Various journalistic and academic accounts of the 1988 election suggest that George Bush's victory over Michael Dukakis should be primarily attributed to Bush's advantage in voters' comparative evaluations of the two candidates' personal qualities, or to citizens' demands to hold the line against any tax increase or to white voters' attitudes towards blacks and policies designed to assist blacks. This article, based on the 1988 National Election Study, presents evidence which contradicts all three of these conventional conclusions. Based on a multi-stage explanatory model, the authors emphasize instead the substantial roles in determining individual vote decisions that were played by preferences concerning policy direction, in arenas other than those about taxes or race, and by voters' evaluations of the Reagan administration's performance. The authors also emphasize the difference between distinguishing Bush voters from Dukakis voters and explaining the aggregate outcome of the election. In their analysis, the relatively modest size of George Bush's victory (in comparison with Reagan's margin in 1984) is given added significance by documenting the continuing role played by approval of the Reagan administration's policies, the virtual disappearance between 1984 and 1988 of the Democrats' advantage in party identification among voters and the increase in the preponderance of self-designated conservatives over liberals that took place during the same four-year interval.

Presidential approval ratings are a frequently used barometer of performance and popularity. However, despite recognition of the media age in which we live, little work has examined the impact of television on presidential popularity. Using a 1980 and 1984 television content study, panel data from the 1980 National Election Study and rolling cross-sectional data from the 1984 Continuous Monitoring Study, I compare two American presidents (Jimmy Carter and Ronald Reagan) commonly seen as having different effectiveness on television. While media effects were not uniformly present during the two presidencies, there was some evidence that television harmed popularity, particularly when the content of news stories and commentary turned negative. I conclude by discussing the ramifications of these results for presidential strategies based on ‘going public’.

This research Note has two complementary theoretical objectives. First, we shall attempt to place the form of interest representation and the involvement of interest groups in policy formation known as corporatism – or as democratic, societal, liberal or neo-corporatism – in a broader political context: is corporatism systematically linked with other democratic institutions and processes? Secondly, we shall try to fill a gap in the theory of consensus democracy. This theory holds that types of party, electoral, executive and legislative systems occur in distinct clusters, but it fails to link interest group systems to these clusters.

In this Note we challenge the claim asserted in a 1984 Wall Street Journal editorial that partisan gerrymandering by Democratic-controlled state legislatures is the principal reason for the inability of Republicans to translate their national share of votes proportionally into seats in the US House of Representatives. In contrast to previous work, we show the critical importance of sectional (South/non-South) differences for understanding the dynamics of electoral change at the congressional level. We argue that the inability of Republicans to translate votes effectively into congressional seats is largely a product of wasted Republican votes in the South, although we recognize that a handful of states (e.g., California) are significantly gerrymandered against Republicans, and we also recognize that part of the reason for the present-day Democratic advantage in the House is an incumbency advantage that benefits the party that controls most seats.