Domestic violence against women remains a pervasive global issue, despite being criminalised in almost every country. According to the World Health Organization, approximately one woman in three has experienced some form of physical or sexual abuse.

The incidence of domestic violence against women is even higher in the developing world. In Turkey, the country that we study in our research, fully 53% of women report having experienced abuse from their partners.

Clearly, criminalising domestic violence, while a prerequisite for its suppression, is not sufficient on its own. What is also needed is a transformation in the minds of perpetrators. To this effect, our study looks at the role of education to see whether additional schooling makes men less prone to abuse of their spouses.

A critical question in this context is one of causality. Education could lower the inclination to engage in domestic violence. But it is also possible that education and abusive behaviour are both driven by a third unknown factor.

For example, older men or those from rural areas might be more socially conservative and behave accordingly. They may also have less education than younger and urban men. Interpreting such correlation between education and abusive behaviour as causal would be incorrect.

Our research analyses a recent policy change in Turkey that resulted in a one-off increase in schooling: an extension, in 1997, of compulsory education from five to eight years. This reform is similar to an experiment that would randomly assign more education to some children but not to others.

Specifically, children born in 1986 or before had to stay in school for five years, those born in 1987 or later were required to study three years longer. It is unlikely that someone born in 1987 or 1988 was exposed to dramatically different social norms or had different upbringing in the family compared to his peers born just one or two years earlier.

Therefore, it can be assumed that the only material difference between them is that one received more education. Importantly, those affected by the reform are now in their 20s and 30s, allowing us to observe how staying in school longer affected their socio-economic outcomes.

The results of our analysis suggest that the extension of compulsory schooling has a marked effect on educational outcomes: among boys born in 1987 or during the following four years, the share of those completing eight years of schooling increased by 18 percentage points (from 63% to 81%), compared with those born during the five years before the reform.

This increase in education is associated with a lower propensity to engage in domestic violence against their wives. Among women whose husbands completed eight years of education, the share of those experiencing domestic violence is lower by 12 percentage points.

This result holds for all main types of domestic violence – physical, emotional and economic – with the only exception being sexual violence.

Schooling also lowers the likelihood that the marriage was arranged against the woman’s will. And it makes men less inclined to engage in socially unacceptable behaviours such as drinking, gambling or drug abuse.

Finally, we find that women whose mothers or husbands’ mothers experienced domestic violence are more likely to suffer violence themselves. This means that lowering the incidence of domestic violence at present should have a positive effect not only for the current generation of women, but also for future generations.