The Troubling Economic Logic of Racially Charged Policies

Could racially charged policies cause you to act racist, even if you aren’t actually a racist? That’s the story two scholars, Stephane Wolton and Torun Dewan from the London School of Economics and Political Science, tell in a recent paper, "A Political Economy of Social Discrimination".

On this episode, we discuss the hidden story behind policies like the Trump administration’s border wall and Muslim travel ban, or Europe’s burqa bans.

Listen on Apple Podcasts or wherever you enjoy podcasts.

Transcript:

William Howell:

Welcome to Not Another Political Podcast. I am Will Howell.

Anthony Fowler:

I am Anthony Fowler.

Wioletta Dziuda:

I'm Wioletta Dziuda.

Anthony Fowler:

We are professors in the Harris School of Public Policy at the University of Chicago.

Anthony Fowler:

As always, we're going to talk today about an exciting, salient, controversial political topic which is the recent surge of populist leaders who are passing potentially racist, potentially discriminatory policies. And we're going to talk about why they do it and what the implications of those policies might be.

Tape (Donald Trump):

We are going to build a great border wall to stop illegal immigration.

Anthony Fowler:

Trump wants to build the border wall and a lot of people think that the border wall won't actually have much significant, direct impact. It's not going to meaningfully change the number of people crossing the border and so forth. And so why would he do it? Why would Trump decide to undertake this fairly costly enterprise? It certainly gets lots of cheers at rallies but is there anything else to it?

William Howell:

Yeah, or the Muslim travel bans.

Tape (Donald Trump):

Donald J. Trump is calling for a total and complete shutdown of Muslims entering the United States until our country's representatives can figure out what the hell is going on.

William Howell:

Those Muslims who wanted to travel into the United States certainly were directly affected, but it had these larger effects by raising the salience of what it means to be Muslim in the United States. We want to then think through why that's consequential for our politics and for our labor markets.

William Howell:

Wioletta, you had a good conversation with our colleague at the London School of Economics, Stephen Walton, about a paper.

Wioletta Dziuda:

Yes. Stephen Walton and Torun Dewan, they are both stationed in London, but Europe has its own share of pretty strange policies, for example, the burqa ban...

Tape:

A European court has ruled that France's ban on full face veils does not breach religious freedom. That means the so-called "burqa ban" is legal and will stay.

Tape:

Denmark has banned the wearing of face veils in public, joining France and other European countries in outlawing the burqa and the niqab worn by some Muslim women.

Wioletta Dziuda:

At the same time, this burqa ban seems to affect very few people, very few women in Europe who wear burqas in public places and seems not to benefit the majority. So, Stephen and Torun were very puzzled by why we see politicians wasting time on these policies that actually do not benefit most of the voters, and they came up with a very interesting explanation.

Stephen Walton:

It has been the spread of these types of policies who seem, maybe not officially, but at least unofficially to target a particular minority.

Wioletta Dziuda:

So, what's so special about those policies? Should we think about them as just being racist?

Stephen Walton:

In one sense, we could, but it would be a little bit depressing because, for example, these policies are very popular. So, what we try to do in this paper is to say, "Okay, let's think there is no racial animus." There must be some other things going on behind this and that was kind of the starting point of this paper.

Wioletta Dziuda:

Okay. So, what's your explanation for these policies?

Stephen Walton:

So, our explanation is built on the idea that when you put a ban on the burqa, they become singled out as a minority in a way that they were not before. And we talk about this as raising the salience of some identity trait, of raising the salience of the identity trait of being Muslim.

Wioletta Dziuda:

And why would you want to do that?

Stephen Walton:

The basic idea is before you have these types of laws, you meet me, you're an employer, Wioletta, and you meet me. And you look at me and you say, "Oh, this guy, he seems comparatively productive. I'm going to hire him so I can make some profit."

Stephen Walton:

Another is this type of laws, and it happened that it raised the identity trait, not of Muslim in my case, but of being French. So, you meet me and you hear my accent, and you're like, "Oh, this guy is French," and that's not so great because if I have French people, people are going to think that I'm friendly to French people. And then they are going to kind of like not want to interact with me.

Wioletta Dziuda:

But that sounds racist.

Stephen Walton:

It is racist, but it's not based on any sense of animus. It's based on your fear that because you interact with me as a French, all the people will see you as being tainted by this interaction.

Wioletta Dziuda:

Suppose I'm a majority firm...

Stephen Walton:

Yes.

Wioletta Dziuda:

And I meet you, a French person, and you seem very productive.

Stephen Walton:

Yes.

Wioletta Dziuda:

Why would I not hire you? What am I afraid of?

