Fundación de Estudios de Economía Aplicada

Anthropometry and Socioeconomics in the Couple: Evidence from the PSID by Sonia Oreffice* Climent Quintana-Domeque** DOCUMENTO DE TRABAJO 2009-22 Economía de la Salud y Hábitos de Vida FEDEA – “la Caixa”

June 2009

* Universitat d'Alacant ** Universitat d'Alacant & FEDEA

Los Documentos de Trabajo se distribuyen gratuitamente a las Universidades e Instituciones de Investigación que lo solicitan. No obstante están disponibles en texto completo a través de Internet: http://www.fedea.es. These Working Paper are distributed free of charge to University Department and other Research Centres. They are also available through Internet: http://www.fedea.es. ISSN:1696-750X

Jorge Juan, 46 28001 Madrid -España Tel.: +34 914 359 020 Fax: +34 915 779 575 infpub@fedea.es

Anthropometry and Socioeconomics in the Couple: evidence from the PSID

Sonia Oreffice Universitat d'Alacant

Climent Quintana-Domeque Universitat d'Alacant & FEDEA

First draft: June 23, 2009.

Abstract

We empirically analyze the marriage market aspects of body size, weight and height in the US using data from the Panel Study of Income Dynamics on anthropometric characteristics of both spouses. Gender-asymmetric trade-offs arise within couples between physical and socio-economic traits, but also between anthropometric traits, with significant penalties for fatter women and shorter men. Wives' obesity (body size or weight) measures are negatively correlated with their husbands' income, education and height, controlling for his weight (or body size) and her height, along with spouses' demographic and socioeconomic characteristics. Conversely, heavier husbands are not penalized by matching with poorer or shorter wives, but only with less educated women. Men's and women's height are both valued in the market, with shorter men matched to heavier and less educated wives, and shorter women to poorer and less educated husbands (the latter effect only shows up in 2005).

JEL codes: D1, I1, J1.

Keywords: weight, height, BMI, marriage market.

1. Introduction

Spouses tend to resemble each other on a variety of characteristics including age, education, race, religion, and physical traits such as height and weight (Becker, 1991; Weiss and Willis, 1997; Qian, 1998; Silventoinen et al., 2003). Specifically, positive assortative mating in body weight has been established in the medical and psychological literatures, which document significant and positive interspousal correlations for weight (Schafer and Keith, 1990; Allison et al., 1996; Speakman et al., 2007), and the importance of the examination of both spouses' qualities on their marriages (Fu and Goldman, 2000; Jeffrey and Rick, 2002; McNulty and Neff, 2008).

Recently, Kano (2008) investigates the joint dynamics of spousal obesities in the US and finds that the probability of an individual being obese is positively associated with past obese status of his/her spouse. Belot and Fidrmuc(2009) and Herpin (2005) consider height as a determinant of marriage rates, the former analyzing interracial marriage rates and linking them to gender preference for height differences, the latter showing that the probability of being in a relationship is lower for shorter men.

In this study, we examine the marriage market effects of body size, height and weight, and investigate the spousal trade-offs among these anthropometric traits and socioeconomic characteristics, such as income and education, which men and women face in the marriage market. Body size is measured by means of the body mass index (BMI), which is defined as the ratio of weight (in kilograms) to squared height (in meters squared).

Using data from the Panel Study of Income Dynamics (PSID) on heads and wives from 1986 to 2005, we focus on the within-couple correlation in both anthropometric and socioeconomic measures, controlling for other individual and household characteristics of the couple members. Assessing the extent of marital sorting along the anthropometric and socioeconomic dimensions allows us to analyze the actual marriage market impact that body size, weight and height have on individuals. Does the market penalize obesity and reward height (and income) by matching physically unfit individuals with partners who are weaker along other physical and/or socioeconomic dimensions? What is the relative importance of the physical versus socioeconomic traits in the marriage market? Are these traits assessed differently by gender?

We specifically address these issues and find that female physical appearance (proxied by anthropometric measures, namely body size and weight) plays a more relevant role in the marriage market than men's, as obese women are thrice penalized with husbands of both lower (socio)economic and physical status (poorer, less educated, and shorter). Shorter wives tend to be matched to lower socioeconomic trait husbands (less educated and poorer), although this finding only appears in 2005. Shorter husbands are instead only penalized with lower physical trait (heavier) and less educated wife. Moreover, husbands' weight entails a one-dimensional marriage market penalty in terms of lower education of the wife. These gender-asymmetric cross-effects of body size, weight, height and income have not been emphasized in the literature, where male penalties and female height rewards in the marriage market have been often overlooked.

