Ben Ansell is Professor of Comparative Democratic Institutions at the University of Oxford and a Fellow of the British Academy. He is the author of Why Politics Fails.
In this week’s conversation, Yascha Mounk and Ben Ansell discuss the collective goods that we all desire (democracy, equality, solidarity, security and prosperity) and why political systems struggle to bring them about; the implicit trade-offs in democratic systems; and the institutions we need to create more just and prosperous societies.
The views expressed are those of the speakers, not those of Persuasion. The transcript has been condensed and lightly edited for clarity.
Yascha Mounk: Your book, just out, is called Why Politics Fails. One of the standard answers that a lot of people would probably give is that politicians fail; that politicians are self-serving, evil, conniving, and out of touch.
Before we get into what your answer is, why is that the wrong answer? It seems like a pretty plausible answer.
Ben Ansell: It is a plausible answer to the extent that our politicians share the frailties that we all share. My answer is that, broadly, politics fails because all of us are driven by individual self-interest. That means that we find it hard to coalesce towards the kind of collective goals we might want. Our politicians are representatives of us and they face an even sharper set of constraints and incentives to misbehave, perhaps, than we do. Perhaps they're worse versions of us. That might be right. But ultimately, they disagree, at least in part—and often, mostly—because we disagree.
I construed that in the book by, to some extent, thinking the worst of everybody. But I'm not trying to create a moral message to say that humans are sort of fallen and that we all behave poorly in every instance. There are all kinds of ways in which we can have harmonious relationships with one another. But those usually only occur when we all agree on outcomes to begin with. Or when there aren't any interdependencies among us where, if I get something, it makes it harder for you to get that thing. And the moment we end up in a world where we do disagree with one another—and I think we've seen over the last several decades that we all do fundamentally disagree in liberal democracies and that in part explains our political polarization—we exit the world of harmony.
We do have to recognize and adjust for the fact that people will disagree, on the one hand, so we have to think about ways in which our democratic institutions can manage that disagreement effectively, but also that people will misbehave when they're not monitored tyrannically. And nobody wants to be monitored tyrannically. And we might misstate what we want or we might fail to come through on promises that we make. If we assume that those kinds of dynamics are pretty pervasive, our politics will work better if we try and design institutions that can respond to those incentives.
But the institutions that we have often don't fit the current state of events that we're in. We often feel frustrated because an institution designed for an earlier time is failing to give us what we want today. And that's a dance really. Getting rid of them in a kind of populist bonfire of the institutions creates the dilemma that we now enter a world where nobody can trust what anyone else is going to do anymore and I think is responsible, in part, for some of the volatility we see in contemporary politics. But going along with institutions just because they're there can lead us to the situation we have in, say, the US Senate, where very, very small states with around a million people get the same representation as states with tens of millions of people. Now we have a different problem where the shoe doesn't really fit anymore. Ultimately, our political problems sort of swing between those two poles.
Mounk: Let's talk through some of those traps that you discuss in the book. Perhaps you can briefly list them and then we can start with the security trap.
Ansell: There are five broad goals that most people can generally agree on: democracy, equality, solidarity, security, and prosperity. To give you the example of security that you mentioned, that's a trap where we can't avoid anarchy without risking tyranny, which is precisely the problem that Thomas Hobbes is in part solving, but in part creating, with Leviathan.
I think Western European countries in particular seem to have found the right place within this trade-off. It was successfully resolved in those countries that made the police force and the citizenry less different from one another, creating civil police forces where the police were supposed to be drawn from the local population and to govern the local population. That was the initial idea, for example, behind the London Metropolitan Police Force, and it was very much created in contrast with what we perceive to be the autocratic continental police forces, which are essentially military offshoots—gendarmeries, which are officers wearing military-like uniforms and carrying weapons around. We still see some of that, with the Italian Carabinieri and so on. That type of more tyrannical-looking police force was perhaps a step too far for some countries. But those countries that did develop these civic police forces, I think, did straddle that balance quite nicely. That said, even in the United Kingdom, the London Metropolitan Police force has in recent years murdered an innocent man suspecting him to be a terrorist during the 7/7 bombings, and has had an officer rape and murder a citizen whom he picked up walking home during the COVID lockdown, and has had a serial rapist arrested. Even in the country that seems to have resolved most of these questions about how one avoids anarchy without having an overbearing police force, we still clearly have a number of police officers who are stepping past that mark.
