the development agencies lost the plot completely and became essentially charity organizations focused on mitigating the worst consequences of the fact that many countries had not acquired national development.

There is nothing wrong with that. Mitigating terrible things that happen to human beings because they live in a country that is underdeveloped is a good thing. But it is not development. Too much attention to the latter detracts attention from the core issue.

Mounk: Is the concern here simply that this money is being spent on things that don’t actually help to solve a problem in the long term—that the most effective use of that money would be to invest in things that actually help solve these underlying problems, and therefore make the country more affluent so there is not as much poverty or human desperation that you need to buffer with these donations? Or is it more profound than that—that those donor dollars in some ways make it harder for that process to take place, that they might in some complicated way backfire? Is the concern mostly about efforts wasted, or efforts that might in some complicated way impede the country that is supposedly being helped from solving its long-term problems?

Pritchett: The big problem is more about ideas than it is about the concrete use of money. I have what I call the “bird on the elephant” theory of development. Development agencies are spending all this money, financing projects—some of them economic projects like roads and bridges and power plants, and some of them charity-like projects funding health interventions and the like. But that is in some sense secondary to the fact that this creates a global discourse about development and about how countries can do development.

I am much more worried about the waste of effort on ideas. We have geniuses—truly stunning geniuses—devoting themselves to charity work as opposed to thinking about development strategy. Ideas are supremely important to the fate of nations, and the ideas that get transmitted via a global discourse of research and practice, to government officials, to people in power, to people who have influence—that is a huge deal. Obviously the most consequential thing that has happened in the last 50 years is the leadership of China changing its mind about what to do in China and how to make China a better place to live. That was, fundamentally, a change of ideas. By losing the plot on national development in favor of mitigation, we also draw the discourse, the research, and the ideas away from big questions: How do we get states to be more capable? What is the right sequencing of state capability and democracy? How does democracy interact with the creation of economic growth—does it impede it or not? The number of people in the world who can produce new, original, and correct ideas is very few, and drawing those people onto small issues is a huge loss.

I don’t think aid is often counterproductive. Some of my friends—Bill Easterly, Angus Deaton—think that aid can sometimes foster mentalities and practices that impede development. I think the elephant is mostly neutral. But if the bird, who sits on top of the elephant and sees what is going on, warns the herd of impending dangers, and can provide a vision—if that gets messed up, then the whole elephant is diminished.

Mounk: Let’s take a step back. One of the intuitive things about what you’re saying is that the United States, the United Kingdom, and France didn’t become rich because some much richer countries said, we’re going to give you a bunch of development aid. These were mostly internal processes, obviously with an international component—trade with each other and so on. So it stands to reason that if we want to think about how countries like India or Kenya might become rich in the future, it would probably be by following some of the same kind of processes.

Now, a certain kind of progressive critic would respond by saying that there is an economic structure in the world and that these countries are in some ways interdependent. If you’re one of the most developed countries in the world, you can specialize in high-return services, for example. If you are much poorer than those countries, then you can’t follow the same development path as those that developed historically, because you occupy a different niche in the economy. The other difference, of course, is just the stage of development. A lot of the countries that grew rich in the 19th and early 20th century did so through industry and manufacturing. But nowadays most wealth is not created in factories—it’s created in the knowledge sector, in the service industries. Perhaps the same kind of path to development just isn’t open anymore. When I was in India, a lot of people were worried about whether the traditional path to development is still available in the 21st century.

What do you think about that? Can you broadly follow the same playbook that made the rich countries of the world rich in the past? Or do we actually need to look for a different path, either because of a change in the nature of the world economy or because of the different relative standing of the poorest countries in the world within it?

Pritchett: Your question brought up two very different strands that we shouldn’t conflate. One is the path being organic and being driven by a country’s own dynamic. If you take, say, Dieter Mikloski’s work, the key to all of this was creating an environment in which you can give it a go. That is distinct from the question of specifically what, economically, a country is going to do. The question of whether the path that Denmark took is open to Kenya has two very different components.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

One is the deeper question: can Kenya develop an organic dynamic within its polity, society, and economy such that the actors find a path for Kenya to become prosperous? That is the true endogenous process. The second question is, when Kenya organizes itself to find that path, will it look like the specific economic path that Denmark followed? The answer to the latter is certainly no—and no in both a good way and a more challenging way.

The first way the answer is no: what happened—and I wrote a paper a long time ago called “Divergence Big Time” that emphasized this—is that the rich countries in 1870 were not that advanced relative to the most lagging countries, because there is a floor. There is only so poor you can be. What has happened in the world is that the rich countries have done this through exponential growth. They have created the basic hockey stick graph.

The rich countries collectively grew at 2% a year for 120 years, and the power of compound exponential growth means they are manifold richer than they were in 1870. But in the process, they invented and discovered a great many things—in science, in practice, and in all kinds of domains—that weren’t available in the world in 1870 but are available in 1970 or 2026. This means the countries that enter this process later have the potential to grow incredibly faster than any rich country grew.

None of the countries that were rich in 1970 or 1980 were rich because they grew fast. They were rich because they grew steadily. But that created the possibility that later countries could accelerate their growth enormously. The fast-growing countries in the world are growing much faster than they did historically. China, Korea, Vietnam—growing at 6% per capita wasn’t an option for Denmark, which just had to stay on 2%. That is the good news. Countries that, in whatever way, manage this endogenous process that creates an organic drive for prosperity can discover ways to grow faster than was available to the old countries that had to make their own way.

We have seen incredibly good news, and we don’t want to lose the plot on that. The years since World War II have been the best years for improvement of the material condition of humankind by a multiple of any prior period. I work some on education. The average person in the developing world had roughly two years of education in 1950 and has eight now. From whoever your mythic forebears were—Adam and Eve, let’s say—to 1950, humanity had accumulated two years of education. In just 60 years, it added six. Three times more education was added in 60 years than in all of human history combined. That is true of health. That is true of a whole range of things. Just fantastic progress.

Take child mortality. There were all kinds of countries in the world 50 years ago where basically one in five children was dying before the age of five—a rate of around 200 per thousand. Worldwide, that is now down to around 30. Many people in the world in 1950 were living in material conditions not that much different from ancient Greece or ancient Egypt. Even today, if you compare Egypt in AD 1—when Joseph and Mary went to visit—to some of the poorest African countries today, they are at roughly the same level of GDP. For a very long period of human history there was very little progress, and then it accelerated dramatically. From all of human history, we got to still one in five children dying around 1960, and now that figure has fallen enormously. If you look at the people with access to electric power, people with access to water and sanitation—it has just been an amazingly good run.

These questions always start from the assumption that the global order is preventing progress, and that is just surreal when you look at the world over the last 60 years. There has been amazing progress in many places on many things, and this isn’t just a matter of measuring GDP.

Mounk: One illustration of that from my own life: when I was an undergraduate in England in the early 2000s, there were big debates about the World Trade Organization. The shape of the debate, and I remember this vividly, was always whether this was going to screw over China and India—whether the World Trade Organization was just a smart way for the rich countries of the world to keep the poor countries from developing. Today, when you look at the critiques of letting China into the WTO, it is exactly the opposite. The question is whether America screwed itself over, or screwed its working class over, by letting China in—which led to a huge increase in wealth in China and rapid deindustrialization in parts of the US.

You can take seriously the idea that there were some mistakes made in how that was done and what impact it had on the working class in Michigan, Pennsylvania, and other places. But if you have to choose, from the standpoint of humanity, between those two scenarios, we have ended up with the much better one. We have lifted hundreds of millions of people out of poverty to genuine middle-class status in China. Perhaps we could have done that a little bit better, with less disruptive consequences for the already pretty affluent people in the United States. Perhaps that’s a red herring. But it is striking to what extent the progressive case against things like the WTO in 2000 has completely flipped on its head in terms of how we talk about it today.

Pritchett: One of my favorite colleagues and friends at the Kennedy School was Dani Rodrik. If you look at Dani Rodrik’s intellectual trajectory, his most recent book is about shared prosperity, which is hugely concerned about the prosperity of the middle class in America and how the global system hasn’t been good for it—versus his earlier concerns about the WTO and whether its trade structure was truly open to facilitating the growth of poor countries. It turns out we had the opposite problem from the one we thought we were going to have.

Absolute poverty in the world has just amazingly declined. Every discussion of the world and how it’s going needs to start from that factual basis. Hans Rosling wrote a book called Factfulness—the fact is, on nearly every measure of material well-being, things are just fantastically better. In many dimensions, by the way, better than we would have expected even from the economic growth that we got. Sometimes people say, you economists focus just on economics, and we really should have focused on these other things. But if you look at the expansion of schooling, for instance, schooling expanded not by less than we would have expected from the economic growth we got, but way more. Just getting kids into school was one of the most phenomenally transformative things in history, and it happened more than we would have expected given the growth, not less. All of development was focused on expanding schooling. So you cannot say that economists focused on growth at the expense of these other things. That is just not true.

