Dondena Interviews

Few things are more demoralizing for mothers than the perception that their kids aren’t listening. But José Tavares of Nova School of Business and Economics visited Dondena in March with a message that they can breathe a sigh of relief. It seems, his data suggest, kids are indeed listening. Watching, too. And these cues they learn from their parents shape their world view 20 years into the future.


If somebody’s missing your talk today, tell me, how are they going to screw up the raising of their children because they will not realize when their children are watching them, and when their children are going to behave like them, and when they’re not.

It’s true. It’s interesting that you say this. This might be one of the few papers around that can teach parents something! Let’s see.


Dimmi tutto.

Economists have been looking at what mothers do and how this influences the next generation. Parents are believed to have a strong influence on children, as they interact very closely with children, and they pass on their values. Mothers especially, have been studied as a source of cultural change. What we attempt, in this paper, which is novel, is to identify both what mothers do, and what mothers think. The specific issue we examine is attitudes toward female work outside the home, in the labour market.


We use data from the 1980s, for the U.S., when a lot of women still did not work outside the home. Some women are working and they think it’s not the proper role of women—they are working because they need the cash, or because the husband is unemployed. And we have the opposite, we have women that are not working but they think they should. We look at the opinions of the mother when the children were 2 to 9, and then 20 years later we look at the opinions of those children. So we can identify what is the effect on children of working outside and thinking that is the proper role, all the combinations of what women do and what women think. We find an effect of both. So, children pay attention to what the mother is doing and what she thinks about what she’s doing. And once we control for several characteristics of the mother and the child, and we look at both sons and daughters, we find that the effect is particularly strong on daughters. So if you work outside as a mother and you think that is what you should do, this has an influence on your daughter that is the strongest. Is it the behaviour or is it the belief of the mother? There is evidence that it is both.



This seems like terrific stuff.

I’ve been working on topics related to gender, like the cost of gender discrimination and even the impact of gendered political quotas. It is one of the two or three big topics that I’m interested in. I find, with Tiago Cavalcanti from Cambridge, that huge economic losses are due to gender discrimination in the sense that economies are less wealthy than they could be because of gender discrimination. We mimic the U.S. economy, and then we put into the U.S. economy model the gender discrimination of countries like Saudi Arabia, Egypt, Ireland, so we can compute how much less GDP per capita output the U.S. economy would have with those levels of gender bias. For some economies, the difference is quite substantial, up to 40% of the output difference with the U.S. can be due to gender discrimination.


You’re kidding.

Some economies have only 10% or 20% of the women in the labour market. That is a direct impact: These women are not producing in the market. The second effect we study is the increase in fertility associated with women not in the market. These women tend to have more children, and this also lowers GDP per capita in the long run. So the take-away is that gender equality in the job market is probably one of the very important margins that economies should work on.


That’s interesting. When it comes to studying gender ... I’m not sure how to ask this question ... I understand that we haven’t known certain things up to this point, but you say GDP is lower because women are being discriminated against. Yeah, that makes perfect sense. Are there things in gender research that you’re finding out that aren’t so obvious, so much like “you should know this, that your GDP is lower because you discriminate against people”?

That’s a good point. I mean, the truly important innovation in our paper on the output cost of gender discrimination is that we quantify it. We put a figure to the cost of discrimination. And the cost is substantial.


Everybody wants a number, don’t they? Before they do something about it.

A number is almost a story, a way to convince people and policy-makers that it matters. A model, a number, a story. So, I agree, the result seems obvious, what we do is to produce a model and then neatly quantify its significance, making the issue clear. We have the U.S. economy, make it borrow the gender discrimination of, say, Egypt, and estimate the slump in the long-run? The economy could falter by 2% or by 40%. It goes down by 45% ...


To be clear, I’m not saying “well, you’re doing obvious research here.” What I’m saying is, I wonder, is there a dichotomy where this is how you gotta get in with policy makers. You gotta give them a number they want with something that makes perfect sense in academia. So you get their ear, and then you say “you know what else...” and you go to your next paper on something that maybe isn’t so obvious. So that’s my question: are we doing a lot of research on stuff to make headway and give people the  numbers they need on things that we know to be true, or are we doing more counterintuitive stuff? Finding more, “Ah, I wouldn’t have thought that!” things?

I think we’re doing a bit of both. A related example: I have a paper on the impact of political gender quotas that brings forth the obvious and then uncovers the less obvious, which is the really important. In current conversation many people claim that if you have quotas you will have women coming into politics that are less able and could not get in before, so that you end up with worse politicians. I thought this could just not be true, even in a very simple model of the world. This could only be part of the truth, the obvious part at that! So, with my co-author Paulo Júlio, we devise a simple world where you have both high and low-skilled women, and the same for men, and each decides whether to run for office. When we introduce a quota we can examine who runs for office and who gets elected, for different realities. In this very simple model, actually extended from work by Francesco Caselli and Massimo Morelli, we find that yes, there is a region where if you impose a quota it’s the low-quality women that have the incentive to run and quality suffers. The obvious, this is the region that people expect to find. But there may be other regions where if you introduce a quota, it is the high-skilled women who run for office, pushing away both low-quality men and low-quality women. This quota helps raise politicians’ quality. The non-obvious, that which could only be uncovered by a careful and comprehensive modelling of reality.


So when there’s a high-quality female candidate running the low-quality males and females say, ‘I would vote for her.”

Yes! More interesting: Imagine you have 10%, let’s say, of seats occupied by women in the status quo; now, if you increase the quota to 15%, it may happen that low-skilled women go in and quality goes down; but then increase the quota to 20% and the probability of a woman being elected has now increased enough that the high-skilled women go in and now the quality of the elected increases. So you have a non-linear effect of quotas on quality, something almost counter-intuitive. There may be other reasons to want more women elected. Diversity, for example. A role model effect. Here we just looked at the quality of the elected. High-skilled women have more opportunities in the private job market. If the quota is there but not high enough, they are not willing to play the lottery of as political candidacy, but if it’s high enough they say ‘Ok, I’m going’ and when they go, they push the low-skilled away. We start from a very consensual result that I thought could not be the whole story, and research and thorough modelling uncovers the other part of the story.

When I present this paper, people start obviously from their prejudice. A very strong prejudice. But then when the presentation ends, they say ‘Ok, maybe we need to think about this more,’ and this is one of the things we want to do with social science. Make people think more.


Tell me about that—what do you like about what you see going on in your field, or in social sciences broadly. What do you wish you could see more of as well?

What I love and I think this is the reason I conduct research, is sharing interesting ideas. The motivation to study economics is that you can use smart methods to address complex human problems. You will not solve them in their entirety, but you will contribute to their understanding. I’m not a fundamentalist who is thinking ‘Oh, if we have more and more economics we’ll solve the problems of the world.’ But I think it can improve the conversation. Economists need to be aware that people are not going to be ruled by or going to be understood fully through models and mathematics. But if we acknowledge mathematics is only part of the issue and you use it wisely, you can uncover the non-obvious, but important. This to me is really what makes me tick.


To learn more about José Tavares, click here.To learn more about the Dondena Seminar Series, click here.

Last updated 08 October 2015 - 14:06:37