Stephen Walton:

You're afraid that it would change your work relationship in the future. On the one hand, you could make some profit with me today by hiring me. On the other hand, if it happened that our relationship breaks down, tomorrow it would be much harder for you to find workers because all the American, non-French workers would say, "No, I don't work with this person. She's tainted. She has worked with the French people and if I work with this person I, myself, would become tainted and I'm afraid of this."

Wioletta Dziuda:

But this sounds like a great idea for the minority firms. Now they have access to all the workers and they compete with those majority firms that actually cannot hire whomever they want. Are you telling me those bans, those policies are actually great for minority firms?

Stephen Walton:

No, this is where there is a flip side of it because now let's reverse roles and let's think of you as being an Eastern European minority, and...

Wioletta Dziuda:

Which I am.

Stephen Walton:

Which you are, and let's think of me as an American worker and you as a minority firm in this case.

Wioletta Dziuda:

Yes, I own my own sausage shop.

Stephen Walton:

Exactly. Now, the problem is that you get the same type of relationship from the perspective of the worker. That is, if as a majority worker, I start working with you, I do not get a chance in the future to work with the non-Eastern European firms.

So, as a worker, I also close myself from a large pool of employers by working with you and possibly being tainted by this work that I have with you. So, it's not great for a minority firm because minority firms can only work with minority workers.

And it's not great for minority workers because minority workers can only work with minority firms. And so minority workers end up suffering from more unemployment and longer unemployment spans.

Wioletta Dziuda:

Can you tell us a little bit more about the methodology of your paper? My understanding is you did not collect any data, you did not interview voters, workers, firms, or politicians. After all, would we expect them to tell us the truth? Probably not. So, how did you come up with your findings?

Stephen Walton:

I used what is called in political science "formal theory" and in economics is more referred to as "applied theory." The idea is to think about, first, what are the incentives?

For firms, it's to make profits. And for workers, it's to get employed and get some wages. And from this, we try to think about what the structures of the labor market are as if it has been understood by other people. And when we put the notion of identity on this traditional labor market, we're seeing incentives that are very basic. What is the type of output that we get out of it?

And the goal is to have minimal assumptions that are extremely clear and to let the logic of the model speak for itself and discovering the massive impact that this simple policy can have, obviously theoretically, but still we were shocked.

I will say, we were shocked. We were depressed a little bit and so I hope that this depression can lead to a little bit more hope. I wouldn't say I'm optimistic about it, this to happen, but if it were to happen, I would be very proud.

Wioletta Dziuda:

Stephen, thank you so much. This was a fascinating conversation and I think we learned a lot. Thank you for your time.

Stephen Walton:

Thank you very much. Thank you for having me and thank you for the discussion.

Anthony Fowler:

Wioletta, tell us a little bit more about how the model works and the mechanics of it. In particular, I guess the interesting part of it is, of course, we can all explain a racist policy with people being racist. But Stephen and Torun want to explain these policies through some other mechanism.

Anthony Fowler:

So, tell us about how it might come about that you end up with a burqa ban or a border wall or a travel ban, even if people don't have the strong taste-based discrimination that we might think is driving these policies?

Wioletta Dziuda:

Yes. So, the main point of the paper is to show that, even in the work in which we have no prejudices, suppose that for some reason, I think, if I hire minorities, for example, Latinos or Muslims, then no white Americans would want to work for me. Then I might actually think twice about hiring minorities because it's easier for me to find a white, American worker than to find a minority worker. So, it's this self-fulling equilibrium where I don't even know where those beliefs come from. But once they are there in the society, it's best response for the majority workers and majority firms to behave in this way.

Anthony Fowler:

Okay. So, you're telling a story here where there's potentially lots of possible equilibria where, given a set of beliefs about how discriminatory everyone else is, we can sustain a certain pattern of behavior. And if we all believe that everyone else is discriminatory, even if we're not, we'll never see any evidence to the contrary.

Wioletta Dziuda:

So...

William Howell:

It's not that everybody else is discriminatory, not in the sense that they have a taste for discrimination. It's that they, too, think that when identities are transferred, that then one is tainted, and then should I then interact as a potential firm and hire somebody of a certain type? My ability to hire other kinds of people is going to be affected. That's what... I mean, it runs all the way through. There's no point at which somebody has a taste. It looks like it, right? When you look on the outside and you say, "All these people are behaving in ways that are decidedly discriminatory and prejudiced and racist." The equilibrium, nonetheless, can be upheld just by virtue of these anticipatory effects, no?

