Discrimination: Why Women Outperform Men in Congress

Kamala Harris could be the first woman to ever serve as president or vice president. Why are women so underrepresented in our government? 

In this episode, we discuss a paper from Professors Christopher Berry at the University of Chicago and Sarah Anzia at UC Berkeley that attempts to indirectly assess discrimination against women in the electoral process by testing whether the women who are elected perform better once in office. We discuss their study, alternative explanations of their findings, and implications for the 2020 presidential election and a potential Biden-Harris administration.

The show is hosted by three professors at the Harris School of Public Policy: William Howell, Anthony Fowler and Wioletta Dziuda.

Listen on Apple Podcasts or wherever you enjoy podcasts.

Transcript 

Will Howell:

I'm Will Howell.

Wioletta Dziuda:

I'm Wioletta Dziuda.

Anthony Fowler:

I'm Anthony Fowler and this is Not Another Politics Podcast.

Anthony Fowler:

My mom has been calling me regularly to find out what I think about the Democratic Convention. She's surprised that this bulk of scientists we're not glued to our television nonstop following. What do you guys think? Any thoughts?

Wioletta Dziuda:

What are you saying? We are glued to our television non-stop.

Will Howell:

I can't get enough.

Wioletta Dziuda:

What do you think I do?

Will Howell:

I have a thought—

Anthony Fowler:

Just cable news, 24/7.

Will Howell:

I think it's good when the alternative is balloon drops in convention halls, that's not good TV.

Anthony Fowler:

When you do turn on the cable news as we do every day, as political scientists, we see a lot of men. We see a lot of male politicians, about three and four members of Congress are men. Every president and every vice president in history has been male, but half the population is female.

Wioletta Dziuda:

I want to point out that when we interview the authors of the papers, we mostly interview men too.

Will Howell:

It's true.

Wioletta Dziuda:

It's true.

Anthony Fowler:

Something seems amiss. Something seems amiss in the sense that half the population is female, but not half of the politicians are female. A female has been nominated to run for vice president, which is, I think exciting news for a lot of people who really like Kamala Harris, and we'll get back to her in a minute. But what do we think about this underrepresentation of women in the political process? Do we have a whole lot of empirical evidence that speaks to that, as to why we have that?

Wioletta Dziuda:

Yeah. So if you were cynical, the simplest explanation for why there are so few women in politics would be, women are just not good at politics. So I talked to our colleague, Chris Berry, who has this paper with Sarah Anzia where they say, "Let's just see. Is there any truth to this very simplistic explanation?" They look at women in Congress and they find women in Congress perform better, and I'm going to say much better. We are going to go into numbers later. They perform much better than men on average.

Will Howell:

So let's give the interview a listen.

Wioletta Dziuda:

So Chris, you have a paper with Sarah Anzia, who is a Professor of Political Science at Berkeley, in which you're trying to tackle this question, shed some light on this question. Why do we have so few women in politics? Before we even ask this question and answer this question, you ask an auxiliary question. You say, let's look at the women who are in Congress and let's see how they do. Do they perform better than men? Can you tell me a little bit more about how you measure performance of women in Congress and what you find.

Christopher Berry:

In our paper we used two metrics of performance, which are meant to capture what we think of as the two main facets of the legislator's job, which are bringing home goodies for the district, for the constituents, what's broadly called constituent service and then also legislating or policymaking. So for the first of those, we look at federal project spending that comes to the district during the time of the legislator's term. For the second, we look at a bill sponsorship and co-sponsorship patterns.

Wioletta Dziuda:

What do you find?

Christopher Berry:

We find that women perform better on both of these metrics. So a district receives roughly 10% more federal project allocations during the time when a woman is representing it than a man. Women's sponsors significantly more bills than men. They also co-sponsor more bills and they garner more co-sponsors on their own bills than men do. Interestingly, they also co-sponsor with a wider network of other co-sponsors on their own bills.

Wioletta Dziuda:

So let's talk about possible explanations. My favorite explanation, the first one that comes to my mind, of course, is women are better than men. As far as I remember, that's not the explanation that you suggest in your paper. So can you tell me a little bit about your explanation and why you think that's the right explanation?

Christopher Berry:

Sure. We suggest that women are better because of the process by which women are selected into politics in the first place, because there's discrimination in politics, because the voters discriminate against women, the women who win office have to be better than the men that they ran against.

