<p><em>Romesh Vaitilingam interviews Sonia Oreffice for Vox</em></p>
<p style="padding-top: 0px; padding-right: 0px; padding-bottom: 0.5em; padding-left: 0px; margin-top: 0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0px; "><em>April 2011</em></p>
<p style="padding-top: 0px; padding-right: 0px; padding-bottom: 0.5em; padding-left: 0px; margin-top: 0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0px; "><em>Transcription of a VoxEU audio interview [http://www.voxeu.org/index.php?q=node/6444]</em><b>&nbsp;</b>&nbsp;</p>
<p><b>Romesh Vaitilingam</b>: &nbsp;Welcome to Vox Talks, a series of audio interviews with leading economists from around the world. My name is Romesh Vaitilingam, and today's interview is with Sonia Oreffice, from the University of Alicante. Sonia and I met at the Royal Economic Society's annual conference at Royal Holloway, University of London, in April 2011, where she had presented a paper called &quot;Fatter Attraction: Anthropometric and Socioeconomic Matching on the Marriage Market.&quot; I began by asking her to explain what the paper was all about.</p>
<p><b>Professor Sonia Oreffice</b>: &nbsp;We wanted to analyze and think that people, when they match and marry, they consider several characteristics that their partner has. And specifically, not only socioeconomic characteristics, something related to money or income, but maybe, also, physical attractiveness may matter. And, in particular, we were interested also in seeing the tradeoffs between the two. So, if you are interested in both, so physical attractiveness matters when you look at the potential partner and also the income side, then can you trade off for the two? So, if you are not good at one, so if you are a little fatter, then can you still be considered a good partner if you can compensate with a high socioeconomic characteristic, or vice versa? So if we observe, then, in the real world, with scientific data, that this is what is going on.</p>
<p>The challenge varies, both from the data point of view and from a more economic theoretical framework, because here we are trying to deal with matching with multiple characteristics. And in a sense, that's a challenge, and this is one of the new frontiers of research in this area, to try to handle multiple characteristics at the same time.</p>
<p>And then, data‑wise, because here we have a data set that is nationally representative of the US population, so that's good. But, the challenge is that it's not easy to find a dataset where you have both the physical dimension and the socioeconomic dimension, because, for example, there is research in biology or medicine that looks at this type‑‑for example, physical characteristics, how people match in a family. Are they very similar or not?</p>
<p>But the point is that usually, then, you don't have the socioeconomic dimension. If you have good data of the socioeconomic dimension, most of the time you are missing completely the physical part. So here, at least, it's not the perfect data set, but it was comforting to see that we have a dataset that is nationally representative and that we have both types of information for both spouses, because what was important, what we know for the husband, we also know for the wife.</p>
<p><b>Romesh</b>: &nbsp;So, why don't you explain exactly how you're able to use this data to take the measures of, on the one hand, physical attractiveness, and on the other hand, socioeconomic attractiveness?</p>
<p><b>Sonia</b>: &nbsp;Yes. So, what we do is, for physical attractiveness, we use BMI, which is considered a proxy, one of the dimensions of physical attractiveness...</p>
<p><b>Romesh</b>: &nbsp;That's a measure of weight to height.</p>
<p><b>Sonia</b>: &nbsp;Yeah. So, it's basically a weight, weighted by height. So you should take into account both features. And in the literature, if you read biology or medical literature, it is considered a very good proxy for female physical attractiveness. For male physical attractiveness, it's still OK. It would be better to also have muscular mass or a waist‑to‑chest ratio, but still, it's considered an OK proxy. And then, for socioeconomic, there, since it's like an economics dataset, then, in that sense, we are fine, because we have education, we have wages, so we have plenty of information. So, first, we documented that there is positive sorting also in physical attractiveness. So fatter people marry fatter people. So a fatter woman marries a fatter man, relative to the average population. So, we can document it also, using like a dataset used in socioeconomic analysis. And then we find this compensation going on--that, indeed, it seems like it is possible to trade off a bad physical characteristic with a good socioeconomic characteristic.</p>
<p>And the interesting finding is that we find it also for men. So usually, the general view would be that physical attractiveness matters only for women, so when men look at a woman, they value physical attractiveness. But men do not need to be physically fit; they just need to care about the socioeconomic aspect.</p>