Mounk: Let me push you on that—it’s a question I was going to pose in any case. There are a lot of people who are very skeptical of GDP as a metric. They think we can be incredibly rich and yet people are miserable. Society can be very rich in the aggregate, but only because a few people are incredibly rich and everybody else is incredibly poor. There are these standard examples that come up in conversation all the time. If I’m stuck in a traffic jam burning gas and not getting anywhere, that’s increasing GDP—so GDP is a really bad metric.

You believe that GDP is in fact a very good metric of human well-being and that it correlates very strongly with things that we care about more directly. When listeners next encounter somebody who says GDP is a terrible metric we shouldn’t care about, what should they respond?

Pritchett: I want to be clear: I don’t think GDP is a good metric of human well-being. GDP is a very good metric of the net production of a society, and that production creates the material basis for human well-being. GDP per capita is just factually very highly correlated with nearly everything we care about in terms of material well-being.

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

But here is the response you should make. The relationship between GDP and most things we care about in terms of material well-being is concave—meaning those things get better as GDP gets better, starting from poor levels to middle levels. By the time you reach a GDP per capita of roughly $40,000, you have met most of the basics of material well-being, and hence the relationship flattens out. It doesn’t go away necessarily, but it does flatten out.

The real danger in the world discourse is that people unhappy sitting in traffic in Luxembourg—which I use as an example because the only time I went to Luxembourg I got caught in traffic—are right that their overall well-being is not that highly correlated with GDP, because their GDP is so high in the first place. But we shouldn’t extrapolate that back.

Mounk: To put this point a little more polemically: it is very easy—and I grew up in that kind of milieu, not particularly affluent, but among artists and musicians in Germany living in pleasant towns—for people to say that the important thing in life isn’t to drive a big car and go on a fancy holiday, that there are things much more important than material well-being. Of course, if the floor of your society is that you have a decent apartment with heating and running water and you eat three meals a day, then that is probably true. But if you are sitting in a mud hut in a developing country, that is most assuredly untrue.

Pritchett: One of the fundamental insights of economics is declining marginal utility—the more you have of something, the less important it is to you. If you say the West is suffering a paucity of purpose and people are unhappy because they don’t have some driving purpose, well, that is because they have already satisfied the purpose of having material things like a hot shower, a heated home, and the ability to go 300 miles in a few hours. Whereas most of humankind, for most of history and even today, is nowhere near that.

I am happy for German artists to have angst about what their purpose in life is and whether they are really happy and whether more material goods would make them happy—and it probably won’t matter that much to them. But don’t project that back onto Africa, or conclude that India doesn’t need growth, or Bangladesh doesn’t need growth, or that we don’t really need to worry about whether those countries grow.

The right response to the claim that GDP is a bad metric for improving the human condition is to ask: where are you on this spectrum of existing progress? Since you are likely having this conversation with people who are materially fantastically well off relative to any period in human history and relative to most of humanity, they should be sensitive to the fact that yes, that may be true for them, but it is not true for six billion other people on the planet. Don’t project your life and concerns onto theirs, because you have diminishing marginal utility precisely because you have so much.

Mounk: Let’s go to those poorer people in the world. You are pointing out that GDP per capita is very strongly correlated with the other things we care about—child mortality, life expectancy, education, and so on. What does that tell us about what we should aim for? Does it suggest that we should just aim to increase GDP per capita and assume the other things are likely to follow? Or could the association go the other way around—that what we need to do is have all of those specialized interventions to improve the local hospital and the local school, and that is what will then correlate with increases in GDP? Which way around do we read the correlation?

Pritchett: This depends on where you are in the spectrum. I have done a paper where I try to disentangle this question. One of the things that GDP does, by the way, is create a broader tax base. People often get engaged in a very strange discussion as if focusing on GDP growth means ignoring the need for government services. But you cannot have government services unless governments have revenue, and government revenues are mostly tax buoyant—meaning they grow more than proportionately with growth.

Mounk: So GDP growth actually is what enables you to have all those government services?

Pritchett: At one point I did the calculation: Ethiopia’s government spending per head is around $300 per person per year. What can you do with $300 per person per year? Ethiopia’s GDP per capita is so small that if the government started taking bigger proportions of it, they would be eating into expenditures on food. Sixty percent of a poor household’s budget is on food. You cannot just say Ethiopia’s government should fund all these specific interventions—how are they supposed to do it?

When I argue in favor of GDP per capita, it is not because GDP per capita exclusively funds private goods. It is also the basis for funding public goods and public service provision. Getting that mix right is complicated, but economists are not out there saying there should be no government and everything will take care of itself if we just have high GDP per capita. The argument is that high GDP per capita reflects a high productivity economy, and a high productivity economy creates the material basis for both private and public expenditures.

National development is about the process of creating both the material possibilities and the mechanisms for doing that—hence my fourfold definition of development. You could call me a growth fanatic, but I am not a growth-only person. You also need a decent government. In all of the empirical work, you can look at Equatorial Guinea—a few kleptocrats dominating a bunch of oil, with the rest of the economy completely disarticulated from that. Can you get to high GDP per capita and still have low levels of living? Yes, if all of that is being captured by relatively few people.

Mounk: You probably need a natural resource for that as well. It would be very hard to imagine a case where 0.1% of people have an incredibly productive company that doesn’t rely on natural resources and doesn’t rely on broader education and so on. It probably takes a somewhat special case—like kleptocratic control over natural resources.

Pritchett: The way I like to describe the goal—the way we should think about how countries can make progress—is what I call “inclusion into productivity.” The reason human beings are this amazing species is that we have learned how to cooperate to create value. A large part of the development process is getting more and more sophisticated ways in which people can cooperate over time and space to create value. When I look at a great big corporation with tens of thousands of employees, this is a mechanism of cooperation to create value—a cooperation across people with all kinds of different skills and contributions.

Subscribe now

The process of development is getting more and more people out of being engaged in activities where they work a plot of land and attempt to eke out a living in low cooperation, and into more and more sophisticated value chains. That is inclusion into productivity. What I am actually concerned about is the productivity of individuals—but they are going to be more productive not by being more separated from a sophisticated modern economy, but by being more embedded into it. That dynamic, which leads to indicators of inclusive growth, is how I think about the fundamental dynamic on the growth side. Growth should be a process of more and more of the population engaged in higher and higher productivity, with more and more people included into these mechanisms.

Mounk: To go back to those two different challenges I posed earlier: one is whether poor countries can still grow at the same speed that others did, and your answer is yes—in fact, a lot of them are growing more quickly. There is a strange thing in the premise of that argument, which somehow implies that we haven’t seen examples of poor countries developing very quickly. That is partially because people who think about this tend to drop successful countries out of their sample. They don’t remember that China in 1980 was an incredibly poor country. Today we no longer think about China as being part of that sample because it has grown out of it. That in itself is evidence that rapid development is possible. If China was able to do that, there shouldn’t be an in-principle reason why Kenya or India couldn’t achieve the same feat.

The other question is what that looks like in the 21st century, when some of the historical development path is no longer available. My understanding is that your answer is going to center on those four development factors. Why are those so hard to implement in places where they are not in place? Ideas are really important, as you said earlier. If only the right people adopted the right ideas, they should be able to put those four things into place. But clearly it is more complicated than that. It is not just that the rulers of these countries have never had the right ideas. There must be obstacles beyond that.

Pritchett: One of the obstacles is that the process of growth of inclusive productivity is a transformational process. A transformational process requires winners and relative losers. In many places, economic, political, and bureaucratic structures congeal. They congeal because there is a certain way in which a country produces value, and that way of producing value often produces certain ways in which the government extracts value. That can lock into an equilibrium that resists change rather than encourages it.

There are low-level equilibrium traps—ways in which countries get stuck. We already alluded to one of those. In resource economies, if you are relatively well endowed with certain point-source resources—and by point-source I mean not land, which is diffuse and requires a geographically distributed population, but oil or diamonds—you can see how the people who mine the diamonds and the people who control the country get embedded in a relationship where, as long as that elite bargain can survive off the extraction of value from diamonds, it is not looking for anything else. It is not looking to solve the problem we are talking about, which is how a country comes to be embedded in more and more sophisticated value chains. If an elite can generate a bargain that sustains itself politically over time with just diamonds, it is not looking to do anything else, and that becomes an obstacle.

Economies and countries get stuck in an elite bargain that is more worried about the threats from new industries creating new power structures than about stagnation around the parts of the productive structure they are already in. You can easily get stuck in the dynamic between economy, polity, and bureaucracy in which an elite bargain undermines the rule of law and undermines the expansion of new opportunities, in the interest of playing defense around where they are. Finding countries that can handle this dynamic of change is very hard. That is why we see so few transformational successes, even though the global order makes them possible.

