The single subject requirement is a technical rule that is often used to invalidate voter initiatives, either before or after they go to the ballot. Critics of the rule have claimed that it cannot be enforced in an objective, consistent way because the definition of a “subject” is infinitely elastic. Our article, based on our examination of 765 judicial votes in single subject cases during the period 1997–2006, strongly supports the critics: in states with aggressive enforcement of the single subject rule, judicial decisions are strongly predicted by whether or not a judge is likely to agree with the substance of the initiative under review based on his or her partisan affiliation.
The novelty of this evidence is that the influence of a judge’s political preferences grows as enforcement of the rule becomes more aggressive, and that the effect is huge — perhaps unprecedented in the literature on judicial voting which typically finds modest effects. Our evidence does not imply that partisan decisionmaking is deliberate or inherent to the judging process, but rather that it is an outcome that emerges when judges are put in position where neutral principles are not available to guide their decisions.
The question, in light of this, is where do we go with the single-subject rule? One approach might be to try to design more objective rules for counting the number of subjects, as recently attempted by Robert Cooter and Michael Gilbert. (See also our short piece critiquing Cooter and Gilbert, and their reply.)
We are glad to see creative efforts that go down this path, but are pessimistic about its likelihood of success. Judges and legal scholars have tried for decades to come up with objective tests, but our evidence shows that these efforts have failed to produce a doctrine that can be enforced in a neutral and consistent manner, and indeed, existing attempts to discriminate have led to decisionmaking that looks increasingly partisan.
We are also pessimistic about efforts to craft a “better” single subject rule because it is not even clear that restricting initiatives to a single subject is a good thing. We noted in our initial post that existing political economy theory does not support the claim that logrolling is always a bad thing. And it seems noteworthy that there is not a single empirical study to our knowledge that shows the public has been harmed from multiple-subject initiatives.
One possibility is to abandon the single-subject rule. The politicization of judging that accompanies aggressive enforcement of the single subject rule undermines the rule of law. To the extent that decisions depend on the identity of the judges that hear a case, initiative proponents face a form of judicial roulette, which might deter them from paying the high costs to initiate a measure in the first place. Some critics might not mind this outcome, but the single-subject rule is intended to perfect the initiative process rather than suppress it.
By discouraging sponsors from offering the people choices at the ballot box, the electorate will end up with fewer options, and as suggested by a stream of research, the resulting policy choices will be less in line with the will of the majority. Politicization of the rule also threatens to undermine the direct democracy process itself by undermining the belief that the initiative process is equally available to people of all political stripes.
If the single subject rule is to be retained, as there is every reason to believe it will be, our article suggests that neutrality and consistency would be better advanced by adoption of a restrained or deferential posture. In some sense, this would bring the single subject rule for initiatives into conformity with legislative practice. Most states have some form of a single subject rule for state legislatures. Judges, wisely in our opinion, have tended to approach that requirement deferentially, in part because of a recognition that logrolling is not always a bad thing, and in many situations is necessary to form coalitions that allow legislation in the public interest to be advanced.
We believe a deferential approach to the single subject rule is a prudent course of action at this point, given the current lack of objective standards that can discriminate between good and bad logrolls. The dangers that the single subject rule is purported to address are exaggerated, and the hope of alleviating these modest dangers is unlikely to outweigh the costs of aggressive enforcement.