Wioletta Dziuda:

Absolutely. But I share Anthony's concerns. Although the paper is phrased this way, that actually there's zero taste-based discrimination, it's done this way because it wants to show us starkly that those forces might exist in the society. But I think the best way to think about this paper, what this paper tells us about, is that suppose we start with some level of discrimination, true discrimination.

For example, 50's, 60's, 70's discrimination against women. We did think women were not as productive as men. Then it's very easy to envision this discriminatory equilibrium, of course. Women are not as productive. We are not hiring women. And then imagine the beliefs changing. Imagine men actually learning that women are as productive as they are. What this paper tells us about, is that it's going to be very hard to move away from an equilibrium that's discriminatory, to an equilibrium where actually women are hired the same way as men are.

Because right now, even though I completely abandoned my false belief and let's say as a lawyer, I would be thrilled to hire Ruth Bader Ginsburg to work for me, I worry that other people might not want to come and work for me if they see a woman working in a law firm. They're going to perceive maybe this law firm as tainted or maybe weak. I don't have anything against women but I'm going to behave as if I did. So, I think for me, it's a story about how hard it is to move away from discrimination, even if the basic reasons for discrimination that we have...

William Howell:

Are shown to be false.

Wioletta Dziuda:

Exactly.

Anthony Fowler:

Can we talk about the burqa ban? So, this is kind of the leading example in their paper. So, the first question is, is that a fair assumption about the burqa ban, that mechanically what's important about the burqa ban is it just raises the salience of Muslims? And so, it raises the cost of being associated with Muslims in France and when the burqa... I mean, people know who is and isn't Muslim, for the most part. And if anything, wearing burqas makes it easier to identify who is and isn't a Muslim.

So, you might actually think burqa bans might actually go the other way. It actually makes it harder to distinguish between these groups and might, in the long run, lead to more assimilation and so forth. So, that seems like one reason why this is a strange example.

But it's also not obvious to me that... It seems like for the burqa ban to even get some political support, there already have to be lots of people who don't like this group or think of them as an outside group. Think about the wall, for example. Do we think the wall... Do we think Trump coming out and proposing a border wall, does that generally make all Americans less willing to be associated with Mexicans?

William Howell:

To the extent that it raises the salience of being from Central or South America, it's about the symbolic qualities of building a wall as to say, "You are on the other side," and who the "you" are is a social marker that then may have implications for hiring and firing decisions that are above and beyond the changes in the composition of the workforce that also may be affected by the introduction of a wall.

William Howell:

I think that's a story at least... That's the part of the story that Stephen and Torun are trying to grab ahold of.

Anthony Fowler:

Here's... Just out of curiosity, so here's one possible interpretation of what happens when Trump proposes the border wall. You can tell me how well does this fit with their paper or not.

Suppose what happens when Trump proposes the border wall is one, there are a bunch of liberal progressive white people who say, "I don't like that," and if anything, it might actually make them want to go hire Mexican Americans and want to help this discriminated-against group.

Another thing that can happen is there could be a bunch of people who already have discriminatory attitudes toward Mexicans, let's say, and people from Central America that are coming in through the Mexican border. And they now feel empowered to speak openly about those discriminatory views in a way they didn't before. When Obama was president and this wasn't out there, out in the open, they kept quiet. They kept it to themselves. And now, "Hey, Trump is our president. He's saying it on TV. I can say it, too, and let's go have the march. And I'm going to yell when I see someone I don't like on the street." I think that's a pretty common interpretation of what's going on right now in some segments of American society. Could that be fit into their story? Or is it a totally separate story?

Wioletta Dziuda:

I think you're absolutely right. I think one of the effects of these policies is to make people come out and say, "Yes, I actually don't like minorities," and so on. But I think the interesting insight that the paper gives us is as follows. If indeed those symbolic policies push us towards discriminatory equilibrium, if indeed minorities lose at the end, we might conclude that actually most of Americans are racist. We might say, "Well, if we weren't racist, we wouldn't discriminate."

But what they are saying is that perhaps, actually, the situation is not as bad as we think it is. Perhaps what Trump did with his wall is he woke up a few racists who express their racism explicitly.

But at the end of the day, we all fell in line because we felt that if we don't, we are going to suffer economically. I think what I like about this paper is that, yes, I think it doesn't disprove all these explanations that you are putting forward. And I agree with you that they probably are valid explanations. But I think what they tell us is that we don't really have to have a lot of racism, a lot of prejudice in society.

William Howell:

In order to be incredibly racist.

Wioletta Dziuda:

Exactly.

William Howell:

This is...

