Search

In 2010 and 2011, highly visible national surveys documented frequent failure among the public, especially among Republicans, to acknowledge that Barack Obama was born in the United States. However, different questions yielded strikingly different results. The highest rate of partisan division was generated by a CBS/New York Times closed-ended question that included potentially leading introductory sentences. The smallest partisan gap in apparent misinformation was generated by an ABC News/Washington Post open-ended question that did not ask a follow-up that was needed to gauge public beliefs about whether Mr. Obama was born in the United States. This paper reviews the polls on birtherism and describes an experiment embedded in a nationally representative sample survey testing whether methodological features of these two questions might have distorted their results and caused the apparent discrepancy between them. A version of the closed-ended question including the leading introductory sentences yielded a much larger degree of apparent partisan division than did a version of the question without the introductory sentences. Following the open question with another question that clarified people’s beliefs (asking about whether Hawaii was part of the United States at the time of Barack Obama’s birth) did not alter the conclusions supported by that item about the accuracy of public understanding. This study therefore illustrates how different question wording caused different polls to produce notably different results and provides a caution about the use of potentially leading wording. [ABSTRACT FROM PUBLISHER]

This paper explains the surge and decline of presidential approval following rally events. We develop a theoretical framework grounded in psychology and political entrepreneurship to explain not only who joins and leaves the rally, but why. We argue that threatening events such as 9/11 heighten the link between national identity and presidential approval, but this link fades as political entrepreneurs make partisan identities more salient. We find support for these propositions using analysis of pooled cross-sectional survey datasets and content analysis of elite discourse. [ABSTRACT FROM AUTHOR]