This is why the paper is titled, The Jackie and Jill Robinson Effect. We are obviously making an analogy to Jackie Robinson, the first African American player in Major League Baseball. Part of the story there was he had to be so much better than the next best white player in order for owners and fans and other players to have accepted him into baseball in the first place. We're making a similar argument about women in politics.

There's a second explanation, more of a psychological explanation, and this is associated with Jen Lawless, who's one of the leading scholars of women in politics. She suggests what she calls a perception gap between men and women, which is just that women underestimate their own qualifications for office. She has reached this conclusion through surveys that she's done, where she asked them if they've ever thought of running for office before or whether they believe they're qualified or not. Women have the same objective qualifications systematically consider themselves to be less qualified than men do.

That would be another mechanism that would make the women who get elected of higher quality. That is, if women systematically underestimate their qualifications, then those that do run are going to be more qualified and those that get elected will be of higher quality.

In our paper, we don't take a position on these two particular mechanisms, but we do suggest it that either one of them would result in the findings that we get.

Wioletta Dziuda:

But you have some additional evidence that I think speaks to which of these two phenomena all care or which one's more likely or that they both all care. So the first one is, you ask a question on whether women perform better if they're elected in districts in which we would expect is anti, higher degree of gender discrimination. Can you talk to me about this a little bit?

Christopher Berry:

Yes, and I should say, you're forcing me to put my cards on the table, which is great and I should. I personally lean towards the interpretation that discrimination is the more powerful mechanism here. One of the things we do is we have another measure of voter discrimination that we use, which is just the conservativeness of the district. We find that women who come from these more conservative districts perform even better.

In other words, there's a correlation in the performance of women and the conservativeness of the district where the more conservative the district was that the woman was elected from, even better that she performs. That's certainly prediction of the discrimination mechanism that we talk about.

Now, I think one could say that if you really believe that there's this psychological mechanism going on, where women underestimate their own qualifications, then you might argue that women who are coming from these more conservative districts are more subject to that psychological mechanism internally. Maybe they've been raised in an environment that's more conservative telling them that women are less qualified. So maybe that mechanism is also stronger in districts that are more conservative.

I happen to lean with you on this particular one that this seems like very clear evidence that the discrimination mechanism is at work, but it could be interpreted as being consistent with the self-selection on quality mechanism too.

Wioletta Dziuda:

Okay. Fair enough. This is support of the joint story that you're proposing, but it's hard to use this piece of evidence to differentiate between those two. What about a different thing that you do in the paper, which is, you look at women who became Congress women because they succeeded their husbands in Congress. You compare them to women who just became elected in a competitive process and you find something interesting there. So can you walk me through that?

Christopher Berry:

Sure. It is true, particularly in the early years of women's emergence in politics and in Congress, a very large proportion of the women that were elected to Congress were elected to succeed their husbands who had occupied the seat and died in office. So often the widow would either be appointed or run for office and win. So we look at the performance of those women, the women who are the widows of Congressmen and ask whether they perform as well, and we find that they do not.

The logic there, we think, is that those women didn't have to overcome the discrimination that a new de novo female candidate would have to overcome in order to win office. So if it really is this discrimination mechanism that accounts for the difference in performance, we wouldn't expect to see that for widows. We don't find it for widows. So that is another piece of evidence for the discrimination mechanism.

Again, if I were going to play devil's advocate, I would point out that you might think also that widows are less subject to the psychological mechanism that leads them to underestimate their own qualifications. So you can tell a story where that's also consistent with that mechanism again.

Wioletta Dziuda:

Interesting. So let's talk a little bit about alternative explanation. So let's try to put ourselves in the shoes of someone who doesn't really buy your finding that women actually perform better in Congress. So let me try a few things. So the first one is, of course, you are comparing women to men. Perhaps women just happen to be elected in districts which receive more money from the federal government. I can tell many stories for why that would be, what if that's what you're seeing. They are just not better, they just happen to be in better district.

Christopher Berry:

I think that's the most obvious alternative interpretation to our findings is because it is true that women are not selected at random in districts and women do represent districts that are different on average from the districts represented by men. So, for example, women are more likely to come from districts that are more urban, have a higher poverty rate, have different racial composition on average than the districts that elect men. So you might think, "That's the reason why those districts are getting more money, not because of the person they elected, but because of features of the district that just make them more eligible to attract federal money."

That's why our research design is based on, let's just say, within district analysis where our comparisons are really coming from the same district that has been at different years represented by a man versus a woman and we compare the amount of money flowing to that district in the years represented by a woman versus a man. That way, because it's within the same district, it can't be the fact that women represent different districts that would explain these results. Because really here, we're just talking about how much money does the same district get in years when it has a woman versus a man in office.

