A Rationale for Marriage

Yoram Barzel (University of Washington)
Aurora Stephany (University of Washington)
Qing Zhang (University of Washington)

Abstract: We argue that the rationale for marriage is to provide paternity assurance so that men have an incentive to contribute to the upbringing of the children. Without the prospect of material support, women will simply seek the highest standing man they can reach to father their children. The society will then be similar to that of sea lions or chimpanzees. However, material support enhances the survival of children; an arrangement that creates incentives for males to provide support is likely to benefit the survival of the species. We propose that the institution of marriage, along with the constraints it imposes on women, serve the purpose of assuring paternity so that men are willing to spend resources on their offspring. We conduct two empirical tests for the hypothesis. The first one relates to the ease of determining paternity and the constraints on the wives. Our hypothesis predicts that the easier it is to determine paternity, the fewer the constraints. The emergence of genetic paternity tests offers an opportunity to test this implication. We show that states that accepted genetic testing as a legally valid way of establishing paternity early experienced an increase in women’s labor participation sooner than states that initially ignored the new technology. Our second test compares the relative achievement of children of intentional single mothers with that of children of unintentional single mothers. A clear implication of our hypothesis is that women who plan to be single mothers will seek men of higher standings than them, while women who plan to marry in order to obtain support will marry men of the same standings as them. And we expect that children of the two groups of women will perform differently. We show that the children of intentional single mothers outperform their mothers by a greater magnitude than do the children of unintentional single mothers, as predicted by our hypothesis.