Our empirical analysis reveals that a one-pound increase in average wife's weight is associated with being married with a husband who is 0.14 inches shorter, a physical penalty. Thinner women tend to be married with taller men: an increase of one inch in average husband's height is associated with being married with a wife who weighs 2 pounds and a third less. We also find an economic penalty for heavier women: thinner women tend to be married with richer men. An increase of in average husband's earnings is associated with being married with a wife who has a BMI of almost less, who weighs 5 pounds and two thirds less, and who is one fourth inch taller. Moreover, a one-percent increase in average husband's earnings is associated with a 5-percent lower probability of being married with an obese wife.

We are not aware of any study looking at the within-couple correlation between anthropometric and socioeconomic characteristic using PSID data. The only two related studies using PSID data to analyze the weight correlations are as follows. Kano (2008) investigates the joint dynamics of spousal obesities and finds that the probability of an individual being obese is positively associated with past obese status of his/her spouse. However, he controls only for own individual socioeconomic characteristics and total household income, without including the corresponding spouses' variables. It is also worth noting the article by Conley and Glauber (2007). They use siblings rather than couples from the PSID, and find that, for women, BMI is negatively associated with family income, the likelihood of marriage, spouse's occupational prestige and spousal earnings. However, for men, BMI is positively associated with spousal earnings.

There is also a large body of literature using National Longitudinal Youth Study data linking women's weight to lower spousal earnings or likelihood of being in a relationship (Averett and Korenman, 1993; Averett et al., 2008; Mukhopadhyay, 2008). However, these data provide only anthropometric measures of the respondent, so that the trade-off weight-income is estimated without controlling for husbands' physical attributes.

The findings presented here are consistent with the marriage market reinforcing the negative effects of women's weight by sorting obese women with poor, less educated and short men. Moreover, they provide empirical support for female physical appearance (proxied by body size or weight) to play a more relevant role in the marriage market than men's, accounting for spousal cross-effects in both anthropometric traits and income, as heavier women are thrice penalized with husbands of both lower socio-economic and physical status, while men's weight is penalized only by matching with less educated wives. Men's and women's height are also valued in the market, with shorter men matched to heavier and less educated wives, and shorter women to poorer and less educated husbands. Female height seems to be appreciated in the marriage market; specifically, it is twice rewarded through both the socioeconomic characteristics of the husbands. This positive assessment has not been emphasized in the literature, but it seems to be very recent, since it only appears to be statistically significant in 2005.

Our findings can also be contextualized in the economic research agenda on the effects of anthropometric measures. Many economists have been working on assessing the effects of height, BMI or obesity on labor market outcomes. The general consensus seems to be that BMI or obesity has negative effects on labor market outcomes, such as hourly wages and the probability of employment, particularly for women (Han, Norton, and Stearns, 2009), while height has a positive effect on wages, perhaps reflecting the fact that taller people are more likely to have reached their full cognitive potential (Case and Paxson, 2008). On top of these labor market effects, and the well-know negative health effects of obesity, we provide evidence of additional consequences that obese individuals may experience in the marriage market.

The paper is organized as follows. Section 2 describes the data. Section 3 presents the empirical results. Section 4 offers some robustness checks. Finally, Section 5 concludes.

2. Data Description

Estimation is carried out on data from the Panel Study of Income Dynamics (PSID). The PSID is a longitudinal household survey collecting a wide range of individual and household demographic, income and labor market variables. Additionally, in the survey year 1986, and in all the most recent waves since 1999 (1999, 2001, 2003 and 2005), detailed information on weight and height of both heads and wives is available, which we use to construct measures of body mass index (BMI) and obesity for each spouse. The former measure is defined as the ratio of weight in kilograms to height in meters squared, while the latter is a dummy variable which takes value of one if an individual has BMI of 30 or above (WHO, 2003).

In the PSID all the variables are reported by the head of the household, including the information on the wife. Although it is well-known that self-reported anthropometric measures are likely to suffer from measurement error, the error seems to be constant for the 25-55 age group according to the analysis in Thomas and Frankenberg (2000) and Ezzati et al. (2006). However, notice that in the PSID the household head is reporting both his own and his wife's height and weight. Hence, it may not be appropriate to rely on the above measurement error findings. We will discuss the implications of this feature in the robustness checks section.