We restrict those comparisons to within the course of a decade. So it's also not the case that the district could have changed radically between the time it had elected a man versus a woman. These are relatively small blocks of time of which we're making these comparisons.

Wioletta Dziuda:

Is it true that a majority of women who are in Congress are coming from the Democratic Party?

Christopher Berry:

Yeah, absolutely. There's certainly been years during which virtually all the women in Congress are Democrats. That's changing a little bit, we're starting to see some emergence of more female Republican candidates. So there's still a huge majority of them are Democrats.

So that might be another concern that you would have with these results, which is, in an extreme example, if all Republicans were men and all Democrats for women, then the change from a man to a woman would also imply a change in party. You might think that Democrats, just due to that party's platform, are more likely to want to bring home, for example, federal projects pending and more likely to want to pass legislation.

We can control for party. Again, when I'm talking about within district, we can look at a district that's represented in both of the periods by a Democrat, but one of those Democrats is a man and the other is a woman. We still find these results. So I think that's another pretty obvious concern that people would have knowing that women are more likely to be Democrats.

Wioletta Dziuda:

So is there anything in particular you think we haven't covered that you would like to mention? I just have one more question. These were the issues I wanted to cover.

Christopher Berry:

Yeah, I think one interesting thing, when the paper first came out and one of the common misunderstandings that I think a lot of reporters had that we were claiming that women are inherently better, kind of your first explanation that you led with I think maybe jokingly was that we had found that women are somehow inherently better.

Wioletta Dziuda:

I wasn't joking.

Christopher Berry:

Okay. But anyway, I had a number of reporters that said, "Oh, so what your paper is telling us is that we should all go out and vote for women because they're better and they're going to provide better representation for us." In fact, our prediction will be just the opposite of that. If everyone decided that women were better and went out to, in other words, developed a bias in favor of women, we would actually expect their quality advantage to go away. In our account, women are better precisely because they had to overcome discrimination. So if that discrimination were to go away or to be reversed so that people favored women, we would get exactly the opposite prediction.

I have a perverse way. One of the marks of the success of women in politics will be when our finding goes away, when it's no longer the case that the women elected to office perform better than men, that would actually be a sign that things have become equal and they no longer had to overcome this discrimination to begin with.

Will Howell:

Terrific interview. I think we need to address three issues. The first is that, we need to interrogate the particular explanations for why we see the effects that we see. Why it is that women are performing better than men in Congress. Second, I think we should take some time to reflect upon Kamala Harris's selection as Biden's VP and what Anzia and Berry's findings have to say about the significance of Harris's selection.

Third, let's take some time to reflect a little bit more generally about the relevance of discrimination as a political phenomenon in affecting different people's ability to perform better or worse, to think beyond simply issues involving sex, and sex and race, to think more generally about how discrimination in selection bares upon performance.

Wioletta Dziuda:

I think it's very important that we explain why they even started thinking about this explanation. So if you just look at their result, the first big empirical result that women perform better than men, that the first obvious explanation that we should jump at is women are just better.

So it's worth to point out that the reason why we are searching for some other explanation on this most straightforward one is that we have other pieces of evidence that they put forward in their paper that leads us to question this explanation as the most obvious one. So those two pieces of explanation is the first one that we mentioned at the beginning. There are so few women in politics. If they were so much better than you would expect that there would be more of them.

The second piece of information is that, if you take two incumbents, one male, and one female, they actually, when they run, they are equally likely to win. So it seems to be that conditional on running for office, women are equally likely to be elected. So they just started to think, how can we reconcile those three pieces of information that we have into a coherent story.

The story of just women being better explains why women perform better in Congress, but doesn't explain why we have so few of them. Why isn't that the case that voters actually are more likely to elect a woman expecting that she will perform better.

Will Howell:

Some of the findings that they put forward are incompatible with this essentialist claim. So for instance, the fact that women from particularly conservative regions of the country disproportionately do better, unless you believe that somehow in conservative parts of the country, the women all have to higher quality stock. Why would we believe that?

But what they do suggest and what we could see is that they're held to much higher standards and that the discrimination against women is even greater in those domains. So the hurdle that they have to clear is even higher, which is then pointing towards this selection mechanism is being crucial.

I think my favorite cut of the data is when they look at the widows of male legislators, who for a good portion of congressional history would take over the seat of their husbands when they died. What they find is that widows don't perform any better along the metrics that they offer.

