The Polls: Policy-Specific Presidential Approval, Part 1

Abstract

Listen

Translate article

Perhaps more than any other issue, presidents are held responsible for foreign policy. The literature on public approval of the president, however, focuses more on the impact of economics on presidential approval than foreign affairs. Two factors have lead to this dominance of economic factors over foreign policy in the study of public approval of the president. First, theories of political economy naturally lead many scholars to investigate the impact of economics on public opinion. Second, data on the economy and public perceptions of the economy are much easier to come by and are superior to that on public thinking about foreign affairs. For example, series exist that track real-world economic conditions, such as inflation, unemployment, and economic growth, all of which have been found to affect trends in presidential approval. Moreover, series tapping the public's perceptions and expectations about the economy also exist. These have been used profitably in several recent studies, providing a more theoretically appealing account of how the economy affects trends in presidential approval (MacKuen, Erikson, and Stimson 1992; Clarke and Stewart 1994; Norpoth 1996). In comparison, most studies use a dummy variable to indicate a foreign policy event to measure the impact of foreign affairs on presidential approval (e.g., Brace and Hinckley 1992; Ostrom and Simon 1985). There are many issues, conceptual and methodological, that such a research design strategy poses. For instance, researchers must determine which event(s) they think may affect public consciousness enough to affect presidential approval and how long that event's impact will last. Second, the use of dummy variables across events assumes equal impact of foreign-policy-related events on presidential approval, although it is reasonable to suggest that events will vary in importance and impact on public thinking. Third, for many time periods, no foreign policy event of magnitude exists. This implies that the public's attitude about foreign affairs does not register on presidential approval, but it might be just as likely that the public rewards the president when the foreign scene is quiet. Rather than such periods' having no impact on presidential approval, these quiescent periods may instead lead to high levels of public approval for the president. Fourth, many studies of presidential approval employ quarterly data, asserting that aggregating monthly approval into quarters helps lessen the impact of measurement error associated with single polls. Quarterly aggregations, however, make using dummy variables to measure foreign policy events questionable because, within a quarter, more than one event may arise or the impact of some events may rise and vanish within a quarter. In the case of multiple events, it may be difficult to determine which event affected approval. In the latter case, we may not detect any statistical impact because the impact vanished before the next quarterly approval reading. This may lead us to erroneously understate the impact of foreign affairs on presidential approval. Fifth, the dummy variable measurement strategy does not inform us about the causal process from (1) the real-world event to (2) public thinking about that event to (3) public assessments of the president. Much like the empirical research on real-world economic conditions, the dummy variable measurement strategy makes assumptions about public awareness and understanding about any particular foreign policy event as well as the impact of that event on presidential approval. This focus on specific events makes it difficult to generate general theoretical propositions about the impact of foreign affairs on presidential approval. For instance, do foreign affairs boost or erode presidential approval? Some studies argue that some events have a positive impact while others will have a negative impact, but few have developed a theory to predict what type of foreign affairs event will help or harm presidential standing with the public (for instance, compare Ostrom and Simon 1985 with Brace and Hinckley 1992). …