Working papers

The Role of Evolving Marital Preferences in Growing Income Inequality (with Simon Weber)

Status: revision and resubmission requested at Journal of Population Economics

In this paper, we describe mating patterns in the United States from 1962 to 2015 and measure the impact of changes in marital preferences on between-household income inequality. We rely on the recent literature on the econometrics of matching models to estimate complementarity parameters of the household production function. Our structural approach allows to measure sorting on multiple dimensions and to effectively disentangle changes in marital preferences and in demographics, addressing concerns that affect results from existing literature. We answer the following questions: has assortativeness increased over time? Along which dimensions? To which extent the shifts in marital preferences can explain inequality trends? We find that, after controlling for other observables, assortative mating on education has become stronger. Moreover, if mating patterns had not changed since 1971, the 2015 Gini coefficient between households would be lower by 6%. We conclude that about 20% of the increase in between-household inequality is due to changes in marital preferences.

Available at:

Like Attract Like? A Structural Comparison of Homogamy Across Same-Sex and Different-Sex Households (with Alfred Galichon and Marion Goussé)

Status: revision and resubmission requested at Journal of Political Economy

In this paper, we extend Gary Becker's empirical analysis of the marriage market to same-sex couples. Beckers's theory rationalizes the well-known phenomenon of homogamy among heterosexual couples: individuals mate with their likes because many characteristics, such as education, consumption behaviour, desire to nurture children, religion, etc., exhibit strong complementarities in the household production function. However, because of asymmetries in the distributions of male and female characteristics, men and women may need to marry "up" or "down" according to the relative shortage of their characteristics among the populations of men and women. Yet, among homosexual couples, this limit does not exist as partners are drawn from the same population, and thus the theory of assortative mating would boldly predict that individuals will choose a partner with nearly identical characteristics. Empirical evidence suggests a very different picture: a robust stylized fact is that the correlation of characteristics is in fact weaker among the homosexual couples. In this paper, we build an equilibrium model of the same-sex marriage market which allows for straightforward identification of the gains to marriage. We estimate the model with 2008-2012 ACS data on California and show that positive assortative mating is weaker for homosexuals than for heterosexuals with respect to age and race. Yet, contrarily to previous empirical findings, our results suggest that postitive assortative mating with respect to education is stronger on the same-sex marriage market. As regards labor market outcomes, such as hourly wages and working hours, we find that the process of specialization within the household mainly applies to heterosexual couples.

Available at:

Work in progress

Marriage, Divorce and Wage Uncertainty

The economic theory of the family suggests that divorce may occur following a negative shock on economic or non-pecuniary gains from marriage. A wage cut, for instance, is likely to have important consequences on household stability. Wage shocks may partly destroy gains from marriage by reducing disposable income, public good expenditure and consumption complementarities. The household sees its Pareto frontier shrinking but is still able to react by switching to a different allocation, so that the spouse who experienced the labor market shock is partially insured. However, if the shock hits too hard, the other spouse might decide to walk away rather than insuring his/her partner. When divorce costs are low and agents know that they can remarry, the lack of commitment represents a limitation for the insurance mechanism implied by the marriage contract. Moreover, forward-looking agents will adjust their marriage and divorce choices according to the wage distribution they (and their partners') face. Hence, understanding how families react to changes in wage risk along the business cycle is of primary importance for policymakers.
In this paper, I model the marriage market in a search and matching framework where agents experience stochastic changes in their wage rates. Married couples face two layers of uncertainty: about the quality of their match and about the wages of the spouses. When hit by a shock, a couple switches to another allocation on the shrunk Pareto frontier so that both spouses' participation constraints are satisfied. If no such an allocation is available, the couple splits. Reservation values correspond to the lifetime utilities as singles, and are endogenously computed taking into account agents' remarriage prospects. Such a characterization is only possible by solving for the search equilibrium of the model. Hence, this setup allows to study how household formation and stability are affected by (1) wage realizations and (2) different levels of uncertainty (an agent's future wage distribution). I show that the underlying production function is identified with matched data on partners' characteristics. Data on marriage and divorce flows help identify additional search parameters.

Sequential Matching Markets and Endogenous Reservation Value

It is well-known that any optimal assignment problem (Shapley and Shubik, 1971) can be expressed so that the reservation value is normalized to zero, although the latter may differ across individuals. Nonetheless, when agents are allowed to participate to several markets that open in a sequential fashion, they will ponder on the outcome of the next market before deciding whether and whom to match with. It follows that, while the economic value generated by a match only depends on the available technology, the surplus on the current market is endogenously determined after that individuals have figured out what their outside options are. When introducing uncertainty on the competitive environment that agents will face in future markets - notably on the marginals - there one has a general but insightful economic problem: how will different degrees of uncertainty affect the current matching? I solve the problem under simple assumptions (large markets, i.i.d. shocks) showing the relationship of the assignment problem with supermodular games. Finally, the conditions that bind surplus and reservation values are insightful to understand the similarities with search and matching frameworks, a topic that certainly deserves to be explored further.