Questionnaire Design Effects in Climate Change Surveys

Abstract

Despite strong agreement among scientists, public opinion surveys reveal wide partisan disagreement on climate issues in the United States. We suggest that this divide may be exaggerated by questionnaire design variables. Following a brief literature review, we report on a national survey experiment involving U.S. Democrats and Republicans (n = 2,041) (fielded August 25–September 5, 2012) that examined the effects of question wording and order on the belief that climate change exists, perceptions of scientific consensus, and support for limiting greenhouse gas emissions. Wording a questionnaire in terms of “global warming” (versus “climate change”) reduced Republicans’ (but not Democrats’) existence beliefs and weakened perceptions of the scientific consensus for both groups. Moreover, “global warming” reduced Republicans’ support for limiting greenhouse gases when this question immediately followed personal existence beliefs but not when the scientific consensus question intervened. We highlight the importance of attending to questionnaire design in the analysis of partisan differences.