<p>Instead, here we find that these trade‑offs go on also for men. So it seems like, also, when a woman looks at a man, she also values physical attractiveness, in the sense that if the man is, let's say, fat, or much fatter than the average, then, to be considered a good partner, he should have a high socioeconomic status, meaning that physical attractiveness and BMI in the sense of a man matters. So it's not true that it's gender asymmetric.</p>
<p><b>Romesh</b>: &nbsp;And is the trade‑off pretty much the same, do you think? Will a man compensate by having higher earnings for being fatter in the same way that a woman would?</p>
<p><b>Sonia</b>: &nbsp;We don't get to that stage. But it's true, because we use sometimes different measures for men and women in terms of the socioeconomic measure. But if we use the same... So, for example, using education for both, as a measure of socioeconomic characteristic, I think we would still find that for women, it's a little bit higher. But I think here the main message is in terms of the qualitative result, that there is this trade‑off that we documented, and before it was not documented. And also, that it's happening for both genders.</p>
<p><b>Romesh</b>: &nbsp;Are you able to look at it, in a way, and say there's a different trade‑off depending on where you are in your marriage, how long you've been married, or if there are differences between generations?</p>
<p><b>Sonia</b>: &nbsp;That's a very good question. So, our framework is thought to consider people who have just gotten married, because we have a framework where basically we think of how the matches happen, so how people meet and decide to marry, and find what we would call the equilibrium in the marriage market. So, you're right. The key point here is looking at people who are recently married, because we want to see, at the start of the relationship, of a serious relationship, a marriage or a cohabitation, what are the characteristics. Because you're right. I mean, over time then things could evolve. And that is another research question, to look at, the dynamics of how people vary their behavior or their characteristics over time throughout marriage.</p>
<p><b>Romesh</b>: &nbsp;This is US data you're looking at. Do you think similar things would apply to European data?</p>
<p><b>Sonia</b>: &nbsp;No, no. Definitely. So here, we could also make a cross‑country analysis, or looking at other European countries. So, what is good is that we set up a theoretical framework with some predictions, in terms of these trade‑offs, and then this could be checked for any country. So with British data, actually, there is the British Household Panel, that has information on BMI and socioeconomic characteristics, or other household surveys data, or also a European cross‑country analysis.</p>
<p><b>Romesh</b>: &nbsp;Is there a kind of rule of thumb that you can draw from this, that someone might actually think, &quot;Well, I could put on a few extra pounds, as long as I earn a bit more, to keep my ability to catch the man or woman that I want?&quot;</p>
<p><b>Sonia</b>: &nbsp; No, here it's like a rule of thumb. So, by collaborative calculations with our figures, we would find that for a man, an extra kilo, you should pay to it like four dollars per week, of a workweek, for like working 40 hours a week, average weight/height. So, it's like four dollars per kilo.</p>
<p><b>Romesh</b>: &nbsp;And then, could we apply that to the royal wedding that's coming up?&nbsp;We say Prince William's estimated net worth is about $40 million, so how fat would that allow him to be and still get the girl?</p>
<p><b>Sonia</b>: &nbsp;No, no. There could be something going on in terms of trade‑offs. I mean, physical attractiveness, maybe it's not only BMI. It's also facial attractiveness or other dimensions. But, we could also expand it to think of--There is BMI. There is face. There could be, I don't know, intelligence, or something like that. So, these tradeoffs, money, and then other dimensions that maybe sometimes are hard to measure. But I would say, for example, for me, what was nice to work on this is that sometimes you look at couples, or what you see in life, or what your mom tells you about, like people she knows, and then sometimes it matches. So, what you find in terms of these tradeoffs, that sometimes there is something that compensates for some features that you see. Sometimes something is more observable or less observable, but sometimes you can tell.</p>
<p><b>Romesh</b>: &nbsp;You're looking at marriage markets, as economists call them.&nbsp;But, presumably, people's relative attractiveness also has an impact on their labor‑market outcomes, how much they earn and what kind of jobs they get.</p>
<p><b>Sonia</b>: &nbsp;Yeah. On this, there is already a lot of research. And it's actually an interesting area, basically, labor‑market discrimination, and especially looking at physical characteristics, so like obesity or other characteristics, not only physical‑‑for example, sexual orientation or other dimensions. We call it marriage market, but I would say, what we are trying to do is basically analyze what's happening in the formation of families, because basically the marriage market is how families form. So, there is a man and a woman, and then, OK, homosexual couples, but let's say traditionally a man and a woman, who get together and form a family. And this, from a society point of view, is very important, because then it can increase or decrease the inequality that there is in the society, meaning that the characteristics of that man and that woman, when they get together, will determine, basically, not only their well‑being throughout their life but especially the children's outcome.</p>