Again, if you had just a straight essentialist argument, you'd say, "Well, women are women”. So they should be performing higher regardless of what their pathway was to securing the seat. But precisely because the pathway for a widow to reclaim the seat of the husband who dies is smoother than it is for a woman running from the outside. It suggests that she doesn't face the discrimination that many women do face. Therefore, we wouldn't expect her to be of higher quality. She doesn't have to clear that threshold, even though you what you might say is that by virtue of being the spouse of a legislator, you're privy to all insider knowledge, you're given certain advantages that would directly bear upon your ability to do your job well.

But the discrimination that they're talking about is in the background of the analysis. They don't go out and measure it directly. They don't have a score for each legislator that characterizes how much discrimination she/he faced and then looks for heterogeneity and the effects is a function of that observed phenomenon.

What they say is, is that, "Well, if you believe that women from conservative districts face higher levels of discrimination, or if you believe that widows face lowers levels of discrimination that has implications for the size of the effects that we have to observe." Sure enough, that carries through in ways that makes sense, but they don't directly observe the discrimination.

I think it's something of an open question too, about when we say discrimination, where is that discrimination? Is this a story about voters? It's about voters who look more skeptically upon women who run for office? Is it a story about donors and their willingness to get behind women versus men or never mind their own individual views about the leadership qualities of men and women? Is it about networks and structures that advantage men versus women? We don't know. I think there's just an abundance of possibilities here, but none of this has pinned down for us.

Wioletta Dziuda:

Yeah. I think I would like to ponder over that a little bit more. So I like their story, but the way we talk about it, it seems like they are these vicious voter, they are these bias voters, bias lobbyist, there's this evil thing that's preventing women from being elected to office.

It's possible that there is, but there's also a little bit more nuance or like more innocent explanation, which is, for so many years, we've lived in a patriarchy. As a result, it might be that we are all conditioned, we men and women, everyone in the society is conditioned to think about women as less skilled given the same set of characteristics.

So we look at a man and a woman with the same education and the same experience, and there's something in us that says, "Well, but she's a woman. She probably is not as," and here you put your favorite adjective as ambitious, as skilled, as a trailblazer, as a man would be. It might be completely innocent. We're just bad at judging the quality of women because they behave differently.

There are a lot of stories in psychology that, for example, when we see someone who is loud, who interrupt other people, we say, "Oh, this person is confident. This person must be actually good." When we see someone a little bit more quiet, we say, "This person is probably not skilled. They're probably not really competent." Even though they're the people who might be actually as competent or it might be reversed. There's a lot of things that we are all bad at judging women's quality, given the external characteristics.

That story is a little bit more innocent and it could explain their data perfectly. When we see a woman running, we perceive her as worse. So we don't elect her so frequently. Women actually look at themselves and they say, "Well, I'm not as high of a quality as some of my male colleagues, I'm not going to run." As a result, you see only the highest quality women who cannot actually deny that they're so high quality running for office.

So the outcome is the same, but the story is a little bit more nuanced. It's not a story of like the sinister discrimination, "I don't like women. I don't want to have them around. I don't want to see that in politics. I'm not going to vote for them." It's more like, "I would like to have a woman as much as I would like to have a man. It's just I look at this woman and she doesn't seem competent." So it seems to be this implicit bias as opposed to this explicit bias.

But there's a second difference between what I'm saying and what they are saying is that, in their model, in their story, women judge themselves correctly. They realize how skilled they are and they decide not to run because they recognize they will be discriminated against. In my story, when a woman decides not to run, the reason why she decides not to run is because she herself thinks she's not good enough to be a politician. She doesn't believe she will be elected. Not because she'll be discriminated against, but because she doesn't have enough of ability.

Will Howell:

Yeah, and I guess this is where I want to underscore the ambiguity about the phenomenon that we're describing. It seems to me, you could have the discrimination work outside of any story about psychology or any story about the individual opposition of any individual to get behind a female candidate.

It's that the networks of fundraising are built around men or the networks of endorsements and personal connections are built around men. That even if you had perfectly well-meaning men who occupy those positions of power, that leads to this support that provide the support you need in order to succeed in a campaign, women to break into those structures need to be of higher quality. That's what drives it.

Anthony Fowler:

Right. But it could be that a lot of women want to run and they go talk to some people and they try to get the signatures and they can't get them. They can't get the money, they can't get the endorsements. We never hear about them because they had a failed campaign before they even appeared in any public database or anything like that.

