Consistency of Point Estimates and Subjective Distribution of Inflation Expectations

The issue of reporting errors in the survey results of inflation expectations should hardly be ignored. Recent studies focus on the differences in survey responses depending on variations in the survey questionnaire. However, no consensus has been reached with regard to the appropriate way to ask people about their inflation expectations. This paper analyzes the possible impacts of survey questions and their responses, through implementing an inflation expectation survey by asking both point estimates and subjective distributions as well as by means of executing economic experiments. Around 50% of consumers surveyed provided consistent responses to the two questions, while their point estimates were greater in their levels, variances, and reporting errors than was the mean of the subjective distribution. We also found that the average level of these consistent responses was close to the realized inflation level, with small variances. The responses to both types of questions are modified through economic experiments in a consistent manner with economic theory. We conclude that both questions, particularly the one on the subjective distribution, are useful as they allow researchers to check the consistency of the responses.