The 1992 election was influenced by a number of factors. The personalities, or characteristics, of the candidates clearly played a role, as did the kind of campaign they ran. Economic conditions were particularly important. However, it is interesting to note that statistical models by economists predicting election results, which emphasize economic trends, were less successful in 1992 than in previous years. Political scientists were more successful.

Models and Perceptions

Four years ago, Ray Fair, an economist at Yale, was hailed for an incredibly accurate prediction. Applying a model that did not include data more recent than six months before the election, he anticipated the outcome of the 1988 contest within .3 of 1%. This model did not include Willie Horton or even Dukakis. It simply analyzed the way in which parameter data anticipated support for the incumbent and the opposition. Other economically based models were also quite accurate.

Successfully applying such models is enough to make journalists and politicians ill, as it suggests that everything they do or write about is useless—that the outcome of elections is determined by pre-existing structural factors, not by candidates, campaign tactics, or ideology. This time Fair was wrong! He predicted that Bush was going to win, since the economy was really not that bad. The 1991-92 downturn was more moderate than the Reagan 1981-82 recession, or the Carter 1977 decline, while the misery index—unemployment plus the inflation rate—was much worse in 1979-80. But Bush was never able to get this message across.