Our main sample consists of white married couples with wives being between 25 and 50 years old, the age group for which the effects of physical appearance (proxied by body size or weight) on economic outcomes should be more relevant. A couple consists of the head and his wife. We include intact couples only if both the head and the wife are actually present. In our sample years, all the married heads with spouse present are male and the wives are female.

We run regressions of spouses' physical traits, separately by gender and by year, controlling for both spouses' physical, demographic and economic characteristics. Our dependent variables are, in turn, BMI, a dummy variable for being obese, weight or height. Weight and height are reported in pounds and inches respectively in the PSID.

The other regressors are age and education of husband and wife, the latter defined as number of completed years of schooling, and top-coded at 17 for some completed graduate work, number of children in the household under 18 years of age. Income variables include individual earnings, as total individual income is not available in the PSID in recent years. We focus on working couples, however we also perform our estimation including non-working individuals, and results are robust. Labor supply is defined as annual hours of work (the income and labor variables refer to the year before the interview). From the health status originally recorded by the PSID as a 5-category variable (from excellent to poor health), we create a dummy variable for being in good health (1 if excellent, very good, or good; 0 if fair or poor). State fixed effects are included to capture constant differences in marriage markets across geographical areas in the US.

We exclude observations from the top and bottom 1% of the income distribution. We also discard the couples with extreme values of height and weight, as customary in the anthropometric literature (Conley and Glauber, 2007). Thus, we exclude couples where the weight of either spouse is greater than 400 pounds or less than 70, and height greater than 84 inches or less than 45 inches. Finally, household weights are used.

Table 1 presents the descriptive statistics for the husbands' and wives' main variables. On average, wives are younger, a little bit more educated, earn lower income and work fewer hours than their spouses. The first two findings may be the result of our age restriction 25-50 for the wife. The prevalence of obesity among the husbands is , while for wives is . These results are in line with Kano (2008) and contrast with Ogden et al. (2006), who estimate that the US adult prevalence rate is for male and for female in 2003-04, using data from the NHNES. As Kano (2008) points out this difference might stem from the fact that we focused on married couples, not on the general US population.

[Table 1 about here]

3. Results

Table 2 presents the results of several regressions where the dependent variable is wife's BMI. The first column shows a strong correlation between wife's and husband's BMIs controlling for state fixed effects. Column (2) adds wife's age and her completed education level as well as the number of children in the household. Higher educated wives, on average, tend to be thinner. Moreover, the strong correlation between wife's and husband's BMIs persists. Column (3) uses the corresponding variables for the husband, revealing that higher educated husbands, on average, tend to be married with thinner women. Column (4) controls for both wife's and husband's characteristics confirming the previous relationships, but with smaller magnitudes. Once we control for wife's log earnings, column (5), the correlation between wife's BMI and her completed education level decreases, and the correlation with her log earnings is negative: low-earnings women tend to be heavier than high-earnings women. In column (6) we add husband's log earnings: high-earnings men tend to be married with thinner women. Finally, column (7) adds hours of work for both wives and husbands and indicators of their self-reported health status. The results emerging from this final column are worth noting. Wife's BMI is explained only by her spouse's characteristics other than health, and her self-reported health status: healthier women tend to be thinner, and better educated and/or richer husbands tend to be matched to thinner women.

[Table 2 about here]

These results point towards two kinds of relationships: one responds to economic behavior, the other to biological mechanisms. First, we can think of biology as providing the link between health and BMI, as our data show a relationship between wives' BMI and their health status (both reported by their husbands). Second, we also find relationships between wives' BMI and their husband's BMI, education and income, which seem to respond to socio-economic mechanisms. As emphasized by Carmalt (2009), the strong relationship between wife's and husband's BMI may arise from three sources: (1) active assortative mating (selection of a partner based on phenotypic preferences), (2) social homogamy (selection of a partner from within one's own social setting or geographical area), and (3) convergence (the tendency to become similar in weight due to sharing a common environment). Importantly, controlling for state fixed effects can be thought as an easy way to account for social homogamy, albeit rather crude. In a similar vein, one might think that controlling for age could help us to control for the tendency of couples to become similar over time. We interpret the BMI correlations as an overestimate of the extent of assortative mating. Perhaps, the panel structure of the PSID can help us to test for this convergence. Nevertheless, our focus is on the spousal trade-offs between BMI, education and income: it seems that husbands with better socio-economic characteristics tend to match with thinner wives. This is not necessarily evidence of either husband's income or education leading him to marry a thinner wife. Better educated or richer husbands might be also smarter or more sociable, or different in unobservable characteristics, which may make them more attractive in the marriage market.