Mounk: Is there some way to change that, to contribute to that? If economists at the moment are focused far too much on whether giving people in this village one kind of intervention or people in that village another kind of intervention will lead to a little bit more growth—and you think that is fundamentally the wrong question to ask—do we have better answers to how we can help a country like Nigeria, with a very fractious political system and an elite bargain that is very bad for the country, get out of that trap?

Pritchett: Many of your questions have two sophisticated halves to them. Let me dwell on the first half—the premise that people are wasting their time. I want to emphasize that they are not just focused on the wrong question. They are focused on ontologically the wrong question.

The word ontology is not one I like to use, but what it means here, at its fundamental roots, is this: the problem with the world is not poor people. The problem is that people are in poor places. If you are studying the dynamics of how to make people less poor, you have to be ontologically studying the characteristics of the system, not the people. The methods being deployed by economists to study how to make people better off are focused as if the person were the unit at which we should be studying this. That is ontologically wrong.

Mounk: Let’s stay with this half of the question before moving on to the other half. Take a step back and explain to us what the dominant paradigm in development economics has been for the last 20 years and how that dominant paradigm is particularly vulnerable to the critique you’re making. I take it you’re mostly talking about what are called RCTs. What is an RCT? How did it become so dominant? Why do you think that’s the wrong way to ask this question?

Pritchett: If you characterize what development actors are doing as carrying out interventions, then you can get obsessed with understanding whether the particular intervention you’re undertaking has a truly causal effect. We can’t just evaluate a project on a before-and-after basis—we need to really separate out the causal impact. For separating out the causal impact, doing a randomized controlled trial is the best way to do that.

The problem is, if I’m trying to do a randomized controlled trial, I need some group of individuals to receive the treatment and some individuals to be the control group. That means I have already ontologically assumed that the important causal effects are individuated, as opposed to being country-wide or market-wide phenomena. This is as plain as I can make it.

Mounk: Let me try to put this in plain terms to see whether I understand correctly. The idea is: I study Village A and Village B, I give Village A deworming medicine and Village B nothing, or perhaps some other kind of treatment. If what I’m trying to figure out is how to spend a hundred million dollars for a charity, that is a very reasonable question to ask. The problem you are pointing to is that what actually explains why both Village A and Village B are very poor—versus Village A and Village B in England being relatively affluent—is these country-level characteristics: in particular, whether they have managed to figure out the rule of law and all the other things we have been talking about. You are defining the interesting stuff out of the question if you are just comparing two villages within the same country that are meant to be as similar as possible, and then seeing whether this or that kind of treatment is going to marginally improve the lives of people in each of them.

Pritchett: Let me take what I consider a paradigm example—one that, when you encounter it, should make you realize something has gone badly wrong. There was a paper in Nature—the most highly reputable scientific journal in the world—about carrying out an experiment in Niger where a cash transfer was given to some people and not to others. Bundled with that cash transfer was a psychosocial intervention, and the study then looked at whether this psychosocial intervention independently and causally caused individuals in Niger to see their incomes rise.

This was Nature magazine’s characterization of what development economics was doing, with 11 prominent development economists as co-authors. When you see that, you should immediately think: this is madness. People in Niger are poor because they are in Niger. Niger, on every indicator of national development, is a basket case. If you are not fixing Niger, thinking that you are doing good in the world by making tiny tweaks to psychosocial interventions at the individual level, you are assuming that a large part of the low standard of living there must be a characteristic of the people in Niger. That is just wrong—wrong by orders of magnitude. People in Niger are poor because they are in Niger. If you allow a person in Niger to move to France, their income will converge to that of people in France almost instantaneously, because France is a high productivity place and Niger is a low productivity place.

If Niger as a country had national development at the level of France, all of these problems would be addressed, because national development is a machinery for endogenously identifying and solving problems. If you don’t have that, attempting to solve these problems at the individuated level is wildly, ontologically wrong. You are not taking into account that the outcomes individuals have are determined not by their own characteristics, but by the characteristics of the political, organizational, economic, and social system they live in.

Mounk: The other thing that strikes me about this study is the theory of action lurking in the background. I have no doubt that people who encounter psychosocial problems are less productive, and I am moderately optimistic that the right kind of psychosocial intervention might reduce those problems and therefore make people more productive. But how on earth are you going to deliver psychosocial intervention at scale in one of the poorest countries in the world? It is not going to happen because an army of therapists is flown in from Brooklyn, New York to treat everyone.

How do you actually reduce the amount of psychosocial problems that people have? How do you increase access to therapy and other support for people who do have serious psychosocial problems? By making the country a lot more affluent. If the country is a lot more affluent, perhaps your child doesn’t die at age three and you have fewer psychosocial problems to begin with. You are less depressed about the fact that your child just died. You will have the money to afford a therapist, or perhaps the government will have the money to fund a mental health service. You are asking the question from the wrong end.

Pritchett: Then you get obsessed with cost effectiveness, which means reducing both the numerator and the denominator. In this study, per capita income went from $1.80 a day to $1.85 a day. If you look at Vietnam, in 1990 it had a per capita income of about $1.80 a day—similar to where Niger is now—and it has since gone up to around $7.50 a day. That is development. None of that in Vietnam came from people becoming psychosocially more capable first, with everything else following from that.

Mounk: It’s not that USAID or the Ford Foundation sent an army of psychologists to Vietnam.

Pritchett: It was because the development actor and discourse helped Vietnam change its mind about its development strategy. Vietnam said: we can embody Vietnamese labor into global value chains in a way that is going to radically change the productive possibilities and the inclusion into productivity of Vietnamese people. They did it successfully, and extreme poverty disappeared in Vietnam right before our eyes through a process that involved almost no direct anti-poverty interventions.

So that is the first point: this isn’t just wrong, it is ontologically wrong. Ontologically wrong means you cannot fix it by doing slightly better experiments on this or that. It is looking at the wrong set of ways in which prosperity actually happens. Prosperity happens through the inclusion of individuals into productivity—a cooperative endeavor involving artifacts and institutions whereby large numbers of people can cooperate to create value: markets, governments, bureaucracies, and large organizations. Not whether I am a more go-get-em person. That was the first part.

Mounk: The second part then is: if RCTs are the wrong thing to look at, if that is not how we are going to make change, if we shouldn’t be obsessed with whether it is this intervention or that intervention at the level of the village or the town—but rather should be asking how we help countries choose successful development paths like Vietnam, paths that put in place the institutions needed to actually develop—how can development economists or other social scientists help with that?

Is the problem fundamentally that the people in Nigeria just don’t get it? Or is it, as you were saying earlier, that the people in charge have their own interests and want to preserve the privileges they have? They are worried that in the process of development they might be displaced from power, and so they would rather continue to be the kleptocratic elite in a relatively poor country than be displaced in a much more affluent one. Is development economics, or more broadly the field of social science, actually going to be able to help in that process? Or is that a hopeless endeavor?

Pritchett: It is not a hopeless endeavor, because after all, it has happened. The strange thing about saying it is a hopeless endeavor is that it has to ignore that Korea happened, and Taiwan happened, and Vietnam happened, and China happened, and Indonesia happened, and India happened.

Mounk: But did it happen because those places, for whatever reasons, had more favorable circumstances where they were more able to challenge their elites—or did we just get lucky that they put a few of the right reforms in place? Or did it happen in some way because social scientists helped guide them?

Pritchett: Yes, social scientists helped guide them, because social scientists created the data—like GDP per capita—to show that some countries have had really rapid growth by doing certain things and other countries have not. Deng Xiaoping didn’t just decide on moving towards a path of unleashing incentives for people to engage in private enterprise—perhaps too strong a word for what happened in China in 1978—in an intellectual vacuum. He looked around the world, brought in experts who had studied economics and the relative performance of economies, and was convinced that there was an alternative path from where China was to where China wanted to be.

The idea that ideas are irrelevant and that social science hasn’t had an impact is surreal. In the case of India, which I know relatively well—I have been going back and forth to India since 1992 and have spent time living in different cities there—there was a debate in India in 1991 about how to respond to an incipient macroeconomic crisis. That debate drew on decades of social science about the relative importance of market versus government-led development, and of import substitution versus export orientation. It drew on a body of empirical science that had been generated over time, and it led the country to do different things. You cannot act as if India decided what it did in the reform period in an intellectual vacuum. It is surreal to say that social science cannot in principle affect the way governments act, because it has demonstrably done so in specific instances. Maybe it has not always been as successful as one would hope, and let’s admit that oftentimes social science is pushing against direct material and political interests.

But let me answer your question. The answer is what I call full Trinity growth diagnostics. If development economics were focusing its time, effort, and capabilities on the development of full Trinity growth diagnostics, I think we would be in a radically better place to help countries with their organic, country-level development strategies.

What do I mean by full Trinity? For something to be successful, it has to be technically correct—it has to have a correct causal model of how, if actors undertake this action, working perhaps through a complex adaptive system, a particular outcome will result. It is not direct cause and effect, because markets are complicated and complex adaptive systems. But the technical question is: if I do this, will this be the result?

