This paper studies the impact of the regime shift from maternal preference to joint
custody, in custody dispute adjudication during the 1980s using the one percent
Integrated Public Use Microsample Series (IPUMS) of the decennial Census for
the decades from 1970 to 1990. We focused on children between the ages of 15 to
18, who were living with a single divorced or separated parent and children of
intact families. Educational attainment was used to quantify child outcomes.
Using cross state and year variation in the timing of adoption of those laws, we
found strong evidence that the children of these single parent households, living
in states which adopted joint custody, had a higher probability of high school
graduation by age 18. On the other hand, we found that children from intact
families suffered a decrease in probability of high school graduation by age 18.
This suggests that the law has important unintended negative effects that had been
thus far neglected. The result on children from intact families was replicated using
the IPUMS Current Population Survey Sample, and results concur with the
findings from the census dataset. The results were also replicated when we relax
the distributional assumption using stochastic dominance techniques.