Everything is Uncertain and Uncertainty is Everything: Strategic Voting in Simple Plurality Elections

Intuition tells us that strategic voting is most likely in marginal constituencies where the preferred party is a long way behind the second placed parity. Some formal theories suggest there should be complete desertion of all but two candidates (Palfrey 1989), or additionally that the second and third will have similar vote shares (Cox 1997). Unfortunately, these theories fail to account for uncertainty over the strength of candidates. We present a model that allows for such uncertainty. It generates interesting and original comparative statics. All three approaches are tested against English voting data from 1987, 1992 and 1997. Our model fits the data; the standard intuition and Cox hypothesis do not. Thus formal theory can improve on intuition. But, this depends on the realization that voters are uncertain, and it is only uncertainty that matters for strategic voting.