October 19, 2017
Gender Peer Effects, Non-cognitive Skills and Marriage Outcomes: Evidence from Single-sex Schools in the UK
Katja Kaufmann, University of Mannheim
In this paper we analyze the longrun effects of single-sex schooling on individuals’ marriage market outcomes. We find that men who attended public single-sex schools in the UK are less likely to be married by 42 and more likely to be divorced, despite the fact that they are positively selected into single-sex schools. In other words, individuals selecting into those schools have (ex-ante) individual and family characteristics (such as high performance on cognitive and noncognitive skills, better health, higher attractiveness, more likely catholic, highly educated parents and high family income) which would make them more likely to be married and less likely divorced, as we show, i.e. coefficients are biased towards zero. Therefore the better we control for those characteristics (using regression and propensity score matching techniques), the more negative the effect on marriage becomes (and the more positive the effect on the likelihood of divorce). Interestingly, parents are less likely to send their male children to a single-sex school, if those have only brothers, consistent with parents expecting negative effects of single-sex schools on boys’ ability to interact with the opposite sex and mediating effects of having a sister. For girls who attended single-sex schools we find no effects on marriage or divorce. Moreover, in terms of mechanism we show that single-sex education does not affect individuals' preferences for marriage or family. Instead an important part of the effect on boys appears to be driven by single-sex schooling affecting teenagers’ noncognitive skills. In particular, boys who attended single-sex schools are more timid and more introverted at 16, which negatively affects their marital chances.