Do Surveys Provide Representative or Whimsical Assessments of the Economy?

We argue that survey responses to economic evaluation questions exhibit instability and can be affected by fairly trivial changes in questionnaire wording. Our analyses make three empirical contributions to this area of survey research. First, we demonstrate that within the course of the interview there is considerable instability in economic evaluations. Second, one source of this instability is cues regarding economic performance, such as those provided by the media. We find that respondents can be persuaded to change their economic evaluations if they receive contradictory cues. Finally, we demonstrate that question placement can affect economic evaluations. More specifically, we demonstrate that proximity to political questions can contaminate economic evaluations. If economic evaluations closely follow political preference questions, respondents have a tendency to give economic responses that are “consistent” with their political responses. Our empirical analysis is based on economic evaluations of respondents to the Hungarian Markets and Democracy Survey administered during December 1997.