Wioletta Dziuda:

Yeah, I think you're right that we should definitely think about networks and about what happens before a woman even shows up on a ticket. These are the conversations that we have very frequently in academia because we have a similar problem in many departments, in many fields, we have many fewer women than men. Then there's this question? Why is this the case? Yes, you can talk about explicit discrimination, but a lot of people recognize that networking is very important. You go for a beer with someone after a conference, you go golfing with someone, and then you start talking about—

Anthony Fowler:

If you're invited to the golf attic, just to be clear.

Wioletta Dziuda:

Let see how this all plays out. So yeah, you start talking. This maybe sounds sinister. It's like, "Okay, we are just forming connection and that's why some people get promoted," but it's also less sinister. We start golfing and we start talking about research, and then you say, "Oh, actually I like your research. I think it would be a great addition to our department."

So at least there's a lot of narrative in where we live, the world we occupy, that networks, the structure, the way we built this entire promotion machine in politics or in academia is discriminating against women. Completely without an intention, just as a byproduct of us living in the world of men for such a long time.

I just want to say one more thing for all the women out there. We started this conversation by saying that perhaps women are just better than men. That's why we see them performing better. Then we went through this entire story. We know that this cannot be the explanation for everything that we've seen today, but I think we have to be very clear that nothing, in what we've discussed so far, and nothing in the data says that actually women are not better than the men.

That their story can be at work. It's possible to work, it's very plausible. But at the end of the day, we don't know what would happen if you removed selection and discrimination or any of the story that we put forward. The question is still out there. We actually don't know what the difference between genders is.

Will Howell:

That's right.

Anthony Fowler:

That's true. Although you wouldn't be spending so much time on this topic if they had found that men are better than women. We wouldn't be saying, "Well, maybe men really are better than women." I think this is actually a tricky challenge in any test of discrimination. There're results that are more politically desirable than others. But usually, our first inclination isn't to think, "Oh, there must be some big biological difference that explains this phenomenon."

Wioletta Dziuda:

I'm not saying that's my default. I just want to put it out there that there's no piece of evidence so far put forward that there are no differences. I completely agree with you, especially as a researcher, we should start with the presumption of, there are no differences and see what else is happening in the world. But I just wanted to put it forward especially because we have a VP, that's a female. I would like to think the best thing about the future of this candidate and the future of this country.

Anthony Fowler:

So let's talk about Kamala Harris and let's talk about what some of these different explanations might imply potential for a female vice president. Kamala Harris, if you buy this story has throughout her political career faced a higher burden than a male candidate would have otherwise faced.

If I just told you that someone was Attorney General of California and a US Senator and had the following qualifications, when I also tell you that person is a female, and you've read and understood this Anzia and Berry story, you should say, "Oh, they must be especially high quality," not because they're a woman per se and there're biological differences between men and women, but because she must have faced higher hurdles along the way.

So she must be really competent, really smart, really ambitious, really outgoing, really creative, et cetera. She has got something going for her that makes voters support her despite the fact that she's a woman. Do we buy that story with Kamala Harris?

Will Howell:

Well, I think there's this other dimension, which is that, what's significant about Kamala Harris is not just that she's a woman, she's a woman of color. I think we need to have some accounting of how race also can serve in all the ways we've just described, as another source of discrimination, another impediment to one's ability to get traction on a national level.

So then, to the extent that race interacted with sex should lead us, if we believe the Anzia and Berry story, to expect Kamala Harris to just be ready to kick some ass. There is this one piece though that I think complicates the story, which is that her parents are from Jamaica and India. So they're both immigrants. They both also, it's worth noting, have PhDs. But put that aside for the moment, they're both immigrants.

So the work that her race and immigrant status is doing is not just about erecting barriers. It may also be informing the character that she has and the qualities that she brings to bear and the ways that her parents raised her and that those things too may maybe contributing to her success in important ways. It's not just in spite of her immigrant status that she's able to overcome barriers that are put before her, but because of it, that imbued her with a strength of character.

Anthony Fowler:

It's a complicated story, because if you think, there may be on some dimension she's disadvantaged, on other dimensions maybe she's very privileged. She did have two parents with PhDs and not everybody has that. So maybe some poor white man from Appalachian who'd had no economic or educational opportunities. They also would have to overcome a ton when they become US Senator, and the Anzia and Berry story would apply to that person as well.

So in some dimensions, Kamala Harris has privileged, in some dimension she's underprivileged and on net, how does that shake out for how we think of her ability? It's a little bit hard to say.