The second element of the Trinity is organizational. If I am asking a government to undertake actions to promote a given outcome, does the government have organizations capable of doing that? If I say I am going to do industrial policy to promote high-tech industry, do I have government organizations capable of identifying and promoting high-tech industry?

The third element is that it has to be politically supportable. I have to be able to assemble and sustain a political coalition to support the implementation of these actions.

The first step of this is a growth diagnostic—a process that I and colleagues at the Kennedy School developed over time—of asking, in a given place, what are the binding constraints to having more rapid growth than we are currently experiencing. The binding constraints matter because when we ask what is good for growth, we tend to produce a very long list. If you say “get to Denmark,” you can say Denmark has this and Denmark has that. But you cannot do everything that Denmark now does. You have to do the right thing. Identifying the right thing involves an analytical process of prioritizing among the many things you could do—asking which would have the biggest impact on the prospects for a rapid, sustained, and inclusive episode of growth. We have mechanisms for trying to adjudicate among contesting claims about what the binding constraint is.

But then we have to supplement that by asking: of the things that should be priorities and would have a big impact on economic growth, which can we actually do? Saying that if you had Korea’s current capabilities you could do a certain thing is not very helpful if you have Niger’s capabilities. We have to ask which of the binding constraints can realistically be addressed. Then you have to have a realistic, positive model of the politics—which of these can the existing elite be persuaded to engage in? None of those questions poses an intellectual challenge anything like reconciling general relativity with quantum mechanics.

Mounk: But it’s an art rather than a science, I imagine. What might that look like concretely? Let’s say there is a new government in some country—genuinely reforming in certain ways, perhaps a little corrupt in others. They have just come to power, perhaps through election, perhaps in some other way. They can make some changes, but they cannot make a million changes at the same time. It is hard to drive change across sectors simultaneously, and there is a limited amount of money to invest because it is a very poor country. Should they invest in education or in the judiciary? Should the first big push be to reform the judiciary—so that who wins a dispute is driven more by the facts and creates stable expectations, rather than by bribes or family relations? Or should they start with a big push on schools, making sure teachers are competent and actually show up? I imagine that is the kind of choice a government might realistically face. How should you go about making that decision?

Pritchett: There are analytical tools for addressing this, and this is where we get into the mechanics of what a growth diagnostic would do. Partly what a growth diagnostic does is ask: if it were the case that a proposed action were the binding constraint to growth, what should we observe about the economy?

There are four or five things we look for. First, if we think X is a binding constraint—say, corruption—then when we see relaxations or improvements in corruption, we should see more growth. If I have my arm in a sling and the movement of my arm is constrained by that sling, then if I remove the sling, I should be able to move it a lot more. If I can’t, maybe it was the shoulder injury all along and there is a different cause. Changes in the constraint should cause changes in outcomes.

The second is what we call bypassing. If finance is a constraint and I could otherwise be productive and profitable, we should see firms actively engaged in creative ways of raising finance. We should see enterprises actively adapting around the constraint.

The third involves what we have given the quirky name of camels and hippos. Different industries in an economy are more or less intensive in a particular proposed constraint. If we say water is a constraint, then we should observe camels—animals well adapted to a lack of water—and we should not observe hippos. If we look at the economy and the industries that are thriving are ones that economize on the proposed binding constraint, that is evidence in favor of its being the binding constraint.

These are plausible, sophisticated ways of analyzing the current situation of a country. They can produce answers like: putting more kids through school really isn’t a binding constraint here, because we don’t see changes in that producing changes in growth, we don’t see firms desperately engaged in training because they are short on skills, and we don’t see labor-intensive, non-skill-intensive industries thriving because they are economizing on the lack of skill. That means we need other explanations.

This technique has been deployed and it produces interesting results—not clean RCT-looking results, nor should one expect that. To me, the key test of a growth diagnostic is whether it comes to different answers for different countries. If you come to the same answer for every country, you have an ideology. If every country you go to produces the answer that reforming the trade system would accelerate growth, you are an ideologist. Maybe you are right sometimes, but you are not always going to be right, because there are going to be societies in which that is not the most important binding constraint. When we do these growth diagnostics in different countries, we come to demonstrably different answers. The data and the evidence line up around different things.

This is a promising technique, but it has been radically under-invested in—because the resources in the world have been devoted to deworming studies rather than to creating a global community of practice that does technical growth diagnostics. More support to a global community of practice focused on honing the art and science of growth diagnostics would be a tremendous way the world could support the organic process of countries finding their own development paths. It would allow you to engage with countries and actors within countries in a radically different way than coming to them and saying: you lack these 50 things, all of which would be good if you did them—when you know they are going to be able to do two, maybe three.

In the rest of this conversation, Yascha and Lant discuss migration, how to build a fair asylum system, and what the destruction of USAID means for the future of development. This part of the conversation is reserved for paying subscribers…

Read more

Predicting Donald Trump’s political demise has typically been a fool’s errand. Some of my smartest friends have declared his definitive fall from grace again and again, only to be proven wrong each and every time.

If you watch MSNBC or listen to NPR, you may over the past decade have believed that Trump’s presidential campaign is a hopeless publicity stunt; that the Republican Party is about to turn on him because of the Access Hollywood tape; that he has no chance of winning against Hillary Clinton; that his presidency will be so chaotic that he’ll be forced to resign within his first year in office; that Robert Mueller’s investigation into his relationship with the Kremlin will result in his impeachment; that his mishandling of the COVID pandemic will make him toxic to voters; that his loss against Joe Biden has ended his career for good; that he is about to be impeached over the January 6 riot at the Capitol; that he is sure to lose the race for the Republican nomination against Ron DeSantis; that he is sure to lose his bid for reelection against Joe Biden; that he is sure to lose it against Kamala Harris; and so on.

Ten years into Trump’s political career, the most avoidable mistake pundits can make is to underestimate his powers of survival and resurrection.

And yet, I have come to the tentative conclusion that this time may, finally, be different. For the past decade, Trump has dominated American politics like no other president in living memory; now, signs of that era coming to a close are suddenly multiplying. It is, as Saturday’s appalling assassination attempt on the president reminds us, impossible to see around the next historical corner. But it sure seems as though Trump’s hold over the country is finally slipping. This, to misquote Winston Churchill, no longer feels like the end of the beginning; it may be the beginning of the end.

A minority of Americans has always been drawn to Trump because of his most extreme actions and statements. They loved his coarseness, reveled in his taunts, and unhesitatingly embraced his radicalism. This group made up a significant share of his most devoted base—but it was never big enough to explain how he could have won two presidential elections.

Many of the voters who twice put Trump over the top have, all along, had a more conflicted view of him. Trump swore that he would make Americans far richer. He would cut taxes and curb inflation. The costs of health insurance would fall. There would be peace in the Middle East. The country would return to its former grandeur. It is not hard to see why those who were inclined to believe that he might actually turn these promises into reality, at least to some extent, found them to be very enticing.

Want to hear me in conversation with leading thinkers of our time? Sign up for ad-free access to full episodes of The Good Fight by becoming a paying subscriber today!

To listen on your favorite app, click “Set up podcast” below. Once your personal link is properly installed, you should see “PREMIUM FEED” in the top-left corner of the podcast icon. If you are having trouble setting this up, please email support@substack.com.

Set up podcast

During his first term, Trump did celebrate some genuine successes, from Operation Warp Speed to the Abraham Accords. But when he predictably failed to bring about most of his outsized promises, he proved shrewd at making up excuses. He had only just taken power. The deep state was standing in his way. The “Russia hoax” had made it impossible for him to govern. The global pandemic had messed everything up. The share of Americans who were genuinely excited about Trump shrank rapidly towards the end of his first term; and yet, the thought that it might be worth giving him a second chance in 2024—even if he just delivered on some tiny fraction of his promises—lingered in the minds of a surprising number of voters.

But the fulfillment of promises can’t be deferred forever without voters starting to lose patience. As Viktor Orbán learned to his chagrin in Hungary, there comes a time when leaders are measured by their results rather than their rhetoric. And that time has now come for Donald Trump.

The immediate reasons for Trump’s travails lie in his ill-fated war with Iran. The contention that foolish “foreign entanglements” had repeatedly led America astray was central to his political persona from day one. In his second inaugural address, he announced that “we will measure our success not only by the battles we win, but also by the wars that we end, and perhaps most importantly, the wars we never get into.” This makes it especially damaging that he pursued a war of choice in Iran without bothering to make a coherent case for it to the American public or ensuring that there would be a real exit strategy. The one major promise that Trump actually honored in his first term was that he would start no new major wars; that too now looks like empty self-promotion.

