Bad Weather Lowers Voter Turnout

Abstract
While political experts have long claimed that bad weather lowers voter turnout, the impact of weather on U.S. election outcomes remains unclear. The most rigorous work to date found that precipitation benefits Republicans and suggested that Florida rains influenced the outcome of the 2000 presidential election (Gomez et al., 2007), but a recent analysis finding that precipitation only lowers turnout in uncompetitive election states calls this claim into question (Fraga & Hersh, 2010). Here, we reanalyze the 1972-2000 U.S. presidential elections with a focus on supporters of non-major party candidates, an oft-overlooked contingency. We propose that bad weather affects election outcomes not through its effect on turnout—as has long
While this popular belief made its way into the political science literature as early as the 1960s when “the fisherman who expects to return from the sea before the polls close but is held up by a storm” was cited as one of many plausible reasons why inclement weather might suppress voter turnout (Andrews, 1966), only in recent years have researchers been able to analyze the impact of weather on election outcomes using rigorous empirical methods. Most notably, Gomez, Hansford, & Krause (2007) utilized meteorological data from over 22,000 U.S. weather stations and, using GIS interpolations, produced estimates of Election Day rain and snow for more than 3,000 U.S. counties for 14 presidential elections. In analyzing this novel dataset, the authors found that Republican vote share increases with precipitation, leading them to conclude that “Republicans should pray for rain.” In fact, the 2000 election was so close that their model predicted that the Democratic candidate, Al Gore, would have won if it had rained less in Florida on that Election Day. However, recently this result has been called into question because the model assumed that voter turnout was equally affected by rain in competitive and uncompetitive states. Specifically, Fraga and Hersh (2010) found that while rain leads to lower voter turnout on average, it does not lead to lower turnout in competitive election
To understand, consider the classic “calculus of voting” model, R = PB - C + D, first proposed by Downs (1957) and later adapted by Riker and Ordeshook (1968), where R is the reward gained from voting in a given election, P is the probability of a vote influencing the outcome of the election, B is the differential benefit of one candidate winning over the other, C is the cost of voting (e.g., time and effort spent), and D is the psychological benefit of voting derived from things like feelings of civic duty or altruism. Most scholarship on voting has assumed—either explicitly or implicitly—that inclement weather impacts elections only through increasing C, the actual costs associated with voting (e.g., through causing poor driving conditions), and therefore that inclement weather might only impact election outcomes through its effect on voter turnout. In contrast, we argue that inclement weather also impacts voting behavior through its effect on D, the psychology of the voter, and through the individual’s perception of the values of P and B. Therefore, we contend that precipitation might impact not only voter turnout, but also voters’ choices at the