A Comparison of the Voting Preferences of Likely Voters and Likely Nonvoters

Abstract

In elections in the United States, even national general elections, relatively large proportions of the population old enough to vote do not do so. In the six off-year congressional elections from 1950 through 1970 inclusive (that is, congressional elections not coincident with presidential elections), the percentage of the population of voting age not voting for House candidates averaged about 56 percent. In the six presidential elections from 1952 through 1972 inclusive, the comparable percentage who did not vote for President averaged about 39 percent. Among these six elections the lowest turnout occurred in 1972, when it appears, at the time of writing, that about 45 percent of those of voting age did not vote for President. Because such a large share of the voting-age population does not vote, election research has been concerned with the problem of identifying likely voters and likely nonvoters in advance of an election. This in turn has raised the question of what a comparison of the voting preferences of likely voters and nonvoters would show, which is the subject addressed by this paper.