Subscribe now

The knock-on effects for Trump’s other areas of traditional strength have been brutal. Americans voted out Joe Biden’s Democrats in good part because of the persistently high level of inflation after the pandemic, which had been fueled by the administration’s generous stimulus programs. Now, Trump’s failure to anticipate that Iran would choose to block oil tanker traffic through the Strait of Hormuz has led to a renewed spike in inflation, putting the president’s approval ratings on inflation and the cost of living underwater by a remarkable 40 points.

Trump is also in trouble in some historic areas of strength that are less directly connected to the war in the Middle East. Most Americans grew furious with Biden’s inability to control immigration at the southern border. But in his second term, Trump has embraced a deportation policy that is so pointlessly cruel that, in many polls, a clear majority of Americans now disapproves of his handling of the issue.

The result is becoming increasingly clear in the data: Overall support for Trump is at or near record lows.

Trump has often been far more popular with the American public than his detractors cared to acknowledge; today, his approval ratings are genuinely dismal. Nearly 58 percent of Americans disapprove of Trump’s job performance (most of them strongly) while only 39 percent approve (most of them weakly), according to statistician Nate Silver’s polling tracker. His net approval is as low today as it was in the immediate aftermath of the January 6 assault.

In the past, Trump has been hated by liberals, seen as divisive among independents, and (the complaints of a small band of principled columnists notwithstanding) enjoyed popular support among conservatives. Declines in Trump’s poll numbers were usually precipitated by independents abandoning him. Today, Trump remains toxic among liberals, has come to be viewed negatively by most independents, and is newly divisive among conservatives.

Trump’s ironclad grip over the Republican base is starting to loosen. In the past, conservative critics of Trump have usually complained that he has sold out the views and values associated with figures such as Ronald Reagan. Now, criticism of Trump within the conservative camp is for the first time being framed as a betrayal of the supposed values on which the MAGA movement was founded. Some of the biggest influencers on the American right, such as Candace Owens and Tucker Carlson, have recently expressed regret for supporting Trump. For the first time since 2016, his hold over the MAGA movement may actually be weakening.

The political costs from these developments are likely to compound over the course of the coming months. Betting markets give Democrats about a 6-to-1 edge to win the House of Representatives in the midterm elections this November; despite a daunting electoral map, they also have slightly more than even odds to take control of the Senate. If Trump’s party really does suffer a serious shellacking in the midterms, his inability to push major legislation through Congress and the impending end of his term will further weaken his control over his own party. With attention turning toward the 2028 primaries, the White House may suddenly see its power slipping away, as happened after the 2006 midterm elections, in which Democrats took control of both houses of Congress, rendering George W. Bush largely powerless during the last years of his term. Sooner than we can now imagine, Trump may come to be seen as a lame duck.

When Trump was reelected with a bigger, younger, and more diverse electorate in 2024, it seemed as though he might actually manage to impose his vision and his values on the country. In the first months of his second term, the administration was moving with impressive speed. Resistance to its ascendancy was conspicuous by its absence. It felt as though America might stand at a genuine tipping point.

The window of opportunity for Trump to reshape the country in a significant way was, I think, real. But he responded to the cultural excesses of the Democratic Party—and the broader progressive establishment with which it is increasingly associated in the public mind—with even more extreme cultural excesses of his own, provoking a broad counter-reaction which extended well beyond those who partook in resistance marches during his first term. Therefore, it now seems increasingly safe to say that he has squandered it. Trump’s second term will leave behind an America that is weakened, cheapened, and fractious; but it seems increasingly unlikely that he will leave behind an America shaped in his own image.

Subscribe now

This is cause for optimism, an indication that America has proven to be more resistant to the appeal of authoritarian populism than many feared. It would take someone who is much more popular and disciplined than Trump to change the country in a fundamental way.

And yet it is far too early to celebrate. Trump will, after all, remain in office for another 32 months. That is enough time to do a lot of damage to democratic institutions, to engage in a great deal of corruption, and perhaps to start more reckless wars. In all likelihood, a President Trump who is starting to sense that the tide is turning against him will turn out to be more, not less, dangerous to the American republic—and the world.

Some danger will persist even after he leaves office. When demagogues leave office—even when they are booted from office in disgrace—it rarely spells the end of their movement. Brazil’s Jair Bolsonaro lost his bid for reelection and was imprisoned for trying to impede the peaceful transfer of power, and yet his son Flávio has close to even odds of becoming the next president of Brazil, according to prediction markets. Alberto Fujimori was hounded out of Peruvian politics due to massive corruption and human rights abuses nearly three decades ago, and yet his daughter may be about to lead the country.

In Brazil, Peru, and many other democracies around the world, voters may decide to give populist movements a second (or third or fourth) chance because they were so disillusioned with the hapless alternatives. Given that the popularity of the Democratic Party remains at record lows, it would be deeply naive to rule out a similar future for the United States.

Trump looks likely to start fading from American politics over the coming years. But the broader threat of Trumpism may well outlast its creator.

This piece was originally published by the Dispatch.

David Bromwich has taught literature at Yale University since 1988. His books include Hazlitt: the Mind of a Critic, The Intellectual Life of Edmund Burke, How Words Make Things Happen, and Politics by Other Means: Higher Education and Group Thinking.

In this week’s conversation, Yascha Mounk and David Bromwich discuss why Americans have lost faith in universities, how grade inflation and political conformity undermine academic credibility, and whether the opacity of elite admissions processes can be reformed.

Note: David Bromwich asked us to be clear that the views he expresses are his own and not those of any institution or group within an institution.

This transcript has been condensed and lightly edited for clarity.

Yascha Mounk: I’ve been hoping to have you on the podcast for a long time, and we have a good occasion now because you were on the faculty committee at Yale University that was tasked with trying to figure out why people have lost faith in higher education in significant numbers in the United States, and what universities—and particularly Yale—can do to regain the trust of the public.

What do the findings of the committee suggest lies at the heart of this loss of faith in universities in the United States?

David Bromwich: Several things. You use the word “faith.” The official name of the committee was Trust in Higher Education. But trust, as anyone who has studied moral philosophy or just speaks English knows, is mutual. It depends on a shared understanding of what the purpose is of this or that institution, this or that custom or ritual.

The understanding that was lost is how public higher education—and specifically liberal arts education—prepares you for life in a way that will serve students well in getting jobs, but also make them thoughtful citizens. There is some effect from the education that they wouldn’t get just from reading a lot of books or even watching a lot of television.

Some of the causes of waning trust that we looked at were the process of admissions. Yale is an elite school, so this is particularly dealing with the loss of trust in schools that have that sort of prestige. The price of the school is very high, although most people don’t pay the sticker price. That relates to a lack of transparency in things like how you pay, what tuition costs are, how it may be deferred, what loans are available, and how much tuition can be covered for people who aren’t rich.

Recently—and this overlaps with the work of the committee—an announcement came out about a month or two before our report was put into its final draft. The president of Yale, Maurie McInnis, announced that families making $200,000 a year or less, per household, would not have to pay any tuition. If you know how low on the scale of disposable income $200,000 for a whole household actually is, that’s not surprising. It’s not extraordinarily generous, but it is more generous than most people imagine the Ivy League schools are.

Mounk: For probably most of the households that send kids to Yale, that threshold is above their income. The median household income in the United States is something like $70,000 a year, so for the median American family that is able to get their child into Yale, they are going to have a free ride.

Bromwich: There is what we should call opacity rather than transparency. I like the 18th century word for it: publicity—making public, in a neutral sense, the criteria that are used by a university like Yale to accept or not accept students. This is all the more important because we’ve acquired an excess of bogus prestige from rejecting a lot of students. There are more and more applicants, more and more people think they might make it, but admissions are now lower than 5% at Yale and at places like Yale.

So how to account for the distrust? There seems to be a kind of false advertising in which the institution has been unconsciously indulging.

Mounk: The report is also quite explicit about the problem of grade inflation. In Yale, the median grade is now an A or an A minus, I believe. How is that contributing to the loss of trust in universities?

Bromwich: A student reporter writing for the campus newspaper, the Yale Daily News, found in an article published about a year and a half ago that the average grade given at Yale was in the A family—meaning As or A minuses—something like 70% of grades. As a teacher in the humanities (and we are usually charged with being the great culprits on grade inflation) I was shocked to discover that even in the natural sciences, grade inflation of this sort prevails. In order to distinguish among students—non-invidiously, but to discriminate—some much greater separation of degrees of distinction would seem desired.

That is one of the suggestions made in the final section of our report, which is called “Recommendations.” There are 20 recommendations, and one involves making it easier to calibrate how well students are doing against the cohort of people in that class. If a class is 80% A’s, your A in that class is going to count less than an A in a class where there are 30% A’s. That can seem like a small recommendation, but it may mean that students demand classes where distinction can show up. If they demand classes like that, they may gravitate more towards courses that have some real rigor and be a little less shameless about attending big lecture semi-gut courses in the social sciences or even the humanities that assure them of a good grade, thereby mingling with the respectable crowd of Ivy graduates.

