The Micro Foundations of Presidential Approval

The “miracle of aggregation” has become the key to understanding how an uninterested and uninformed electorate can produce systematic and responsive presidential approval. This paper, however, presents a set of theoretical considerations, which predict that not only do the least informed contribute to aggregate measures of presidential approval, but that the most and least informed update their attitudes in tandem. To test this hypothesis of uniform opinion change, I analyze subgroup and individual level presidential approval with quarterly, monthly, and daily data. The results suggest a reconsideration of how we think about public opinion in the United States. Although subtle differences exist across information groups, all segments of the public translate economic and political information into their presidential evaluations. Peter K. Enns, Department of Government, Cornell University, [email protected]. A previous version of this paper was presented at the 2007 Annual Meeting of the American Political Science Association, Chicago, IL. I would like to thank Chris Anderson and Greg McAvoy for helpful comments and Matthew Martell and Patrick Healy for research assistance. This research received financial support from the National Science Foundation (Grant 0617156) and the University of North Carolina’s American Politics Research Group. “Do you approve or disapprove of the way [the incumbent] is handling his job as president?” is the longest running and most frequent poll question asked to the American public. And answers to this question exert a tremendous influence on U.S. politics. Shifts in presidential approval ratings help steer behavior and outcomes in all three branches of government (Abramson, Aldrich & Rohde 1995, Brace & Hinckley 1992, Brody 1991, Canes-Wrone & DeMarchi 2002, Edwards 1980, Fiorina 1981, Gronke & Newman 2003, Gronke, Kock & Wilson 2003, Marra & Jr. 1989, Neustadt 1960, Rivers & Rose 1985, Rohde & Simon 1985, Sigelman 1979, Yates & Whitford 1998, Yates 2002). Yet, the linkage between presidential approval and political outcomes may be less democratic than appears. The existing literature suggests that only a small minority of the public—the most informed and politically sophisticated—move presidential approval. Erikson, MacKuen, & Stimson (2002, 107) conclude their analysis of presidential approval by stating, “Although we have presented evidence of an impressive economic intelligence on the part of the U.S. electorate, it is worth repeating that this result depends on the powers of aggregation.” Similarly, recent experimental research finds that the most informed are the most likely to update their presidential evaluations in response to media coverage of an event. Miller and Krosnick (2000, 312) state that for media priming to influence presidential evaluations, “one must have the requisite knowledge to interpret, store, and later retrieve and make inferences from news stories they see, hear, or read.”1 When politicians react to changes in presidential approval, the voice of the most informed may be all that matters. The finding that information elites dominate presidential approval is perhaps not surprising, given the dominant influence of the economy and international events on presidential evaluations (Brody 1991, Erikson, MacKuen & Stimson 2002, McAvoy For additional perspectives on who is most susceptible to media priming see (Hill 1985, Iyengar 1979, Iyengar & Krosnick 1984, Iyengar & Kinder 1987, Krosnick 1990, Krosnick & Brannon 1993, Weaver, Graber, McCombs & Eyal 1981).