By Mike Dorf
The juxtaposition of political yard signs with skeletons and jack-o-lanterns got me wondering whether there is a "Halloween effect" that arises from the close temporal proximity of said holiday and our general elections. Herewith a few thoughts:
1) As an essentially pagan holiday, Halloween creates anxiety for the most religious elements of our society, especially evangelical Christians, many of whom view it as a celebration of witchcraft, the occult, and devil-worship. These are strongly negative associations (at least when Christine O'Donnell isn't on the ballot), and so, I wonder whether the activation of such fears so close to the election mobilizes turnout of a constituency that, over the last couple of political generations, has tended to vote Republican. I leave to those with better empirical skills the task of figuring out how to test the possibility of a Halloween effect.
2) While I'm speculating about the Halloween effect, I may as well raise the question of whether the Presidential elections are susceptible to an "Olympics effect." In the U.S., at least, the Olympics are a time of nationalism bordering on jingoism. (U-S-A! U-S-A!). If it has any political impact, the stirring up of such sentiments is likely to benefit political conservatives, who are more comfortable and better at wearing their nationalistic chauvinism on their sleeves than are their liberal/progressive counterparts.
3) The Olympics effect may be mitigated somewhat by the more than two months that passes between the end of the Olympics and the general election, but perhaps not always. A well-timed convention could probably harness Olympic spirit and perhaps derive a bigger "bounce" as a result.
4) It's also possible that the Olympics tilt pro-incumbent rather than (or in addition to) pro-conservative. A good performance by the Americans in the Olympics could create a general sense of well-being that would benefit the incumbent, of whatever party. Conversely, a poor showing in the Olympics could help the challenger in heightening overall malaise and the perceived need for change.
5) Here too, it's quite hard to know how one would test any of this or even what counts as a good American showing in the Olympics. Consider 1972. Mark Spitz won 7 gold medals and set 7 world records in the course of doing so; count that as a positive. The USA basketball team was robbed of a gold medal by horrendous officiating in the championship game against the Soviet Union; that's a negative, which might harm the incumbent (Nixon), but it was cold-war-salient, so it would help the more conservative candidate (also Nixon). And all of the action on the field was overshadowed by the kidnapping of Israeli athletes, which may have been salient to the election, but not qua sports.
6) Or consider 1980. The USA men's hockey team pulled off a "miracle on ice," but that was in February, far enough in advance of the November election to have little obvious effect. (It was also before the winter and summer olympics were staggered to occur in different years.) The USA (and other countries) boycotted the 1980 summer games in protest of the Soviet invasion of Afghanistan; and the Soviet bloc boycotted the 1984 games in retaliation. There ought to be some way to use these natural experiments as a control in the tallying of the Olympics effect.
7) In the speculative spirit here, what else might have an effect on election outcomes? Anticipation of Thanksgiving? The end of daylight savings time? Clearly, there is an ambitious research agenda here for someone with the right tools. Just footnote me if you undertake it.