Table 3 presents the corresponding regression results where the dependent variable is now husband's BMI. The results are dramatically different: First, unhealthier and/or richer husbands tend to be heavier. Second, healthier and/or less educated wives' tend to be married with heavier men.

[Table 3 about here]

The previous Tables look at the determinants of wife's and husband's BMIs, estimating averages of wife's and husband's BMI conditional on several characteristics. However, means are not necessarily informative about the tails of the BMI distribution, so that we also explore the determinants of the likelihood of being obese for both wives and husbands. This analysis is presented in Tables 4 and 5.

Table 4 highlights the results of several probit regressions where the dependent variable takes the value 1 if the wife is obese and 0 otherwise. The results closely match those reported in Table 1. More importantly, the qualitative results, column (7), are basically the same: wife's obesity just depends on husband's characteristics. On the one hand, obese men are more likely to be married to obese women. On the other hand, higher-earnings and/or unhealthier husbands are more likely to be married with non-obese wives.

[Table 4 about here]

Table 5 presents the corresponding regression results where the dependent variable takes the value 1 if the husband is obese and 0 otherwise. Again, the results closely match those reported in Table 2, except perhaps for one important difference. While younger and/or unhealthier husbands are more likely to be obese and healthier wives' tend to be married with obese men, now higher-earnings husbands are not more or less likely to be obese. Hence, although higher-earnings husbands tend to be heavier, they are not more likely to be obese.

[Table 5 about here]

Tables 6, for women, and 7, for men, present the results of several regressions where the dependent variables are weight and height.

Table 6 shows that heavier women are thrice penalized with husbands of both lower (socio) economic and physical status (poorer, less educated, and shorter), columns (1)-(3), while shorter wives tend to be matched to lower socioeconomic trait husbands (less educated and poorer), columns (4)-(6). Heavier women are more likely to be married to heavier men, to earn a lower income and to be in worse health, as found in the literature. Tall women are more likely to be matched to tall men. An interesting pattern arises from the analysis of wives' height. Columns (4) to (6) show that on average, higher earnings and/or better educated men tend to be married to taller women. This positive correlation seems to go against the traditional view that the height of a woman is an undesirable trait in the couple, as reported for instance in Belot and Fidrmuc (2009). However, we must emphasize that, first, these estimates show average effects and, second, they hold only for the most recent year 2005. Indeed, preliminary results using quantile regression show that height only matters for taller women. The reward to tall women in terms of husband's earnings and education is only present in couples where the wife belongs to the upper tail of the height distribution, while there is no evidence of any such compensation or penalty for short women in the lower tail of the height distribution. It is worth noting that the weight quantile regressions show instead that female weight matters along the entire distribution, the negative correlation with husbands' earnings and education being significant for both thin and heavy women, even though the effect increases with wives' weight.

[Table 6 about here]

Table 7 shows that husbands' weight entails a marriage market penalty only in terms of lower education of the wife, columns (1)-(3), while shorter husbands are instead penalized with lower physical trait (heavier) and less educated wife, columns (4)-(6). Heavier men are more likely to be married to heavier women, to earn a higher income, to be in worse health and less educated. Tall men are more likely to be matched to tall women and to be more educated.

[Table 7 about here]

A common pattern across specifications is the lack of an effect of own age on BMI, obesity or weight. We relate this statistically insignificant correlation to the low weight variation in the younger age group 25 to 50 years old. As a matter of fact, the age effect is significant when older couples are included.

Carmalt et al. (2008) use a very young sample of couples from Add Health, aged 18-24, and similarly find that obese individuals, women especially, were less likely to have physically attractive (interviewer-rated attractiveness and grooming) partners, and that more education increased the probability of having a physically attractive partner. However they do not find any impact of individual income on matching probabilities. The comparison between their age group 18-24 and our 25-50 year is consistent with the fact that at very young ages income cannot yet be a great predictor of mate quality and economic prosperity as not enough time elapsed to establish social status, whereas in adult individuals it is.

