The (non) impact of education on marital dissolution

Abstract

Despite the relevant role attributed to education on marital outcomes, literature does not show a generalized consensus regarding a positive or negative effect from education on marital decisions with findings differing across countries and linked to differences in several factors such as the labour market conditions, social policies and strictness of divorce legislation. In this paper I investigate the impact of education on marriage dissolution exploiting a change in the length of compulsory education in Mexico in 1993 as an instrument for education. The federal government increased compulsory education from completion of primary school, sixth grade, to completion of secondary school, ninth grade, at a national level. In the first part of the analysis, the probit models show education is significant and negatively correlated to the probability of marital breakdown. An additional year of education is associated with a decrease between 0.6 and 0.9 percentage points in the probability of marital disruption for the 2002-2012 period. However, the results using the instrumental variables methodology indicate an additional year of schooling has no effect on the probability of marriage dissolution. This finding demonstrates the relationship between education and divorce is not causal in Mexico, suggesting although higher levels of education are an undeniable trait observed in non-broken marriages, it is not education by itself one of the mechanisms leading to better marriage outcomes in a Latin American country.