Will Howell:

I don't know. I don't want to get into an essentialist argument. I don't want to do that. I think it feels like we flattened out the phenomenon, but all we say is that, depending upon which set of boxes you check you, face higher or lower levels of discrimination. The significance of those boxes is just about discrimination.

But now I'm drifting into Wioletta Dziuda land and saying, maybe this sounds a little bit like women are better. Women are better. That's probably are making an essentialist argument. So I guess I'm feeling—

Wioletta Dziuda:

Uncomfortable making this argument?

Will Howell:

Well, I'm uncomfortable trying to sort out what the significance is of the descriptive characteristic they're identifying. I want to say, it's more than just the implications that has for discrimination in selection. It's a much bigger significance than just that, like to say, "I expect Kamala Harris to be of higher quality because I expect her to have faced a lot more discrimination than a white man," is to give short shrift to the significance of her family background and some of the strengths that she draws from it and the learning that came along with it and the identity that came along with it. It's not a, "Well, I'm just like a white man, but I'm just of higher quality, and I know that because I had to clear a higher hurdle."

Wioletta Dziuda:

But I think you're inverting the story a little bit. We have not been saying all along that it's because women face discrimination, they are of higher quality. It's just some women happen to be of higher quality, that happens because they have very supportive parents, very educated parents, they had great education and only those women we are going to see actually succeeding at the top of political selection process.

It's not that because she faced all these hurdles, she decided to be a better quality. She was just of a better quality because of her background, her IQ, and other characteristics. Just for us, because we observe her in politics, we observe her having been successful. That's a clue for us that she's one of those. Otherwise, she wouldn't have succeeded as much as she did. Yeah, it's discrimination that made her. No, she made herself. But we only see her—

Will Howell:

Yes, and she needs her—

Wioletta Dziuda:

... and the fact that we see her tells us that she made herself and she's great.

Will Howell:

There we are. But that she make herself and her family making herself and her community making herself, is itself imbued with the force of sex and race that also bears upon the levels of discrimination that she faces and the hurdles that she has to clear. So in that sense, the relevance of sex and race is not just confined to the establishment of higher hurdles. It bears upon and informs the individuals themselves who are coming forward. No? Yes.

Wioletta Dziuda:

Yes.

Anthony Fowler:

I want to turn from one controversial conversation to another. There is one interesting aspect of this particular presidential campaign that is unusual and it's unprecedented, which is that Joe Biden, when he was running early in the primary, he pre-announced that if he won the nomination that his vice presidential pick would be a woman. It was even understood that it would probably be a woman of color.

One interesting aspect of that is that although all women of color who have made it through the political process, probably have a lot, you can say, about their quality and about their competence and their leadership and so on. The fact that Kamala Harris was selected as the vice president to the nomination, doesn't give you a whole lot of information over and above that because Joe Biden already had said he was going to select her.

You could even imagine Donald Trump and Trump's allies using that argument against Kamala Harris by saying, "Look, she wasn't actually the best person. She was just the best female," or "She was just the best woman of color." Could that actually end up backfiring against her? This is somewhat related to what Wioletta and Chris talked about at one point in the interview, which is, "If at some point voters shifted the discrimination in the other direction. If actually voters, all of a sudden started preferring women, then that would actually reverse this finding and then on that then."

So you can see how all of this plays out. Is that relevant at all? So one question is, was it a mistake of Joe Biden to pre-announce that he would select a woman because that changes this whole selection story. Will Donald Trump use this against Kamala Harris? Does this change the way we think about having a female vice presidential candidate?

Wioletta Dziuda:

I was actually about to ask a different question, which goes with your question very well. My question would be, why did Biden decide to select a woman? Why did he decide to announce he's going to select a woman? Why did he actually select a woman? I think it—

Will Howell:

There's no great mystery to this, right?

Wioletta Dziuda:

No, but—

Will Howell:

He's a 78-year-old white man. We start out with a pool of candidates who are all seeking the Democratic nomination, who are much more diverse among both sex and racial lines, and we settle on a 78-year-old establishment white man who's going to be the head of a party that is a great deal more diverse, and takes a good deal of pride in that. It was smart of him politically to commit to diversifying the ticket.

Wioletta Dziuda:

But was it after all voters in the primary selected an old white man, even though we had an amazing lineup of minority candidates?

Will Howell:

Barely.

Wioletta Dziuda:

Is it really true? Because if it's true, then the Anzia/Berry story is no longer true?

Will Howell:

I'm saying that it was good for him in the primaries wherein he needed to signal that he got the significance of race and of gender to people who were concerned about the democratic party putting forward an all-white ticket. Now, how that translates into his fortunes in the general election, I think is less clear.