Wioletta Dziuda:

Exactly, in order. So, Trump doesn't have to believe that most Americans are racist. He just has to believe that some of them are and they are going to come out, and then they are going to snowball this effect. They are going to now make us even more attuned to racial differences.

William Howell:

Differences.

William Howell:

And...

Wioletta Dziuda:

Yeah.

William Howell:

And assign significance to those differences.

Wioletta Dziuda:

Yes.

William Howell:

And in this sense, it's not clear that this is a good... This is not a happy story at all, right?

Wioletta Dziuda:

No.

William Howell:

Right? This taste-based story would suggest that well, no, there are lots of people who want racist policies and therefore, to the extent that we want to combat the racist policies, we need to attend to this massive population. But what they're coming in and saying is that, "Well, no, no, all you need is a few of these people," right? Potentially to get it going, to have there be some sanctioning effect. And that can carry through the entire population, right?

Wioletta Dziuda:

Yes.

William Howell:

It's really... It's pretty awful actually.

William Howell:

Given that a symbolic policy is adopted, that they recognize that there could be a non-discrimination equilibrium or a discrimination equilibrium, but we have no sense of the likelihood of one or the other, or how we move from one to the other.

Wioletta Dziuda:

Yes.

William Howell:

So, how as a game theorist do you think about equilibrium selection?

Wioletta Dziuda:

So, that is actually one part of game theory where I don't think we have made a lot of progress for the last few decades. Once we are in equilibrium, we are in equilibrium it's optimal for me to behave the way I behave, and it's optimal for you to behave the way you behave. And unless one of us sacrifices themselves and changes our behavior, we are stuck. So, it's very hard to think about how we move from one place to another.

William Howell:

Or how do we begin playing one versus another?

Wioletta Dziuda:

Yes. We have some theories on that, but there's nothing that everyone would agree with. I think that's the biggest weakness of the paper, that they try to tell this story that there's this symbolic policy that moves us from one equilibrium to another. But I agree with Anthony, that that's just a story, that there's really nothing, no theory, no science behind it. But I want to point out that their paper goes a little bit beyond what we have discussed so far.

So, for example they say, "Look, if you buy our story for a second, that it's this symbolic policy that moves us to the other equilibrium, then what you should find is that politicians who propose those policies at the same time will also propose to lower redistribution, lower taxes, and lower governments' spending on social programs." If we believe that the symbolic policy is going to move us to this discriminatory equilibrium, the minorities will suffer but the majority will benefit economically.

Once there's higher income inequality, the majority does not see it useful to pay taxes because those taxes will not support whatever they want to get out of the government. But a big chunk of the taxes is going to go towards minorities. So, there isn’t much support for redistribution. We don't know whether those symbolic policies do change the equilibrium, but we do see, at least the authors have some examples, that politicians propose symbolic policies, and at the same time, propose lowering taxes and lowering redistribution. So, I find and suggest that there's something to their story. It's not a proof, but at least it's something worth thinking about.

William Howell:

Mm-hmm (affirmative).

Wioletta Dziuda:

I just want to bring your attention to another aspect of the paper. So, what they found is that this discriminatory equilibrium can occur only if the minority is sizable, but not too sizable, and if the minority is not well-integrated economically.

What that means is that we should see that there might be some minority firm owners, but not too many of them, okay? And one can see the logic of that, if there are too many firm owners that are minorities, then actually it's even harder for the majorities to find jobs if they cannot work for minorities. So, they are not going to find it profitable to shun the minority firms.

I think this actually lets us understand certain historic facts. So, for example, if you think about the discrimination against women or discrimination against African Americans, this discrimination persisted for a very long time, long after you would think that we got rid of most of the taste-based or statistical discrimination. And that would be consistent with their model. They say this equilibrium just exists, even if we don't have any statistical discrimination. When you think about other minorities that were historically discriminated against based on taste discrimination, such as LGBT community, for example, you see that actually the labor market discrimination against them disappeared very quickly.

I'm not saying that it disappeared completely, but you sort of see differences between African Americans, women, and LGBT discrimination in the workplace. And you can say, "Well, it's consistent with the model. They are a small minority, so shunning them out of the market is not really going to help the rest of us." At this time, they are economically integrated. They are firm owners. They are powerful people in entertainment or in politics. So, they say, "Yes, so this equilibrium is not going to benefit anyone, hence it's not going to be preyed. It's not going to be an equilibrium in the first place."

And again, I don't want to say that it explains the history of discrimination, but I think it's yet another point that makes me pause and say, "Yes, I share Anthony's concerns about different parts of this paper, but it's actually speaking to me in some sense." It's making me think about certain facts from our history differently.

