Gender Peer Effects, Non-Cognitive Skills and Marriage Market Outcomes: Evidence From Single-Sex Schools in the UK

In this paper we analyze the long-run effects of single-sex schooling on individuals marriage and family outcomes. First, we show that individuals positively select into single-sex schools, i.e. they have (ex-ante) characteristics (such as higher cognitive and non-cognitive skills, better health, more likely catholic, more highly educated parents and higher family income) which are positively correlated with marriage (negatively with divorce). Despite positive selection, we find that singlesex education negatively affects men’s likelihood to ever having been married by their mid-forties and increases the likelihood of separation/divorce and we show that the estimated coefficients are likely to be lower bounds (in absolute value) of the true effects. For women on the other hand we do not find any effects. In terms of mechanisms, we show that single-sex schooling does not affect individuals’ preferences/aspirations for marriage suggesting negative welfare implications since men who attended single-sex schools are less likely to reach those goals. Instead, likely channels are fewer (romantic) interactions with the opposite gender during teenage age (even outside of school) and effects of single-sex education on boys’ non-cognitive skills (such as becoming more cautious and less aggressive) which negatively affect their marital chances. Lastly, we find that the likelihood of having a child is the same, while the likelihood of a (stable) marriage (conditional on having a child) is reduced with potentially important negative consequences for those children.