Measuring Expectations of Inflation : Effects of Survey Mode, Wording, and Opportunities to Revise

Several national consumer surveys aim to elicit consumers' inflation expectations. Median reported expectations have been shown to track objective inflation estimates over time, although respondents display relatively large disagreement. Observed medians, however, tend to differ between consumer surveys, possibly reflecting survey design differences. In this paper, we examine the importance of three survey features in explaining these differences: question wording (‘prices in general' vs. ‘inflation'), interview mode (face-to-face vs. web), and the explicit opportunity to revise responses. We find systematic effects on item non-responses, reported inflation expectations and their dispersion. We discuss implications of our findings for survey design