<p>There are already a lot of findings showing that what we call parental background has a lot of relevance in how the child develops. So, here, what we are emphasizing is the fact that in the formation of families, when a man and a woman marry is not random. So, there is a choice involved. We want to study how this choice goes on, and in particular, study how the characteristics of the two partners correlate, because the point is, most of the time, the more they will be similar, for example, then it could be that then this has strong effects on children.</p>
<p>We also have another project where we actually look at smoking. So, if we consider socioeconomic dimension and then the smoking status, and to see how people match. So, there it's more about health, but it's relevant because there is evidence that shows that if your parents are smokers, you are more likely to be a smoker.</p>
<p>So, here, we want to go to the onset of things, so to the formation of family, and document how related the characteristics are, because there could be like a reinforcing effect if only one of the two spouses has a characteristic. But then, what if both have that characteristic, and how strong? So, these issues about BMI or being fat or not, or smoking or not smoking. Then, of course, behind there can be health habits.</p>
<p>So, we believe that's important, because if we understand how the formation of the family happens, then this has long‑term implications, for the couple, per se, but most of all for future generations, because what we call the human capital formation happens mainly in the family. So, we want to see how the families form and what are the characteristics. And over time there can be variations, also how people match, and these can affect inequality in the society.</p>
<p><b>Romesh</b>: &nbsp;Well, I was going to say, this must have some kind of policy implications, some material that policymakers can take from this when they're thinking about anti‑obesity campaigns, or they're thinking about issues around social mobility, or they're thinking about things about children's education.</p>
<p><b>Sonia</b>: &nbsp;Yeah, yeah. No, exactly. So here, basically, in terms of these health behaviors... So, we are adding one dimension, that is also a part for like labor‑market discrimination or the others, that also look at, in the formation of family, there will be a given factor. So here, for example, we say that, OK, everything else being equal, the fatter you are, it will be harder for you to find a well‑fit partner. And this could mean that then it will be more challenging to have, for example, a very healthy child, or something like that. Or if you marry a smoker, and maybe you too are a smoker, then you should worry about the health of your child. So, that's in terms of formation of families. And that definitely has, also, concerns in terms of social mobility or inequality, because the characteristics of both partners do affect the characteristics of the children.</p>
<p>And what is interesting, and this has been found in several disciplines‑‑ in medicine, in biology, in economics, and sociology‑‑ that the matching, so how people marry, it's not random. So, it's clear that there is a choice behind. So, if we look at the data, I mean, in any possible data set you would have, you would find that there is a clear pattern of similarity of characteristics between the men and the women who marry. So, it's not random. So it seems like there is something going on in terms of preferences or constraints people have in terms of who they meet. So, that's the part that we are analyzing in these projects.</p>
<p><b>Romesh</b>: &nbsp;You've presented the paper here at the Royal Economic Society Conference. &nbsp;How did it go down? It seems the kind of subject where everyone will have a view, because they think about their own relationships and they think about their own weight and education.</p>
<p><b>Sonia</b>: &nbsp;Yeah. It was great, indeed. So there was a lot of feedback from the audience. So, I really enjoyed giving the presentation here. It was a very nice invited session. Yeah, you always get good feedback.</p>
<p><b>Romesh</b>: &nbsp;And where are you going next? Where are you taking this research next?</p>
<p><b>Sonia</b>: &nbsp;Next, so now we are looking at smoking. So, we have some theoretical challenges, because smoking is more of a discrete pattern, and then also the point is that people may differ on how they assess. So, if you are a smoker, you don't mind having a partner who smokes, maybe. But if you are not a smoker, then you would mind. But, then we are thinking maybe, also, at other characteristics, like race or other dimensions, that could affect formation of family. So, yeah, we hope we can go on in these topics. And actually, it's fun, also, for what you say, because you can relate it to many things that happen in your everyday life.</p>
<p><b>Romesh</b>: &nbsp;Sonia Oreffice, thank you very much.</p>
<p><b>Sonia</b>: &nbsp;Oh, thanks to you. Thank you.</p>
<p>&nbsp;</p>