Mounk: One topic that has been broadly discussed in the media, but also felt by many professors—including left-leaning professors at universities—is a sense of political conformity and a fear that one may, in some explicit or implicit way, be punished for expressing unpopular views. The report mentions that the number of students and faculty members at Yale who fear sharing a political view is quite substantial. How has the problem of political conformity contributed to a loss of trust in universities?

Bromwich: It is a reductive way of framing this, but not without its own revelations, that 90% or more of faculty at universities like this tend to be registered Democrats and to have, let’s just say, left-liberal politics. Why is that? That’s a long story. I have my own speculative explanations. But there’s no doubt about the enormous, sharp divide in the political culture of the United States right now. It’s a 51-49 politics, counting it simply as Democrat versus Republican, and it keeps going back and forth. So it’s about even, and what these two parties stand for is not quite clear very often, even to themselves. To have it all leaning with great imbalance on one side doesn’t seem an adequate representation, or an adequate preparation for society, among the students who go to college.

That’s not to say that there’s a great deal of political indoctrination that is just part of the ethic of campus life. I don’t think that’s true at Yale. I think it’s been overplayed as a problem about universities in general. But there’s no doubt that the actual disproportion of political tendencies among faculty members was a factor in creating more distrust.

We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!

Subscribe now

Set Up Podcast

If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community

Then there’s the question of free inquiry and free speech—related to politics, but related to cultural and social issues too. Are there questions that could be important moral topics for discussion in a university that are pretty much off limits, or where it’s understood—with nobody having to say so—that there’s one right answer, one right response, one right side to take, so that students are, imperceptibly but nevertheless, discouraged from getting into animated discussions about these things in class or outside of class?

Mounk: Part of what makes students afraid to speak up in class is the threat of social media. If a student says something that a classmate finds offensive, that classmate may go on TikTok or some other social media platform, perhaps misrepresent what was said, and call it out—leaving that student with no friends, with nowhere to sit in the dining hall, and so on.

Broomwich: Students are wary of saying something controversial because it might be reported in such a way as to harm their reputations early on, and how that affects people’s willingness to speak is pretty easy to guess. One of the recommendations—and it is one I’m most proud of on behalf of this committee—is a no-gizmo classroom: no laptops, no iPhones, no recording. It was suggested that we propose Chatham House rules for all classes, which means you can report things that were said, but you must never report who said it. I don’t think we went that far, and I don’t think it’s necessary to go that far. But you want to stop well short of permissiveness towards creating gossip around comments made in class, either by a teacher or a student.

Mounk: I want to double-click on a few of these because there are a lot of interesting things in there. I’m struck by your first observation that there is a lot of opacity. I don’t think that opacity is by design exactly—in some areas perhaps more than others—but it has been created over time.

It’s true that a lot of the things that universities do, that supposedly are meant to serve worthy goals like equity, in fact have the result of favoring the people who know how to play the game. Personal statements are meant to give admissions officers a richer view of a personality and allow students to share when they’ve dealt with some kind of genuine hardship. In reality, it is often the most privileged students, from the most privileged backgrounds, who have the cultural knowledge to understand what you do and don’t say in a personal statement, and who have had the money to go volunteering on some wonderful project somewhere.

With the financial aid system, there are some really strong reasons why universities have embraced the models they have. The logic is that students who come from very rich households pay a lot, and students who don’t come from rich households are not charged any tuition, or are perhaps even given full living expenses. The top universities with big endowments are now very generous in that regard. But the result is that if you come from a genuinely underprivileged background, you may not know that. You may only have heard in general conversation that going to Yale now costs about $100,000 a year all in. So you may not even apply.

Not to mention that there are surely some parents who are smart about how they plan their finances and income—if you’re self-employed, you might frontload some losses to your business in the year your kid goes off to one of these schools, and because you really know the system, you’re able to come in under $200,000 that year, even if in the years before and after your kid attends college you’re way over $200,000 a year.

These are all examples in which a system—often designed quite earnestly, for example to make college accessible, which it now factually is for most people who are not very affluent—remains so opaque that people may not know it. If your kid gets into Yale, Harvard, Johns Hopkins, and so on, you do actually get tremendous aid. But the system is so opaque that people may not know it. How did that come about, and is it fixable?

Bromwich: I don’t know how it came about. As the scholarship process became more intricate, with more kids from public schools being admitted to elite schools—the Ivies, Chicago, and so on—I suppose they found it difficult to explain the intricacies. There is also the gradual advent of what came to be called the holistic approach. This is what Harvard got hit very hard for in the fairness in admissions case before the Supreme Court, because the holistic approach could be shown to disfavor Asian students on personality criteria that are very obscure. They kept receiving low grades on those criteria to compensate for their high SATs, so that the university wouldn’t have to admit too many.

Mounk: To expand on this for a moment, there is a dark history here. Those personality scores were explicitly introduced to keep down the number of Jews at these schools some 50 years ago, and now they are being used to keep down the number of Asian students. The figures were just remarkable. I believe it was on a five-point scale, and the personality of the average Asian applicant was determined to be more than a standard deviation lower than that of any other racial group. Basically, Harvard University’s institutional judgment is that, on average, Asian students have remarkably terrible personalities—and all of this is just part of this opaque system.

Bromwich: It’s all a cultural and racist cliché, but it fits all the old models. Black people are just more interesting, more animated, more lively than everyone else—they have personality. Asian people, on the other hand, are harder to read; they all seem to have, somehow, not enough personality. White people are somewhere in between. It was absurd what came out of that.

All these schools have been practicing some version of holistic admissions. That system depends on the spontaneous judgment and tact of the people working in admissions when reading an application. They should presumably be people who have been taught some basic rules: what are the good schools, what are the hard schools, what are the districts where, if a student has really done something exceptional, it means a lot—and so on. Nevertheless, that system, because it is so personal and so subjective, is liable to particular abuses. No doubt about it.

Auf deutsch lesen 🇩🇪

Lire en français 🇫🇷

A well-meaning democratic desire to present a welcoming face to students of all kinds is also largely responsible for some of the obscurity of the process. One recommendation from our report is that schools not advertise sticker price alone, but actually go into some of the intricacies, showing just what kind of chance an applicant has. But a subjective element remains. There will be students who feel prompted to try their chance at Yale, Harvard, Princeton, or elsewhere, but who, on objective measures such as the SAT, aren’t quite up there with the best students—students who do come from more privileged backgrounds but are more academically prepared for a difficult university.

Mounk: One of the interesting things that the report suggests is that over time the mission of the university has broadened explicitly, and with it the criteria for the kind of students you want to admit. The report suggests that the university should refocus its mission much more narrowly on the creation, preservation, and transmission of knowledge, rather than those vague goals of educating future leaders and all kinds of other things. Going with that, it suggests that academic merit should be the core criterion for admissions, in a way that evidently it is not always now.

How far should we go with that? Why shouldn’t we do what other top universities in most other countries do? The report acknowledges that in most other countries, universities like Yale and Harvard—as we have recognized in the last year during the Trump administration’s attacks on these universities—are private universities with large endowments, but they do actually receive a lot of public funding. One obvious way to preserve trust in them is to have a very transparent, clear academic metric for who should get in, whether that is a national entrance examination or a university-specific entrance exam—something where people are graded without knowledge of their identities. The people who come out on top are admitted. That is broadly speaking the system that most other democratic countries in the world use. Why not go the whole way and get rid of holistic admissions altogether?

Bromwich: I am probably closer to that view than some of my colleagues on this committee, but we went a distance towards it. As you mentioned, what these mission statements import is a hard thing to say. They started becoming inspirational—part of the brochures sent out to prospective applicants—about 30 years ago or so. For example, all of these elite universities will speak of their virtues as being a “second home.” The word “home” is used, I would say, even half as often as the word “community”—as if the university is a whole separate community. That leads to some fallacies about the sort of concord or comity that ought to exist among everybody, fallacies that are very hard to erase but should be erased. If you want what John Stuart Mill called the clash of ideas to happen in a university, that clash means friction, some abrasiveness—not necessarily wounded feelings, but surprise, startlement, shock, the feeling of being rubbed the wrong way in an argument, hearing for the first time an argument made well that you didn’t think you had to take seriously. You want that to be part of what goes on in universities.

As a University of Chicago philosophy professor once told me, we respect you here if you can defend your ideas. That’s a nice way of putting it—maybe a little less forbidding than “clash of ideas.” Students, by the time they are some way into university life, should be interested in defending their ideas. If they are going to, for example, a lecture or a political speech by a person whom they dislike, distrust, and are opposed to, they should have some pride in their ability to ask a hard question instead of shouting the person down. Why? Because it’s a university.

That seems to me to move towards wanting people who are qualified. Are the people who are qualified the people who score the highest on these merit tests? On the whole, I think that should be a guiding line. I believe it is one of the recommendations in our report: that a baseline be set explicitly—which is not there yet in Yale’s advertisements for potential applicants—that if your SAT score is below a certain threshold, and it’s going to be a pretty high one, you are not encouraged to apply because you’re probably only in for disappointment. They don’t do that yet. In fact, in previous outreach to potential applicants, they say don’t worry too much about SATs.