Our findings are also in line with the evidence from psychology (e.g. Braun and Bryan, 2006), which report that men put more emphasis on female body shape and physical attractiveness, while women value signals of socioeconomic status, and also prefer taller men. Furthermore, our results suggest that physical attractiveness is indeed valued in long-term relationships such as marriage, not only in short-term encounters.

4. Robustness checks

4.1. Collinearity among spousal variables

In the previous section we find that wife's BMI is explained only by her spouse's characteristics other than health, and her self-reported health status: healthier women tend to be thinner, and better educated and/or richer husbands tend to be matched to thinner women. However, because spouses tend to resemble each other on a variety of characteristics, this finding might be the result of the high correlation of individual characteristics within couples. In other words, if there is high collinearity among spousal variables, it might be difficult to determine which independent variable, whether the one of the wife or of the husband, is actually predicting the anthropometric characteristics of the wife. Hence, we need to analyze the correlation among the independent variables. For example, if we are interested in assessing the relationship between and wife's anthropometrics (BMI, obesity, weight or height), we need to be aware about the correlation between and the rest of independent variables (all except ), in particular about the correlation between the wife's and the husband's . If the correlation is high, the variance of the coefficient estimate is being inflated by multicollinearity.

The Variance Inflation Factor (VIF) of the coefficient estimate for variable k can be written as , where is the of a linear regression of on the rest of independent variables . The square root of the VIF tells us how much larger the standard error is, compared to what it would be if that variable were uncorrelated with the other variables in the equation.

Although the VIF is the simplest and most direct measure of the harm produced by multicollinearity, there is no formal theory-based cutoff value for the VIF. Nevertheless, many researchers suggest that, as a general “rule of thumb”, multicollinearity becomes a problem when the VIF value is greater than 2.5 (Allison, 1999). Other researchers, however, have suggested that multicollinearity is not a cause for serious concern until VIF value is greater than 10 (Baum, 2006). In Table 8 we report the VIFs for the main independent variables in the previous regressions. Columns (1) and (2) contain the VIF for the main independent variables of regressions in column (7) of tables 2 and 3 respectively. Columns (3) and (4) report the VIF for the main independent variables in regressions in columns (3) and (6) of table 6, while columns (5) and (6) refer to regressions in columns (3) and (6) of table 7.

[Table 8 about here]

All the VIF values reported on Table 8, but the ones of the age variables, are below 2.5. This suggests that mutilcollinearity does not appear to be the responsible factor for the result about wife's BMI being explained mainly by her spouse's characteristics.

4.2. Disentangling body size effects by age group

In the paper our sample is composed of young couples (between 25 and 50), but we are currently working with a sample of older couples (between 51 and 65 years old) to test for age differences in the responsiveness to spousal characteristics. It is not obvious what are going to be the effects of body size (or weight) for old age groups. On the one hand, we should expect physical appearance (proxied by weight or body size) to be less relevant, implying lower impact of body size on socio-economic outcomes. On the other hand, health problems and costs associated with body size (i.e., obesity) are expected to be larger for the elderly with stronger socio-economic consequences.

4.3. Body size effects by duration of marriage

We consider a subsample of couples who got married the year before the PSID interview, to start exploring the impact of duration of marriage on body size, and disentangling the convergence versus assortative mating forces emphasized by Carmalt (2009). Preliminary results show that we cannot reject the hypothesis that the correlations between husband's earnings and wife's BMI is equal for both the newly-weds and the other couples. However, we find that the penalty for heavier women in terms of lower husband's education seems to be higher for recently married couples. Education appears to be a persistent sign of mate quality. In particular, education is a better proxy for permanent income than current earnings, especially for newly-weds, who are about 7 years younger (median age) than the other married couples in our sample. At younger ages education is a better predictor of mate quality and economic prosperity because not enough time elapsed to establish social status and high earnings.

4.4 Pooled regressions across years

In the main tables, we only report regressions using PSID for the year 2005. However, our qualitative results are robust when using PSID data for the years 1999, 2001, 2003 and 2005 altogether, controlling for state, year and time-state fixed effects (Tables A.1 and A.2). When including less recent years, only the wife's height appears to be no longer statistically correlated to husband's education and earnings.

