Cost and the calculus of voting

Downs's (1957) theory of voting maintains that individuals balance the costs of voting against anticipated benefits in deciding whether to vote. However, most empirical tests of his theory have concluded that costs play little role in individuals' decisions to vote or abstain, and that benefits are the determining factor. Unfortunately, the existing empirical tests of the theory have been inadequate, especially in regard to the measurement of the cost of voting. Using data from the Comparative State Elections Project, we develop an improved indicator of the cost of voting. When this measure of cost is used in a test of Down's theory, we find, contrary to most earlier research, that the cost of voting seems to be a more important determinant of participation than the factors associated with voting benefits in Downs's model.The authors shared equally in the research reported; the order in which they are listed was determined randomly.     

Some effects and noneffects of campaign commercials: An experimental study

This paper addresses the question of how TV commercials can be used to build support for a candidate. Employing experimental methods, the authors tested the independent effects of several advertisement variables. Spot advertisements of an actual congressional candidate were produced by the authors and shown to groups of viewers who had no other information about the candidate. Evaluation of the candidate differed little among experimental groups, and the lack of extreme differences suggests that the extent to which voters can be manipulated by manufactured images is severely limited. The implications of these findings for candidates and campaign professionals are discussed.     

'These boots are made for walking': why most divorce filers are women

Because of the financial and social hardship faced after divorce,most people assume that generally husbands have instigated divorcesince the introduction of no-fault divorce. Yet women file fordivorce and are often the instigators of separation, despitea deep attachment to their children and the evidence that manydivorces harm children. Furthermore, divorced women in largenumbers reveal that they are happier than they were while married.They report relief and certainty that they were right in leavingtheir marriages. This fundamental puzzle suggests that the incentivesto divorce require a reexamination, and that the forces affectingthe net benefits from marriage may be quite complicated, andperhaps asymmetric between men and women. This paper considerswomen's filing as rational behavior, based on spouses' relativepower in the marriage, their opportunities following divorce,and their anticipation of custody.     

Sex Differences in Self-Concept and Symptoms of Depression During the Transition to College

In an investigation of sex differences in adaptation to college, real and ideal self-concept and symptoms of depression were studied longitudinally in a sample of 287 students. Survey data were collected at a summer orientation and one semester into freshman year. No sex differences in self-concept were found before college, but males' real self-concept became more positive over the transition. Females were more depressed than males at both times, although depressive symptom scores increased in both sexes. Real self-concept scores were negatively correlated with depressive symptoms in both sexes at both times, while the discrepancy between real and ideal self-concepts was positively correlated with depressive symptoms among females before college and in both sexes midway through freshman year. A one-year follow-up revealed that females' real self-concept scores increased to match those of males by mid-sophomore year. These sex differences are discussed in relation to psychological development during adolescence.     

Reconsidering pocketbook voting: An experimental approach

Despite its deep theoretical roots, the idea that voters reward or punish incumbents in national elections for trends in their personal financial circumstances has not fared well when subjected to empirical test. This paper poses an experimental test of the leading explanations for the surprisingly weak showing of pocketbook influences on vote choice. According to certain of these explanations, the answer lies in distinguishing between sociotropic and self-interested economic voting, or between retrospective and prospective economic voting, or between perceptions of economic trends in general and perceptions of the electorally relevant component of these trends. However, expectations based on these explanations are generally not borne out in the laboratory setting. Consistent with the observed pattern of effects, however, is the idea that pocketbook voting displays little independent impact because economic perceptions and attributions are epiphenomena—strongly biased by the voter's preexisting political commitments.     

Public opinion and presidential responsibility for the economy: Understanding personalization

The general mechanisms which underlie the psychological phenomenon ofpersonalizing (cognitive simplification and defensive attribution) would seem to have great utility in explaining attributions of presidential control over the economy. Yet previous research has generated inconsistent and inconclusive empirical results. This study identifies several sources of inconsistency and then attempts to clarify the approach by focusing separately on the object and the subject of personalizing. Our findings suggest that the determinants of personalizing to the president are different from the factors that explain personalizing to the incumbent. In addition, we find that the impact of the two psychological mechanisms differs substantially within economic subject areas.Paper prepared for delivery at the 1984 Annual Meeting of the Southwestern Political Science Association, Fort Worth, Texas.     

Publication Bias Reconsidered

In political science and many other disciplines, statisticallysignificant results—rejections of the null hypothesis—areachieved more frequently in published than in unpublished studies.Such "publication bias" is generally seen as the consequenceof a widespread prejudice against statistically nonsignificantresults. I argue that evidence of such a prejudice is in surprisinglyshort supply and that publication bias can occur even in theabsence of such a prejudice and even if the review process isfunctioning perfectly. More importantly, publication bias maystem from dutiful application of standards of scientific inquiryrather than from irrational prejudice.     

Anticandidate voting in the 1984 presidential election

Data from a national survey conducted in 1984 form the basis for a new analysis of anticandidate voting in presidential elections, i.e., voting focused more on a candidate one opposes than on a candidate one prefers. Anticandidate voting is viewed as the end product of a process whereby voters attempt to reduce discomfort that cross-pressures generate within their decision frameworks. In 1984, nearly a third of all likely voters said they were primarily motivated by a desire to voteagainst one of the two presidential candidates, a rate of anticandidate voting similar to that observed in the Johnson-Goldwater election of 1964 but well below that of the 1980 Reagan-Carter election. However, factors related to anticandidate voting in the past were not consistently linked to anticandidate voting in 1984. We conclude that the presence of Ronald Reagan exerted such a strong influence on the 1984 campaign that processes that would normally be observable, such as anticandidate voting, were overridden.