Macropartisanship: A replication and critique

MacKuen, Erikson, and Stimson (1989, 1992) argue that the aggregate distribution of party identification, or macropartisanship, shifts significantly over short periods in response of changes in presidential popularity and consumer sentiment. Their results are based on a subset of Gallup surveys available from 1953 to 1988 and only those CBS/New York Times surveys conducted during the Reagan administration. We replicate this analysis using a more extensive inventory of Gallup and CBS/New York Times data and find considerably less evidence of partisan fluctuation. The amount of partisan change caused by short-term movements in consumer sentiment and presidential popularity is found to be two to three times smaller than initially reported by MacKuen et al. (1989). Our results indicate that macropartisanship adjusts to short-term shocks in a limited and gradual fashion, consistent with traditional views of partisan realignment.