4.5. Sensitivity analysis

The results are robust to the use of earnings rather than log earnings and to the exclusion of observations from the poverty sample SEO or the immigrant sample. It is important to note that in the PSID blacks are disproportionately over-represented in the poverty sample (SEO), so that they are not the focus of our analysis.

4.6. Measurement error in self-reported anthropometric measures

In principle we can adjust the husband's self-reported anthropometric variables (Cawley, 2004), but it is not clear what we can do with wife's anthropometric variables which are reported by the husband. Perhaps, we can assume that they are measured with classical measurement error. This is an issue we are still working on.

5. Preliminary conclusions

We examine the marriage market effects of body size, weight, and height and estimate the trade-offs among these anthropometric characteristics, education and income, that men and women exhibit within couples. Using PSID data on both heads' and wives' anthropometric, demographic and income information, we find that female physical appearance (proxied by anthropometric measures, namely body size and weight) plays a more relevant role in the marriage market than men's, as obese women are thrice penalized with husbands of both lower (socio)economic and physical status (poorer, less educated, and shorter). Husbands' weight entails a marriage market penalty only in terms of lower education of the wife. Shorter wives tend to be matched to lower socioeconomic trait husbands (less educated and poorer), at least in 2005. Shorter husbands are instead only penalized with lower physical trait (heavier) and less educated wife. These gender-asymmetric cross-effects of body size, weight, height and income have not been emphasized in the literature, where male penalties and female height rewards in the marriage market have been often overlooked.

These findings represent strong empirical support for female obesity to play a more relevant role in the marriage market than men's, and for men's height to be perceived as valuable mate quality, while accounting for spousal cross-effects in both anthropometric characteristics and income.

References

1. Allison, D.B., Neale, M.C.: Kezis, M.I., Alfonso, V.C., Heshka, S., Heymsfueld, S.B. (1996) “Assortative mating for relative weight: genetic implications,” Behavior Genetics, 2, 103-111.
2. Allison, P. D. (1999) Multiple Regression: A Primer. Pine Forge Press, Thousand Oaks, California.
3. Averett, S., Korenman, S. (1996) “The economic reality of the beauty myth,” Journal of Human Resources, 31, 304-330.
4. Averett, S.L., Sikora, A., Argys, L.M. (2008) “For better or worse: relationship status and body mass index,” Economics and Human Biology, 6, 330-349.
5. Baum, C. (2006) An Introduction to Modern Econometrics Using Stata. Stata Press.
6. Becker, G. (1991) A Treatise on the Family, Harvard University Press.
7. Belot, M., Fidrmuc, J. (2009) “Anthropometry of love. Height and gender asymmetries in interethnic marriages,” mimeo.
8. Braun, M.F., Bryan, A. (2006) “Female-waist-to-hip and male waist-to-shoulder ratios as determinants of romantic partner desirability,” Journal of Social and Personal Relationships, 23, 805-819.
9. Carmalt, J. (2009) “Assortative Mating for Body Weight: New Evidence from Multilevel Models,” PAA 2009.
10. Carmalt, J.H., Cawley, J., Joyner, K., Sobal, J. (2008) “Body weight and matching with a physically attractive partner,” Journal of Marriage and the Family, 70, 1287-1296.

Case, A., Paxson, C. (2008) “Stature and status: height, ability, and labor market outcomes,” Journal of Political Economy, 116, 499-532.

Cawley, J. (2004) “The impact of obesity on weight,” Journal of Human Resources, 39, 451-474.

Conley, D., Glauber, R. (2007) “Gender, Body Mass and Economic Status,” Advances

in Health Economics and Health Services Research, 17: The Economics of Obesity.

Ezzati, M., Martin H., Skjold, S., Vander Hoorn, S., and Murray C. (2006) “Trends in national and state-level obesity in the USA after correction for self-report bias: analysis of health surveys,” Journal of the Royal Society of Medicine, 99, 250-257.

1. Fu, H., Goldman, N. (2000) “The association between health-related behaviors and the risk of divorce in the USA,” Journal of Biosocial Sciences, 32, 63-88.

Jeffrey, R.W., Rick, A.M., (2002) “Cross-sectional and longitudinal associations between body mass index and marriage market factors,” Obesity Research, 10, 809-815.

Han, E., Norton, E., Stearns, S. (2009) “Weight and wages: fat versus lean paychecks,” Health Economics, 18, 535-548.

Herpin, N. (2005) “Love, careers, and heights in France, 2001,” Economics and Human Biology, 3, 420-449.