This is where we've got things cutting both ways. On the one hand, the Anzia/Berry story carries through that, in fact, Kamala Harris is on average of higher quality by virtue of the challenges that she's faced. And there's discrimination in the electorate and in the political process, which will push back the other way.

There will be discrimination against Kamala Harris and people who will say, "I like Biden, but you know, he's old. Then we could have a President Harris, and I am not prepared for a woman of color to be the president of this country. So I'm going to think twice about a Biden-Harris ticket." Those people, I don't know how many of them are, but they probably exist out there. That's a strike against that choice.

Wioletta Dziuda:

Biden's campaign must have thought exactly about this issue. That even if you think about selecting a woman of color, it's not obvious to me that there is no gender discrimination within the black community and all the same way that there is within the broader society. So on the one hand, because she's black, she's going to bring African American voters. But on the other hand, maybe she's only going to bring female African American voters and female African Americans might also have this implicit biases that everyone has against women.

So he didn't have to go for woman of color, even after the announcement, he could have gone with Elizabeth Warren. He could have gone with some other candidate. So I think that they must have thought about all this, and they must have concluded that Kamala brings enough to the table to overcome these issues that she's going to face. Let's see. I don't know. I don't know Kamala too much—

Will Howell:

That's right.

Wioletta Dziuda:

I haven't been following her career too much, but at least you hope that that gives me hope.

Will Howell:

Yeah, to build off of the story you just told is that, while bringing on a female vice president, who is a person of color, may reduce our eventual win margin. The upside is that we're going to have a person of higher caliber. What I'm really concerned about is what life is going to look like when we cover. I'm willing to make that trade off.

Wioletta Dziuda:

Well, I'm confident that once the public gets to know her and gets to see her during the debates and on the campaign trail from commitment—

Will Howell:

That's right. Yeah, that couples that.

Wioletta Dziuda:

... they'll realize, "Well, yes, maybe I don't like women or women of color, but—

Will Howell:

But boy, this was—

Wioletta Dziuda:

... I think it was amazing and I have to overcome my biases or whatever I have.

Will Howell:

Okay. So can we take another step back? If we take Anzia and Berry story seriously, what it suggests is that any time that there are impediments for anybody to move forward and that we see somebody who has the ascriptive characteristics that are being discriminated against, we see one of those people persevere. We should expect them to be of higher quality.

If we take that story on the road, doesn't that mean that when we are in the hospital and a male nurse comes in, and if we believe that there's a system that... and I don't know that this is true, but let's just give it to me for now. That men have a harder time becoming nurses because there're expectations that we need people to be more nurturing and their set of networks that privilege women.

I don't know that this is true, but let's posit that is true for the moment. But then, if I see a male nurse coming in, that I should take heart in that fact, because they're going against type and they've overcome the discrimination. No?

Anthony Fowler:

I think that's right. I think that's absolutely right. It means that potentially we can apply the same logic to look for indirect evidence of discrimination in other settings, but it is very complicated. It's complicated for a few reasons. One is, it's hard to come up with really good objective measures of performance. Maybe we can even quibble with is co-sponsorship and distribution of federal spending, are those really good measures of performance.

It could be there's even discrimination at that stage. So maybe you get less federal money for whatever reason. You might've actually said, "Maybe that's because the male legislators are discriminating against a few," and there's also multiple dimensions that we really care about. Maybe if we elect a veteran, we're getting someone who really looks out for veterans, but we're getting someone who's lower quality on some other dimensions.

Yes, I think all of these things make this very complicated to apply generally and even to apply in the realm of politics. But I think the story is absolutely right, that we should constantly be thinking about this classic selection problem of who was able to make the achievements they did conditional and the discrimination they face and what inferences can we draw about them because of that.

Will Howell:

This works both ways. This, in some instances, is that you're going against type or that you clearly faced or infer that you faced some discrimination before you entered this space, got this job. That's going to inform my expectations about how well you're going to perform.

It could also be that though there's some people who walk into the position and to the extent that, you mentioned veterans. If you believe that the significance of being a veteran doesn't have any direct effect on quality. By virtue of having served the country, that doesn't make me a better legislator per se.

Anthony Fowler:

But it could, it's possible that it could.

Will Howell:

It's possible that it could, for sure, but that it is an advantage for me, electorally, because I can say I sacrificed on behalf of our country and that becomes a sign to voters about my commitment to country and the values that I hold. Then, if this is true, we might expect that veterans are of lower quality on average than non-veterans because that's an advantaged ascriptive characteristic.

