Presidential Elections: An Explanation of Voting Defection

This paper proposes and tests a modest theory of voting defection, the act of voting contrary to party identification. The relevance of voting defection to popular control of government is clear. Except for the infrequent elections that Key calls “critical,†the identification of people with their parties is very stable over time. The percentage of Democrats and Republicans in the electorate changed only slightly in the four presidential elections from 1952 to 1964. Short term shifts in public attitudes, then, are reflected not in shifts in the distribution of party identification, but in the degree that people vote in accordance with their identification. When they are disenchanted with the President, defection will be high among members of the opposition party and low among members of the party in office. In 1952 people were weary of the Korean War; this weariness was apparent in the massive numbers of Democrats who thought Eisenhower a man capable of ending the conflict and who backed up their convictions with Republican votes. In short, in the rate of defection we have a mirror that reflects popular discontent with the politics of the President. My present concern is to explore some personal attributes of the voters who make up this critical electorate, to augment the propositions surrounding party identification with one explanation of why it is that people vote contrary to their party allegiance.