Kano, S. (2008) “Like Husband, Like Wife: A Bivariate Dynamic Probit Analysis of Spousal Obesities”, Osaka Prefecture University, mimeo.

Mare, R., Schwartz, C. (2005) “Trends in Educational Assortative Marriage from 1940

to 2003," California Center for Population Research, 017-05.

McNulty, J.K., Neff, L.A., (2008) “Beyond initial attraction: physical attractiveness in

newlywed marriage," Journal of Family Psychology, 22, 135-143

Mukhopadhyay, S., (2008) “Do women value marriage more? The effects of obesity on

cohabitation and marriage in the USA," Review of Economics of the Household, 6, 111-

126

Ogden, C.L., Carroll, M.D., Curtin, L.R., McDowell, M.A., Tabak, C.J., Flegal, K.M.

(2006) “The prevalence of overweight and obesity in the united states, 1999-2004,”

Journal of American Medical Association, 295, 1549-1555

Qian, Z. (1998) “Changes in Assortative Mating: The Impact of Age and Education,”

Demography, 279-292

Schafer, R.S., Keith, P.M. (1990) “Matching by weight in married couples: a life cycle

perspective," The Journal of Social Psychology, 130, 657-664

Silventoinen, K., Kaprio, J., Lahelma, E., Viken, R.J., Rose, R.J. (2003) “Assortative

mating by body height and BMI; finnish twnis and their spouses," American Journal of

Human Biology, 15, 620-627

Speakman, J.R., Djafarian, K., Stewart, J., Jackson, D.M. (2007) “Assortative mating

for obesity," American Journal of Clinical Nutrition, 86, 316-323

Thomas, D. and Frankenberg, E. (2002) “The measurement and interpretation of health

in social surveys." Chapter 8.2 (pp. 387-420) in Measurement of the Global Burden of

1. Disease. C.Murray, J. Salomon, C. Mathers, and A. Lopez (Eds.). Geneva: World Health Organization.
2. Weiss, Y., Willis, R. (1997), “Match Quality, New Information, and Marital Dissolition,” Journal of Labor Economics, 15, 293-329.
3. World Health Organization (WHO) (2003) Diet, nutrition, and the prevention of chronic diseases. WHO Technical Report Series 916, Geneva.

Table 1: Descriptive Statistics – Means and (Standard Deviations), PSID 2005.

Note: Family weights are used.

Table 2: Determinants of wife's BMI. PSID 2005. Aged 25-50.

Note: Heteroskedastic robust standard errors are reported in parenthesis. Family weights are used. *** (**) [*] Statistical significance at the 1% (5%) [10%].

Table 3: Determinants of husband's BMI. PSID 2005. Aged 25-50.

Note: See footnote Table 2.

Table 4: Determinants of wife's Obesity. PSID 2005. Aged 25-50.

Note: See footnote Table 2. Probit estimates: marginal effects evaluated at the average sample characteristics are reported.

Table 5: Determinants of husband's Obesity. PSID 2005. Aged 25-50.

Note: See footnote Table 2. Probit estimates: marginal effects evaluated at the average sample characteristics are reported.

Table 6: Determinants of Weight and Height for Wives. PSID 2005. Aged 25-50.

Note: See footnote Table 2.

Table 7: Determinants of Weight and Height for Husbands. PSID 2005. Aged 25-50.

Note: See footnote Table 2.

Table 8: VIF of the main independent variables

Table A1: Determinants of BMI and Obesity. PSID 1999, 2001, 2003, 2005. Aged 25-50.

Note: See footnote Table 2. All regressions include year, state, and state-year fixed effects. Robust standard errors clustered at the household head level.

Table A2: Determinants of Weight and Height. PSID 1999, 2001, 2003, 2005. Aged 25-50.

Note: See footnote Table 2. All regressions include year, state, and state-year fixed effects. Robust standard errors clustered at the household head level.