What I would advise people to do is to always, when thinking about the security problems, consider the individual incentives of citizens to misbehave, on the one hand, and to think about the incentives of police officers to act in ways that maximize the interest of each individual officer. That might just be earning more money. It might be that you joined the police force because you like exerting authority on people. It might be corruption. But the simplest way of avoiding individual incentives creating bad outcomes is to observe people in some way and to set up a framework where we can see what both citizens and the police are doing. In the UK, CCTV cameras have been around since the 1990s. That kind of observation can sort of stop anarchy among citizens. But we now have that type of observation among security forces with body cameras on police officers. And there's lots of evidence that it does tend to moderate their behaviors. They're less likely to draw their weapons, for example. So one could make an argument that observation helps and we certainly have the technologies to do so. But that leads us to a couple of dilemmas. George Floyd was murdered on camera, and not just on the cameras of citizens, but in fact on the body cameras of a number of Derek Chauvin’s fellow officers. The existence of observation alone doesn't prevent bad police behavior—although, it did, in the long run, help prosecute Derek Chauvin. And then we have the broader question, which is how we feel about being constantly observed, and is that a different form of tyranny?
My current view of this is rather than defunding the police, observing the police would be the best way of ensuring public safety but removing or reducing the possibility of egregious police violence.
Mounk: Security is perhaps the most fundamental of your dilemmas. But I think second after security is perhaps democracy.
What is the nature of the collective action dilemma when we're talking about democracy?
Ansell: Ultimately, democracy is about how we make decisions and how difficult it is to make decisions. In a way, many of the problems later on and our inability to achieve, say, our security goals might themselves be reduced to our inability to actually make collective decisions at all. Why can't we make collective decisions? The democracy trap, as I set it up, essentially boils down to the fact that there isn't such a clear thing as the will of the people, so that there's no way for us to know what society truly wants. That doesn't make a great deal of sense unless everybody literally agrees on everything. Instead, we have to have some rules and procedures to aggregate people's differences and come to some kind of collective answer.
The way I talk about democracy is to contrast a situation where you have pervasive chaos and you can't make any decisions at all. And there is no unimpeachable way of taking lots of individual preferences and turning them into something that collectively makes sense and that can indeed happen under certain conditions. I talk a lot about the votes around Brexit and the difficulty that the ruling party had back between the vote in 2016 and 2019 in coming to any kind of agreement at all. But the alternative that we can get into when we resolve chaos, sometimes, is just polarization, which I think is more symptomatic of the American experience over the last 20 years: decisions can get made, but they’re decisions where one half of the population is set against the other half with great venom and discontent.
Mounk: One of the striking things about America is that in public opinion polls, polarization is actually not that extreme on policy issues. When you ask people how they feel about the police, abortion, their view of history or even economic policy, they come up with what to me seem pretty reasonable positions which a majority of people agree with and which are not particularly extreme. But on most of these issues, the two main political parties take stances that both are quite far divorced from where the bulk of the American population lies and extremely far away from each other. How do we get to that outcome?
Ansell: As you say, Americans tend to agree on the specifics, but they often disagree on more general sort of philosophical issues. We see that in the United Kingdom as well. There are really large differences across parties in the UK on questions like the death penalty or whether children respect their elders enough—big-picture social issues. But if you ask people about government spending on education, or even the general tax level, you find a lot more agreement on this sort of specific policy question. So I don't think that the American experience is unique. Polarization between parties reflects in part these more fundamental preferences that people have or claim to have, but I think it does, in the American and, increasingly, the British context, also come from the way in which the leaders of those parties are increasingly elected through primary elections; if the American public as a whole has quite moderate preferences, that's not always true among those people who turn out to vote in the primary, who will tend to have more extreme preferences. It's also the case that moderate congressional representatives or senators tend to win elections in swing states and therefore are less likely to remain in office for longer periods of time, because those are swing states where a Republican is replaced by a Democrat and vice versa. The same is true in the United Kingdom, where there are safe seats that Conservatives and Labour MPs always win. And so the more moderate members of parliament, who represent those more moderate constituencies, are going to be powerless. In part, our political systems are structured to empower the extremists, both among voters and among politicians.
