Recent research has derived estimable non-parametric transferable utility models of the marriage market. One of the major goals of the empirical analysis of matching models is to deal with endogeneity, measurement, and other problems. Accounting for endogenous matching in the empirical analysis of contracts helps explain the apparent clash between theoretical predictions on contracts and empirical observations.
This research will extend the theoretical and empirical literature on matching in three ways. First, it examines ways to empirically characterize bilateral matching markets with heterogeneous participants. Choo and Siow (2003) provide an empirical framework to estimate a transferable utility model of the marriage market without observing the equilibrium transfers. This research will use a unique dataset of about 2,700 marriage and dowry contracts for the city of Florence and the villages in her countryside in the early fifteenth century to empirically test this model. The research will also extend the CS model to allow for dynamic behavior in marriage markets, thus making it possible to consider important decisions made by prospective spouses, such as delaying marriage if expected prospects in the future look better.
The second extension is to use traditional hedonic market models to study marriage markets. Recent work has introduced econometric techniques for non-parametric identification of preferences in hedonic markets. This non-parametric identification is important since it ensures that empirical findings come from information in the data rather than from arbitrary functional forms or distributional assumptions. The research will apply these new techniques Italian dataset on marriage and dowry contracts. In addition this research will extend the hedonic methodology by allowing unobserved "traits", characteristics that not only affect an individual's preferences over potential spouses, but also potential spouses' preferences over the individual but not known to the econometrician, in the econometric model.
Finally, the PIs will study the consequences of matching in marriage markets on fertility and sex ratios by estimating a dynamic model of fertility. We will use data on the number and gender of children born to the couples in the marriage contracts described above. This approach allows the PIs to explain biased sex ratios without appealing to infanticide or differential care for boys and girls. The results of this research, together with the planned educational activities, are likely to have a significant intellectual and broader impact on economic science. Renaissance Tuscany.