1. 2009-22: “Anthropometry and Socioeconomics in the Couple: Evidence from the PSID”, Sonia Oreffice y Climent Quintana-Domeque.
2. 2009-21: “Stimulating Graduates' Research-Oriented Careers: Does Academic Research Matter?”, Mauro Sylos Labini y Natalia Zinovyeva.
3. 2009-20: “Papers or Patents: Channels of University Effect on Regional Innovation”, Robin Cowan y Natalia Zinovyeva.
4. 2009-19: “Innovation, Tangible and Intangible Investments and the Value of Spanish Firms”, Aitor Lacuesta, Omar Licandro, Teresa Molina y Luis A. Puch.
5. 2009-18: “Estimation of Elasticity Price of Electricity with Incomplete Information”, Xavier Labandeira, José M. Labeaga y Xiral López-Otero.
6. 2009-17: “MEDEA: A DSGE Model for the Spanish Economy”, Pablo Burriel, Jesús Fernández-Villaverde y Juan F. Rubio-Ramírez.
7. 2009-16: “Greenhouse gases emissions, growth and the energy mix in Europe: A dynamic panel data approach”, Gustavo A. Marrero.
8. 2009-15: “Impact of the Rise in immigrant unemployment on public finances”, Pablo Vazquez, Mario Alloza, Raquel Vegas y Stefano Bertozzi.
9. 2009-14: “Responding to Financial Pressures. The Effect of Managed Care on Hospitals’ Provision of Charity Care”, Núria Mas.
10. 2009-13: “Domestic Transport Cost Reductions and Firms’ Export Behaviour”, Pedro Albarran, Raquel Carrasco y Adelheid Holl.
11. 2009-12: “Compatibility with Firm Dominance”, María Fernanda Viecens.
12. 2009-11: “Pricing Strategies in Two-Sided Platforms: The Role of Sellers’ Competition”, María Fernanda Viecens.
13. 2009-10: “Scheduled Service Versus Personal Transportation: the Role of Distance”, Volodymyr Bilotkach, Xavier Fageda y Ricardo Flores-Fillol.
14. 2009-09: “Social Preferences and Strategic Uncertainty: An Experiment on Markets and Contracts”, Antonio Cabrales, Rafaele Miniaci, Marco Piovesan y Giovanni Ponti.
15. 2009-08: “Hidden Information, Bargaining Power and Efficiency: An Experiment”, Antonio Cabrales, Gary Charness y Marie Claire Villeval.
16. 2009-07: “Democracy and the curse of natural resources”, Antonio Cabrales y Esther Hauk.
17. 2009-06: “Social Interactions and Spillovers: Incentives, Segregation and Topology”, Antonio Cabrales, Antoni Calvó-Armengol e Yves Zenou.
18. 2009-05: “Chance Constrained Programming with one Discrete Random Variable in Each Constraint”, Emilio Cerdá Tena y Julio Moreno Lorente.
19. 2009-04: “Economic Value of Weather Forecasting Systems Information: A Risk Aversion Approach”, Emilio Cerdá Tena y Sonia Quiroga Gómez.
20. 2009-03: “Population Ageing, Inequality and the Political Economy of Public Education”, Francisco Martínez-Mora.
21. 2009-02: “Real Wages over the Business Cycle: OECD Evidence from the Time and Frequency Domains”, Julian Messina, Chiara Strozzi y Jarkko Turunen.
22. 2009-01: “The Determinants Of Misreporting Weight And Height: The Role Of Social Norms”, Joan Gil y Toni Mora.
23. 2008-42: “Social Security Incentives, Exit from the Workforce and Entry of the Young”, Michele Boldrin, Pilar García-Gómez y Sergi Jiménez-Martín.
24. 2008-41: “The Evolution and Main Determinants of Productivity in Brazilian Electricity Distribution 1998-2005: an Empirical Analysis”, Francisco Javier Ramos-Real, Beatriz Tovar, Mariana Iootty, Edmar Fagundes de Almeida y Helder Queiroz Pinto Jr..
25. 2008-40: “Immigration and Housing Prices in Spain”, Simón Sosvilla.
26. 2008-39: “Modeling the Immigration Shock”, Ana Montes y Michele Boldrin.
27. 2008-38: “Immigration and the Demand for Health in Spain”, Sergi Jiménez, Natalia Jorgensen y José María Labeaga.
28. 2008-37: “Immigration and Students' Achievement in Spain”, Natalia Zinovyeva, Florentino Felgueroso y Pablo Vázquez.
29. 2008-36: “Immigration and Social Security in Spain”, Clara Isabel González, J. Ignacio Conde-Ruiz y Michele Boldrin.
30. 2008-35: “Complements or Substitutes? Immigrant and Native Task Specialization in Spain”, Catalina Amuedo-Dorantes y Sara de la Rica.