William Howell:

It’s something of a trade-off, when we evaluate this kind of work, between on the one hand, being kind of clever and saying, "Look what outcomes I can explain by making a set of claims that have no bearing on the actual world," on the one hand. And on the other hand, the upside is to then reveal aspects of our world that we wouldn't otherwise see or to force us to question assumptions. There are times when looking at a model, you can say, "Look, I can explain the history of discrimination in which there is no racist and there are no racists whatsoever."

And I think one concern is that that's being too clever by half. If you want to get to the crux of the matter when you want to think about racism, and its effects that emanate through politics and the economy, the place to start is with the long legacies of bigotry that are rooted in ideas about what they call "taste-based" or statistical discrimination.

That's the place to start. And that this then can feel at times like... Well, again, I'm being a little too clever by half. Is that...

Anthony Fowler:

Yeah, I've got several layers of things to say, I think. So, one place to start is... One way to think of the model is just as a model of labor market discrimination and ignore the symbolic policies for a second.

You've got a model of how labor markets work, and do we think that's a sensible model or not? It's got some merits to it. It certainly seems to explain some empirical things that we see in the world. It also explains why maybe we used to have a lot of taste-based discrimination and we got into this bad equilibrium, and then it's very hard to get out of potentially. So, I think there's some value in the paper, just in doing that.

If you then want to take that model and say, "Now, I'm going to use that infrastructure to explain symbolic discriminatory policies..."

Anthony Fowler:

Yes. I actually don't think they've made a lot of progress there because that's all a black box. In their model, symbolic policies are just "suppose the government can magically push a button that increases the cost of being associated with a minority group." When would they do it and when would they not do it? Okay, fine.

William Howell:

How do you use the black box?

Anthony Fowler:

Well...

William Howell:

Because they identify particular conditions under which a majority in a population would support the adoption of...

Anthony Fowler:

Yes.

William Howell:

These symbolic policies.

Anthony Fowler:

But I'm not even sure if there is such a button that the government can push. Is it really true the government can just magically push a button that all of a sudden makes it more costly to be associated with this minority group? I'm skeptical that that's even true and by supposing that it's true, you've kind of proved too much. And so that isn't really an explanation for anything. That's just saying "if the government did have such a technology available to them, here are situations under which they would use it." That still doesn't really answer the question of "why does it work for Donald Trump to go race-baiting?" Or does it work? Or...

Wioletta Dziuda:

Or maybe it doesn't work.

Anthony Fowler:

Maybe it doesn't work. Yeah, I don't know.

Wioletta Dziuda:

I don't think they would insist that those policies work. I think what they would say is that now once we know that we can have this equilibrium, we can sustain this equilibrium that benefits majority, and majority of the population would like to play this equilibrium, then the governments will try.

Anthony Fowler:

It's hard to imagine governments doing this without there already being lots of animus in the population to begin with. But if there's no animus to begin with, it's hard to imagine the government just... the president just coming out and making an announcement or passing a policy that's all of a sudden going to...

William Howell:

But they give you a model that shows there are conditions under which majorities will want such policies. They will want it, not because of animus, but because they see them benefiting from the heightened salience of a social marker that they don't currently have and its implications for the labor market.

Wioletta Dziuda:

Part of the reason why it's hard for you to imagine that is because you have not... this model in your head before. So, before...

William Howell:

You need to sink in it. You need to latch yourself in the waters of...

Wioletta Dziuda:

No, but it's...

Anthony Fowler:

Okay, come back to...

Wioletta Dziuda:

But it's true because there was this very easy explanation available to us, which is people are just racist, we fell back on this explanation every time we saw these kinds of policies. And what they do is they come, and they say, "Well, let's pause for a second. We give you a different explanation. You might not buy it, but it at least should make you think twice before you jump to the conclusion that people are just racist and that's why we see burqa bans and walls being built."

They are not trying to tell us what's true and what's the right explanation. But I think they made me at least think a little bit more about the conclusions that I jump to when I see these symbolic policies.

William Howell:

But when you say that they're not trying to show what's true, I think what you mean by that is they aren't trying to provide the single, definitive explanation for why all countries, all the time adopt symbolic policies.

Wioletta Dziuda:

Yes, exactly.

William Howell:

That's not what they're doing.

Wioletta Dziuda:

Exactly.

William Howell:

But the paper is decidedly true in the sense that the propositions are mathematically correct. And there's no arguing with that.

Wioletta Dziuda:

That's correct, yes.

William Howell:

Thanks for listening to Not Another Politics Podcast.

Wioletta Dziuda:

Our show is a podcast from the Harris School of Public Policy, and it's produced by Matt Hodapp. Thanks for listening.