Why is it that people seem to agree on the specifics, but not on the general? Well, I think here parties have been increasingly good at taking their philosophies, which to some degree bubble up on the extremes, and guiding voters who are aligned with that party to think in that way—in other words, the parties are leading the voters. On the broad philosophical issues about fairness or about the role of government—even if, on specifics, people find that they might agree—they often find themselves in disagreement, which I think is quite elite-driven in both countries.
Mounk: I strongly agree with your point about primary elections. It’s also really important to note that moderates tend to come from swing districts and therefore tend to sometimes lose elections, and in a system that's as seniority-based as the United States Senate that actually gives greater power to people who tend to be more ideologically extreme.
But talk me through what you think would be a better constitutional setup. Are there institutional reforms that would help the political parties represent where the center of public opinion seems to lie more effectively?
Ansell: There are revolutionary changes that you could make which would almost literally involve a revolution, but I think are still worth discussing. Both Britain and America have electoral systems that privilege two parties. Those parties can potentially be more easily captured by extremes, but that's not always the case. You can have highly successful moderate leaders who are able to push the party in that direction, but certainly there is a kind of inbuilt centrifugal force that pushes people out to the extremes.
I'm not fully sold on proportional representation, not least because I think of Weimar Germany, the French Fourth Republic, contemporary Israel—none of these are great examples for proportional representation and multiparty systems. So there are weaknesses to that. But I think a system where people's votes translate into seats for parties in a much clearer fashion than is true in the United Kingdom or the United States would have a number of advantages in stabilizing politics. Essentially while lots of new parties can emerge and die in PR systems (and that's not necessarily a bad thing, either, because it means that new ideas can get represented), it's also the case that coalitions tend to be larger than just 50%. Oftentimes, you'll see that there are more moderate parties that are constantly in government. Even when you have a party more to the extreme governing in lots of PR coalitions, it can often end up governing with a whole range of other parties. Mark Rutte has been the Prime Minister of the Netherlands for over a decade now and he's had four separate coalitions in which almost every major party in the Netherlands had been represented at some point. That means that there is more stability in policymaking to the extent that there is the same Prime Minister and many of the same figures in charge, but also everybody gets a turn at some point. It's less slash and burn than in countries like the United Kingdom, where it's usually either the Conservatives or the Labour Party and there's a bridge-burning tendency where you get rid of all the reforms of the previous party and introduce yours as long as you can stay in government. Proportional representation systems tend to have the stability of a coalition that's much broader across the public and that almost always privileges groups somewhere in the middle of the distribution of political preferences. Now, that might be stultifying, in lots of ways, and it can be unstable too, as in Weimar. But in the northern European countries it has been really quite stable. That's been good for things like investment, for lower inequality in those countries, and for consistent funding of their public spending projects and consistent taxation to pay for that spending.
There is a danger for all of us that the grass is always greener on the other side. PR systems, I think, are more likely to create chaos and indecision, things not happening. There was the post-war era in Italy, where there was roughly a new Prime Minister per year. But of course, that's the period in which the Italian economy suddenly zooms to the front of the line in developed country wealth up through the 1980s. In some ways, it's the worst political time for Italy, but economically, it's fabulous for the country, and even the Fourth Republic in France is a period of growth. So sometimes it's hard to separate out these things. It feels unaccountable, because it’s hard to know who to punish, and seems unstable, because you have lots of politicians going all the time. And yet, it's also associated with a period of declining inequality, growing prosperity and the development of the welfare state. Proportional representation is going to work least well where there are a series of chaotic, different preferences that just can't be aggregated easily.
Mounk: I'm going to zoom ahead to the third trap, because you are invoking it, which is that of prosperity. Why is it that a country like Italy was able to sustain all of that economic growth in the 1950s up through the ‘80s, and yet has now had 30 or 40 years of stagnation? Is there a broader story to be told here, about why we obviously all want prosperity, and yet some countries like Italy end up getting so mired and stuck?
Ansell: Broadly speaking, I do think most difficulties of attaining prosperity are about the fact that long-run planning and long-run economic growth generally require some kind of immediate cost bearing by somebody. If we borrow money to go to university, we're making an investment that's costly to us now and will pay off a long time in the future. Many investments have that quality, but political systems are not well structured to support that kind of time-based decision making.