As you know, schools like Stanford, Berkeley, Yale, Columbia, and Harvard, around the time of the George Floyd protests and COVID in 2020-2021, abandoned the SAT for a while and then gradually took it back because it was so impractical not to have it be a factor at all. There shouldn’t be a hangdog attitude about using this test—it’s merely objective. It should be a good test, one that tells you something about the student.

Mounk: One of the interesting things about the SAT test is that it went test-optional in most of those universities, which of course meant that if you had a very good score, you would include it. But if you had a bad score—yet otherwise had a great experience volunteering in Ecuador, or teachers who were really pushing you—you would omit it. The effect of that in the incoming class was, in many universities, very negative: there was a serious drop-off in intellectual quality.

The other thing this speaks to is exactly the nexus of admissions and opacity, and perhaps—I don’t like the term too much—a form of mollycoddling that has just become the background hum of the American elite class. There is a very odd mix of a highly meritocratic culture, with outsized returns for getting into Yale or into the investment bank you want to work for, and a culture that is very reluctant to be explicit about criteria. You see that in admissions and you see that in grading.

Part of the advantage of not having minimum SAT cutoffs, let’s be clear, is that you can do all kinds of social engineering. That’s one reason why universities are reluctant to adopt them. But part of it is simply that it feels mean to tell people that if your SAT is below 1300, you probably don’t stand a chance of getting into Yale. So let’s not say that explicitly—even though in reality, I’m sure these admissions offices that claim to read every file holistically probably discard every application below certain criteria without looking at it for more than two seconds. By not writing it on the website, you are actually inviting lots of students to spend a great deal of time on these applications, to get their hopes up, to think that perhaps they are such a unique student that their wonderfully neglected talents will be recognized—only to be disappointed six months later. It’s a kind of fake niceness.

Subscribe now

That is related to grade inflation. I am a relatively soft teacher in the sense that I don’t want my students to be punished for taking my class rather than the class of some colleague of mine. Because grade inflation has become the practice, my grades are about as inflated. I try not to make them more inflated than those of my colleagues, but I don’t think they’re less inflated either. There is just no individual incentive to hold firm. But I think that’s also fake niceness. My students deserve to know whether the piece of writing they have handed in is poor, decent, good, or exceptional. At the moment we don’t have the signals to send them for them to know that about themselves. At the moment that feels nice, but it isn’t actually nice if we take intellectual development seriously. Universities should have a stronger sense of mission and be able to stand up against this better than they have. But it’s a downstream effect, I think, from a broader set of cultural attitudes that have become very widespread in the American professional-managerial class.

Bromwich: There is a well-meaning and rather innocent democratic idealism reflected here: we want all kinds of people, we don’t want to feel that we discriminate, we don’t want to feel we’re an aristocratic country. As the Ivy League—which was an aristocracy if anything was—began to become more democratic, at least in its surface presentation, holistic considerations began to take hold: find a student who has overcome an obstacle, a student who is a fascinating, strong personality and intellectually good enough. That kind of exception started to be made, and it is one form of diversity.

The word “diversity,” before Trump—who uses it to cover everything having to do with civil rights, gender, and whatever else—really meant devoting special attention to race, gender, ethnic background, immigration status, and cultural and ethnic issues. But diversity is also just wanting different types of people, and that impulse goes far back.

This is touched on briefly in the preface to our report. Anti-intellectualism—or, to put it more politely, non-intellectualism—has been a major strand of American life. It seems wrong for the academy to present itself as participating in this unprejudiced non-intellectualism, but there is a little bit of that too. It leads to unfortunate effects when you try to have a popular face while not really being a popular institution.

Mounk: I find the idea that you need to put together a diverse class very strange. I was an undergraduate in England at Cambridge, and the admissions system is so fragmented that there’s no chance of putting together a unified class for the whole university, because colleges do their own admissions per subject. There’s no real way of doing that. Yet the university orchestra somehow always had a second violinist. The idea in America is that you need to make sure that you have this kind of student and that kind of student. But there is a law of large numbers in statistics: if you admit the smartest students, you’re going to end up having one student who happens to be interested in music and one student who happens to be interested in sports. I just don’t buy the premise that active curation is necessary to create a diverse class. My class at Cambridge was every bit as diverse in terms of interests, talents, and the way students spent their free time as a class in America—without any constituency being able to say, we need to make sure we have somebody who can do this sport or that.

The report rightly calls out all kinds of special categories that continue to be given preferential treatment. These include athletes, the children of alumni, and—this is a minor point numerically, but one I find very striking—the children of faculty and staff, who explicitly get a leg up. That is really quite remarkable.

The one thing the report does not include is a treatment of race. That is partially because officially we no longer have affirmative action at Ivy League universities in the wake of recent Supreme Court judgments. But that is where the opacity comes back in. When you look at some of the amicus briefs that various universities wrote in the Supreme Court litigation about affirmative action, they said that the number of black students would decline radically if they were not able to practice affirmative action. In the case of Harvard, one of the calculations presented by Harvard’s side said that without affirmative action, the number of black students at the university would go down from about 14% to something like 2%. Yet after the Supreme Court outlawed affirmative action, the number didn’t budge—or it budged only a tiny bit. At Yale, it went from 14% before the Supreme Court judgment to 14% a couple of years later, and I think it has since fallen a little to 12%.

Bromwich: The desire to have the percentage of students in an entering class reflect the percentage in the population goes with, I think, a fallacy that Michael Oakeshott talks about in some of his essays on education, which he calls the reflection theory of culture. That is to say, the idea that higher education and universities should, in all possible respects, reflect the society they are meant to serve. But institutions are good at functioning for different purposes. If you’re thinking about an activity—call it an institution—such as classical music, or engineering at a high level of specialization, or the armed forces, the kind of abilities needed for one thing or another aren’t necessarily going to reflect the distribution in the population. Universities are a large institution and you should perhaps strive to make them as representative as possible. But the reflection theory also carries the trouble that it makes the university think it should reflect attitudes in society as well as populations. Again, I think that’s wrong.

Mounk: That brings us to the question of intellectual diversity. You alluded to this briefly earlier: the vast majority of faculty at the leading schools lean left, many of them registered Democrats. That in itself need not be a problem. But I do think it’s indicative of a deeper problem—that the debate at these universities lies, as I see it, between the identitarian left and the liberal left. There is probably a slight preponderance of the liberal left on most campuses, not necessarily in every department or every field. But everything that falls outside of that range is hard to formulate at those universities without experiencing significant pushback.

Some of the solutions to this problem I view somewhat critically. In an ideal world, we wouldn’t hire colleagues at our universities in order for them to be intellectually diverse. First, it’s very hard for one person to be intellectually diverse in themselves unless they hold a very incoherent view of the world. Second, it’s rather odd to have a colleague who is just there to have a different opinion. But the truth is that political criteria are currently a very large part of the selection process, especially in the social sciences and the humanities, where the quality of work certainly matters, but where people whose views fall too far outside the consensus in a particular department often aren’t even considered in the first place. There is an effective application of ideological criteria that is covert and sometimes not even self-conscious—it’s simply that this person falls outside a reasonable fold, and so they are not considered. That is a very effective implicit political filter.

Unless you are able to upend that, perhaps the only alternative is to sometimes hire people who fall outside that consensus. It’s very hard to know how to fix this. As in the case of a broader reluctance to state criteria explicitly, this too is downstream from a larger transformation of the professional-managerial class. A lot of the problem today is that the Democrats have simply become the party of the professional-managerial class. As professors, university administrators, and (to a significant extent) students are recruited from the ranks of the professional-managerial class and its offspring, there is always going to be some lean in that direction, as long as that is the nature of our political cleavage.

So how can universities improve on intellectual diversity and foster genuine debates on campus, in a way that doesn’t itself run counter to the principles of free inquiry and that avoids political litmus tests for faculty hiring?

Bromwich: It’s very hard. What you’re describing is a standoff between two desirable goods. On the one hand, the intellectual autonomy of departments: departments should be able to choose to hire, and then possibly to tenure, the scholars they consider the best in a given field. On the other hand, if you have very little variety of opinion in a field such as politics—where there is in fact great variety outside the academy—there is something lacking in the education students are going to get.

This is called, by people who do the kind of study this report was working at, the pipeline problem. The pipeline problem of having so few black people represented in the academy was addressed by affirmative action, but affirmative action is no longer constitutional and may have run its course in any case.

Where the left-liberal side got its strongest foothold was in certain departments in the humanities and social sciences, but also in the studies programs—black studies, ethnic studies of various kinds, immigration studies, gender studies, and so on.

Mounk: Is that where the left-liberal or the identitarian left got its foothold? My sense—and I know Harvard much better than Yale—is that at Harvard, the government department and the main faculties have a left-liberal predominance.

