What Citizens Know Depends on How You Ask Them: Experiments on Time, Money and Political Knowledge

Surveys provide widely cited measures of political knowledge. Do unusual aspects of survey interviews affect these measures? An experiment on a nationally representative sample of over 1200 Americans provides an answer. Respondents are randomly assigned to one of four groups. A control group answers questions in a typical survey context. Respondents in three treatment groups are given a longer window of time in which to answer questions, a small monetary incentive for answering questions correctly, or both. These variations increase performance significantly for almost every knowledge question we asked. Overall, average knowledge scores in the treatment groups are 11-24 percent higher than in the control group. The treatments also cause significant reductions in the magnitude of respondents errors on open-ended questions. The findings imply that new elicitation strategies can improve our understanding of what citizens know about politics and other socially relevant phenomena. 2 A basic premise of democratic governance is that citizens know certain things about politics. For this reason, scholars and other political observers devote considerable energy to th