President Roosevelt had a massive majority in both houses by a massive margin in the late 1930s and initially wanted to introduce the costs of Social Security immediately, that is the payments, into paychecks but have the first benefits come about eight years later. But after losing a congressional election in 1938, they just reversed it completely—actually, the benefits will come at the same time as the costs. They were unable to separate out the two. And that's a much more general problem that we can see in examples from the biggest of all, which is climate change (taking costs now to get to net zero when we know the payoff may not even happen while we're alive). But it also exists with all kinds of development choices. Putting in hard work for politicians, if you know that you're not guaranteed to be in office three or four years later, is very difficult.
Another example, which was a lovely policy idea, to some extent, was the baby bonds that Tony Blair introduced in the New Labour government of the early 2000s. That was a small amount of money, about one thousand pounds, that everybody would get at birth and be able to access at 18. Now, the problem is very few political parties are still in power 18 years later. So they only came in last year, by which time Blair had not only been replaced but, in fact, we've had four separate Conservative Prime Ministers. That's a sort of extreme example of a policy that might well be beneficial—it would be useful for everybody to have these opportunities when they enter adulthood—but where the political payoffs are just a mile away from the costs that are incurred.
Mounk: Talk us through equality. What is the trap of equality and why is that something that we all desire but that we individually act to undermine?
Ansell: I counterpose the difficulty of achieving equal rights and freedoms, or even equal opportunities, on the one hand, with achieving equal outcomes. What I'm drawing on there is a large range of contemporary political theory where everybody agrees to equal treatment in some regard—from the kind of libertarian view of equal property rights and equal treatment under the law and to dispose of that property in any way you want to a kind of fully egalitarian view that equal outcomes are what is important. I think it's obvious to most people that it's very hard to achieve both at the same time; that to have extremely equal outcomes while letting everybody do what they want and dispose of their property as they want is almost logistically impossible. There's a nice example from the political theorist G.A. Cohen where he argues that you could have a world where everybody gets paid exactly the same and everybody can do exactly what they want in the labor market, and do what they're best at, as long as everybody is willing to have an “egalitarian ethos,” where everybody is willing to accept exactly the same returns for what they do. Now there is no contradiction. But that's a long way away from the world in which we live. In almost all cases, getting towards equal outcomes is going to lead to some kind of restrictions on rights and vice versa.
Mounk: Even at a more fundamental level, it's obvious in the literature that there's just very, very basic conflicts between different forms of equality, which is a point made by Amartya Sen and others. Either you have real equality or similarity in how much money people take home at the end of the month or you have equality in how much they’re paid per hour. But unless there's a very tyrannical state that forces everybody to work exactly 30 hours a week, you cannot have both. Either you’re still going to have some people who are much more affluent because they choose to work a lot more or you're going to have to pay people very differentially for the time they do work. Either way, society is unequal in some very important respects.
Ansell: Equal processes almost always lead to people doing divergent things. And to stop those divergent things from producing divergent outcomes generally requires a restriction on those processes—that trade-off is inevitable and you can end up in in extreme examples, in a kind of Stalinism where you essentially enslave or exhort everybody to work as hard as they can and give them all the same. You can do that but it becomes tyrannical. Or you can end up in a world where you wave your hands in the air and say, “Well, who knows? Everyone should do what they want, no restrictions,” and then, of course, you end up with extremely unequal outcomes. Now, again, those are the extremes of politics failing. In most advanced industrial countries we live somewhere in the middle, where we're constantly engaged in this trade-off between the two.
People really don't like extremes of inequality. But that doesn't mean that they are preternaturally driven to want extremes of equality. That doesn't necessarily hold. People don't care, it seems from most surveys, about equal wealth, per se. But they do care quite strongly about their treatment. People care about being treated equally—no exceptions, the rules apply to all of us. They care less about the more abstract ideas of the overall aggregate equality of outcomes in society.
Mounk: Your last trap is about solidarity—what is the trap of solidarity, how is that different from the trap of equality, and how do we get out of it?