In that sense again, they have a general story. The general story in that it carries over beyond men versus women.

Wioletta Dziuda:

I just want to point out that there's an issue that Anthony already mentioned that, it's very hard to measure quality and also, it's not obvious which measures we should use as a measure of quality. What might be happening in the real world is, we have certain groups that are advantaged and then we start conflating their behavior and their performance with success, with measure of quality.

Because this is a big question that we haven't even addressed if women indeed perform better in Congress. Why isn't it that the voters learn? We've had many years over which women were in Congress, their voters should have observed that, they should have seen that they are better, and we should have already changed our attitudes.

I think the reason why voters are not learning, at least, one of the reasons is that again, we don't know how to measure success. We're just used to successful men and successful men behave certain way. We just say, "That must be the measure of success. That must be the measure of quality." "This person is successful, hence he must be good. Hence, what I see what he's doing must be a measure of quality." I see this woman and she's not behaving the same way, and I'm saying, "Well, perhaps she's not of high quality."

Will Howell:

For sure. We make the wrong inferences. The lesson of this paper is, we make the wrong inferences in lots of ways. And we've seen to—

Anthony Fowler:

One relevant piece of evidence is that voters appear to... they do update their beliefs about female candidates. So Flare and Goku have this paper about female mayors. Their finding is that the sophomore surge, which is essentially just the number of votes you get the second time you run for election, compared to the first time that you won. Women have higher sophomore surges than male candidates do, which is consistent with the story.

Maybe there're other explanations for that as well. But it could be that voters, initially, are for whatever reason, skeptical or discriminatory against women. But then once they elect them, they realize, "Actually, this person's a pretty good mayor," and they vote for them. They could increase their votes for them more than they would for a male candidate.

Wioletta Dziuda:

But they don't seem to update about women, in general, they seem to update—

Anthony Fowler:

Yeah, that's right. The next woman that comes along, right?

Wioletta Dziuda:

... which is a little bit strange, which might be consistent with the story I was trying to say, that I see. I'm actually having a pretty good life under this mayor. So she must be a good mayor. But when I'm evaluating the next woman, I don't have experience with living under this woman as a mayor. So I'm looking at the measures of quality that I'm used to, which is, "How do you speak? Are you loud? Are you super confident in their opinions or their expressions? Do you see the world in black and whites?"

Then again, I fall back on my biases that I have developed, that most of the people that I've seen that were successful were actually extremely eloquent, and they spoke in black and whites, and they spoke over other people. Well, and then I draw my conclusions.

Will Howell:

So what's your bottom-line?

Anthony Fowler:

All right. My bottom-line is, I like this paper a lot. It presents really interesting evidence on a question that is very hard to study directly. You can't just go, fill the survey, and ask people if they're discriminatory towards women. You're not going to get very reliable answers from that. Either they won't tell you the truth or maybe they don't even know themselves. So they're coming up with a very clever way to study a very challenging question.

So I come away convinced by the story, and I come away looking for more signs of it and more ways to think about it, in the context of modern American politics. Count me in, in favor of this paper.

Wioletta Dziuda:

So I also like the paper, a lot. First of all, it shows us that women in politics do quite well, performed quite well, and I didn't have any prior or I didn't really know of papers who carefully study how women do, and this is going to be news to a lot of people.

That's already a great contribution. I also think that by their explanation for that, although, as we mentioned, there might be a lot of other things that are going on. The women might face networking hurdles, they might face higher costs. But at the end of the day, they made a pretty compelling argument that those two stories of discrimination selection seem to be needed to reconcile all the findings from their paper and some other papers. Those findings are very robust.

So I think it gives us a pretty robust way to think about why we have so few women, and what we should see moving forward.

Will Howell:

All right. So we're three for three. It's a terrific paper. We didn't talk much about their research design, but they have a really compelling research designed to back out what the relevance is of electing, at the margin, a man versus a woman to Congress. What expectations flow in terms of their performance in office. There are lots of ways of measuring quality and performance, and success, and they put forward some, and the ones that they do seem solid.

The last thing I'll say is that, it does a remarkable job of underscoring the significance of discrimination for political outcomes, without ever directly measuring discrimination. They don't have data on this, and yet they have a set of findings, as Wiola points out, that can only be or primarily be. The best available explanations point towards discrimination for trying to make sense of the pattern of findings that they put forward.

Anthony Fowler:

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.