Did Unilateral Divorce Raise Divorce Rates? Evidence from Panel Data

Using panel data sheds new light on the impact of the no-fault revolution on divorce rates. It shows that the way one controls for geographical heterogeneity affects the estimation considerably, as Peters and Allen already demonstrated. But further, it shows that controlling flexibly for unobserved attributes of state populations that are correlated with both divorce behavior and the divorce law is crucial. This is made possible by using longitudinal divorce data. A panel of state-level divorce rates is assembled, which measures virtually every divorce in the US over the entire duration of the law changes. The estimation reveals a strong unilateral divorce: divorce rates would have been about 6% lower if states had not adopted unilateral divorce, accounting for 17% percent of the overall increase between 1968 and 1988. The current move to tightening divorce requirements can therefore be expected to lower the number of divorces. However, the estimates also demonstrate that other factors in addition to unilateral and no-fault divorce had a great deal to do with the increase in divorces in the US.