This paper considers whether an amendment to state divorce laws that strengthens its joint custody preference operates as a traditional default rule, specifying what most divorcing couples would choose or as a penalty default rule the parties will attempt to contract around.

While the Oregon statutes that frame our discussion here, like most state laws, do not state an explicit preference for joint custody, shared custody is certainly encouraged by Section 107.179, which refers cases in which the parties cannot agree on joint custody to mediation and by Section 107.105, which requires the court to consider awarding custody jointly. In addition, 1997 legislation noted in its very first section that it was state policy to [a]ssure minor children of frequent and continuing contact with parents who have shown the ability to act in the best interests of the child. The effect of this legislation was to strengthen the power of noncustodial parents, since denial of access to the children would give the right to terminate spousal or child support, change the parenting plan, or obtain an award for makeup visitation. The legislative history for the bill shows that it was a compromise between men's rights groups and those concerned about domestic violence.

After setting out the problem and describing the legislation in some detail, the paper tests whether the change in the Oregon statutes is to what most people would want (in which case there should be a substantially higher percentage of joint custody awards after its enactment than before). If the legislation functions as a penalty default, there should be more mediation after the statute and more filings of domestic abuse petitions to avoid application of the rule. If it does not function as a default rule, one would predict an increase in various kinds of transaction costs, including more court filings generally. We might also find that the legislation acts to disfavor primary custodial parents (largely women) as Mnookin and Kornhauser's analysis would suggest (in which case there should be lower child support or property judgments than before enactment). What we find is some evidence of all these results, with quite strong effects even though this event study is hardly ideal since the 1997 changes in law were subtle and largely procedural.