Bromwich: Correction taken. I meant consciously political and consciously left, but you’re right—that would better be described as identitarian.

In any case, if you want a pipeline going the other way, you can’t do it by political affirmative action. That is, as you said, a litmus test. But universities—Harvard, Yale, Princeton, Berkeley, Stanford, and so on—have created what they call centers or institutes, which aren’t quite academic programs, let alone full departments, but where people can have a permanent presence on campus and teach credit-bearing classes. It is imaginable that, at some places, if done with imagination and scruple, these centers could provide more room for people with academic training to hold views that inform their teaching without becoming dogma—or to hold views that are libertarian in ways that fall outside the usual range. That is one solution, but it is a standoff. It is a hell of a problem, how to change the composition of faculties without somehow compromising intellectual and departmental autonomy.

Mounk: What about the student side of this? The report refers, at surprising length I thought, to the famous Halloween incident, in which there was an email from the office of intercultural affairs urging people to be very sensitive in the Halloween costumes they chose. There was then a response to that by Erika Christakis, who was—as it was then called—the master, along with her husband Nicholas Christakis, the associate master, of one of the residential houses at Yale. Her response suggested that students should be able to think for themselves about these kinds of things. There was a huge eruption of anger that did look rather like some form of cultural revolution—an effigy of Nicholas in the courtyard and so on. A subsection of students, but an influential subsection, saw it as their calling to impose a certain identitarian orthodoxy on campus. That included intimidating senior faculty members, as in this instance, and sometimes intimidating their fellow students. There is some discussion in the report about the norms that should be established so that people feel free to share their opinions in class without the fear of being canceled afterwards on social media or in other ways.

I have personally found that that moment has somewhat passed. For the last few years of teaching, I’ve been struck by the fact that for many students, these ideas are now the received wisdom of what they’ve been told in high school, middle school, and sometimes elementary school. It depends a little on where they grew up—if they grew up, for example, in the suburbs of a major metropolitan area like New York, Los Angeles, or Boston, these are largely just the ideas their teachers have always taught them. As a result, they tend to take them for granted. It’s the world they’ve grown up in. But they no longer think of themselves as bearers of a flame whose goal is to impose it on others. It’s more that this is what their teachers told them and, like most things their teachers told them, they assume it’s true—until somebody has a different opinion, which surprises them, and they find they want to talk about it. It feels somewhat less fraught.

What is your impression of campus culture at Yale—has it somewhat improved? And more broadly, how big is the challenge today and what are some possible solutions? One of the recommendations in the report is to remove electronics from the classroom. The report also entertains but ultimately dismisses the idea of a Chatham House rule prohibiting attribution of remarks made in class. Where do things stand and what can be done?

Bromwich: I agree with your impression that the pressure zone, so to speak, has lifted somewhat in the last two or three years. A person I’m very close to in this household, who is a psychologist, speculated that in the year 2020 this country had a nervous breakdown—the whole country had a nervous breakdown. COVID was an element of it, but so were the George Floyd protests and the disorders in cities. The Halloween event you mentioned came in 2015-16, so that is an earlier BLM moment, and it had been going on for a long time—really from Obama’s second term into recent days.

But I do agree that there is more tolerance and interest in the exchange of ideas, and more animated talk and susceptibility to humor in the classroom, as far as I am able to discern. That is a good thing. It means that younger teachers, who tend to be cautious because they worry about student evaluations, will not constantly develop twenty different tones of voice for saying that’s interesting—and may instead learn how to say, I don’t really think that’s true and here’s why, or that’s absolutely true but let’s go further with it. That kind of give and take in a classroom is very important.

One of the things you could hear from conservative students in the worse recent years—here at Yale, which is not as badly off as other places because it has a conservative political union and a conservative presence in the background, so those students don’t feel entirely isolated—is that sometimes in class, when they made a remark that showed their traditionalist sympathies, they would not get denounced or shamed by anyone in the class or by the teacher, but the teacher would be silent for thirty seconds and then move on to something else. That’s not good. A greater openness that reflects a community with divergent, somewhat disharmonizing points of view is a good thing, to the extent that it really is coming back. I think it may be.

Mounk: Is one challenge not explicitly mentioned in the report that there is a minority of faculty who really do abuse their power in the classroom to impose their ideological views? My impression is that most faculty members do not do that, and I certainly think most of my colleagues do not. But I am struck, speaking to students, that they very consistently bring up experiences of this.

I go out of my way, when I explain what I think a good essay is for my class, to make clear that I don’t care whether students agree with me or not. The last thing I want is for somebody to badly parrot my views back to me—that’s not going to earn a great grade. If you write something I agree with, I want to feel that it is a really strong, interesting representation of that view. If you write something I disagree with, the question is not whether I am going to come to agree with you—that is unlikely. What I want to feel is whether the pull of that argument in my mind is a little bit stronger after reading your essay than it was at the beginning. Have you made a case for a point of view I may happen to disagree with, where I find myself thinking: I still probably disagree, but I see the force of that—I see why somebody might believe that.

Because I am very explicit about this, students sometimes say that they feel they can say what they think in my class, because it is not always the case elsewhere. I always listen up at that. When is it not the case? I remember one recent instance in particular where a student said that in high school, a teacher had a set of views—actually a rather woke set of views, views the student largely shared—but the student chafed at the fact that any disagreement was clearly going to be punished.

Are there ways that universities need to reckon with that, even just in terms of training teachers and setting clear expectations? You don’t want to be too intrusive, and you certainly don’t want auditing of every grade, which would be open to abuse in all kinds of other ways. But even if it is one in twenty faculty members, that is an experience most undergraduates are going to have once or perhaps twice during their college career. Is that one of the things that damages trust in higher education, and something we need to deal with?

Bromwich: I don’t know if it is one in twenty teachers, but I agree that it goes against the ethic of teaching. If you have anything resembling a Hippocratic oath as a teacher of the liberal arts and sciences, it should be that a view that is well defended gets the respect of the teacher, and a view that is earnestly and genuinely meant but not well defended receives a response that is not crushing or personally harsh, but reasonable—an example of how to point out the fallacies or the loopholes in a rival argument. Teachers should be exemplary in that way and it should be part of their training. But what is involved in acquiring a PhD is so specialized in other ways that that element of pedagogy is often neglected.

Some of it is intuitive. Some of it is, as I’m sure you have experienced, getting to know what your classroom presence is like and what it’s like to teach different kinds of classes. I remember in my early days teaching at Princeton—I was in my 20s—being morbidly worried about what I thought of as dead time, what disc jockeys call dead air: nothing being said. If I asked a question and there was no answer for ten seconds, I got very anxious. But as a teacher you learn that if a question was well asked, you simply wait.

Mounk: The best piece of teaching advice I’ve ever gotten was in graduate school. The department put on a little meeting for people who were going to be teaching assistants for the first time, and the best piece of advice—which I follow very often—is this: if you ask a question and you look anxious that people might not answer it, and you look like you’re going to jump in yourself, students think they can just sit back and let the teacher do their thing. What you should do instead is lean back, look comfortable, and let an awkward silence arise. Some student always can’t bear the awkwardness and jumps in, and suddenly you have a real discussion. I don’t generally have trouble getting my students to talk in the classroom, but on the occasions when that happens, I do that. Sometimes I even call it out explicitly: I’m happy to sit here in awkward silence for as long as you like. When you do that, somebody always jumps in.

Bromwich: Sometimes you haven’t asked the question well, and that shouldn’t lead to too much self-consciousness either. You learn to listen to yourself and say, a few seconds later, you know what? I posed that badly. Let me try again. There was a philosophy professor at Princeton—modal logic and philosophy of language and other such things—David Lewis, who spoke perfect paragraphs and was an excellent lecturer. When he would walk back and forth on the stage and, in the middle of a sentence, realize he was about to talk nonsense, he would stop and say so, then go back and say the whole thing again. Teachers, in a quieter way, should be able to do that.

Let me read a passage from a British education master of the 19th century that is quoted by Oakeshott, because I think it speaks to exactly what we’re talking about—the kinds of habits and manners that go with thinking in the context of higher education:

A certain amount of knowledge you can indeed with average faculties acquire so as to retain, nor need you regret the hours you spend on much that is forgotten, for the shadow of lost knowledge at least protects you from many illusions. But you go to a great school not so much for knowledge as for arts and habits, for the habit of attention, for the art of expression, for the art of assuming at a moment’s notice a new intellectual position, for the art of entering quickly into another person’s thoughts, for the habit of submitting to censure and refutation, for the art of indicating assent or dissent in graduated terms.

That last part is obviously what has become most challenging for students in the last generation or so.

In the rest of this conversation, Yascha and David discuss why the number of students choosing humanities subjects is declining and how colleges and universities can regain public trust. This part of the conversation is reserved for paying subscribers…

Read more