Ansell: Solidarity is how we look after one another. In some ways, it applies, I suppose, to some kind of equalization of welfare: if somebody is really poorly off, you're going to make them better off. But it doesn't really require any kind of major narrowing of equality in society. It is, of course, a form of equal treatment. The idea is that we would all like it if, when we fell on hard times, we had somebody, or society in general, help us out. That's what solidarity is. In substantive terms, that means the policies that political scientists call the “welfare state”—health care, education, pensions, unemployment insurance, and so on. All of those things can exist in highly unequal countries like the United States or South Africa and they can exist in highly equal countries like the Czech Republic. They obviously operate differently in those countries and they can be designed differently. But I think they follow a common human impulse, which is, even in highly unequal countries, there is a lot of charitable giving or a lot of religious common feeling. I think that's as true in the United States as anywhere else. In particular, I talk about the difficulty that often emerges of solidarity across a diverse group of people who don't always feel a sense of oneness with one another. I think that is something that is heightened by America's racial politics. But it's not a solely American experience. Lots of ethnically or linguistically diverse countries have had difficulty in creating solidaristic social programs that benefit lots of diverse groups who might not always want to look after each other.
If we were having this conversation in the late 1990s or early 2000s I think the general consensus in political science and economics at that time was that more diverse countries have smaller welfare states and that these things must be related. I think there are increasingly some arguments that say you could have a Dutch model, where you have a kind of pillarisation: you have diversity in the Dutch case, really, in religious terms, and there's a famous line about the Dutch education system which was “If a window broke in the Calvinist school, the Catholic school and the public school would also get new windows.” You would find ways of spreading the money around. So diversity doesn't have to lead to low spending. It's not a mechanical effect. But secondly, I think there's increasingly strong evidence that whatever differences there are in diversity can be overcome by appeals to some kind of broader—and it is, indeed, often national—oneness. If it's about creating solidarity across a group of people, canny political entrepreneurs can reframe that debate. And they can reframe it about the nation as a whole.
I think you can see that to some degree in places like Scotland with the SNP idea of “civic nationalism,” that I think has worked quite well as a unifying political campaign across individual divides in Scotland under Nicola Sturgeon. So I don't think that diversity means doom. I do think that welfare chauvinism—a political science term for wanting public spending only on native-born populations—is popular in Europe. To give you an example of that, anybody used to be able to come to the UK and get treated by the NHS, then that was restricted to only legal immigrants and nationals. And then legal immigrants themselves had to pay a fee. Those were very politically popular decisions that narrowed the scope of solidarity. And we do have to push back against that. But I think there are ways of doing that and I think civic nationalism is one potential mechanism.
Mounk: It seems to me that there is another trade-off, or trap, here, which is that on the one hand, it is understandable that you want to call attention to conflicts and inequalities in a country that over the last few decades has become much more likely to frame public discourse around how some groups are faring less well than others and the ways in which they are discriminated against and so on. But on the other hand, that, of course, seems to counteract exactly that kind of emphasis of “we are all part of the same nation.”
How do you combine those things in terms of how the public discourse is framed today? In Britain, the United States and some other countries, how do you think we're doing at reconciling those two goals?
Ansell: It's an inherent challenge when you have a kind of assimilationist argument up against the multiculturalist argument, where the people who are the proponents of multiculturalism and also the people who are, generally speaking, proponents of solidarism, because it creates a trap for them in the sense that if they are speaking about differences, and people hear that and think “Gosh, those people are different,” that reduces solidarity. And it creates potential for bad faith arguments from assimilationists, who don't really want to spend more on the welfare state—“We would if only you stopped highlighting these differences.” That is very challenging. I do think that most of the change in this will come from the general public blurring existing differences in their minds and simply caring less about ethnic or religious differences within society. Now, that becomes tricky if you're trying also to make up for historic inequalities in those differences. But if the general public stops caring as much about those differences, then they're sort of solving the solidarity problem, in part, themselves, including everybody in a sort of single nation—unless when people become more diversity tolerant they also are still exercised about making up for past inequalities. Then you can lose that kind of reparative effect. And that's going to be a challenge for people who want to make up for the previous misfortunes of ethnic or religious or linguistic communities, if they discover that that moment of reparation never really happens: we get a moment of national solidarity, but there's never a kind of extra push for that group.
It’s much better for us to acknowledge that we will have differences, that we can't always be trusted, and to design institutions that don't treat us as angels, to paraphrase James Madison, but understand that we're not we're not devils, either—we’re people. And we need institutions that are made for people.
And, to receive pieces like this in your inbox and support our work, subscribe below: