Nonresponse Bias in Household Inflation Expectations Surveys

Abstract This paper uses microdata from the Reserve Bank of New Zealand's Household Inflation Expectations survey to evaluate the effects of nonresponses to the inflation expectations question in the survey. We find nonresponses lead to substantial underrepresentation of some demographic groups in the survey: young, female, low‐income, and minority ethnic groups have lower response rates. How the survey is conducted also affects item response rates. The survey response rates increase when the survey is conducted online and when inflation rates deviate from the central bank's target range. Using a sample selection model, we assess whether the survey has item nonresponse bias by comparing the demographic characteristics of responders and nonresponders. After accounting for selection, we find that observed differences in inflation expectations by gender, ethnicity, and income decrease substantially, while differences by age increase. We quantify and demonstrate how to adjust average aggregate and subgroup inflation expectations for bias caused by item nonresponse. We show that there is a positive bias, and the aggregate and subgroup inflation expectation series shift down after the adjustment. We also show that inflation expectations disagreement, both across and within subgroups, tends to decrease with the correction for nonresponse bias. These findings have important implications for survey design and monetary policy communication.