In a collaboration with the Federal Reserve Bank of New
York, this project aimed to improve survey-based measures of
consumers’ inflation expectations. For decades, national consumer
surveys of inflation expectations have been conducted in the US and
EU, to inform central banks’ economic policy. Median responses tend to
track actual inflation, and sometimes outperform professional and
model-based inflation forecasts. Yet, even in periods of relatively
stable inflation, responses tend to show large disagreement, with some
individuals giving seemingly large overestimates of inflation. Our
results suggest that question wording partly drives this problem, with
some respondents interpreting the survey questions about expectations
for “prices in general” as asking about “inflation” while others think
relatively more about personal price experiences — with the latter
interpretation being associated with seemingly unrealistic responses.
We conducted experiments in which respondents were randomly assigned
to different question wordings and found that questions asking
directly about “inflation” yielded less disagreement about question
interpretation and more reasonable response patterns. Our results
have implications for the design of economic surveys.
Presented by:
Wandi Bruine de Bruin (Carnegie Mellon University & Leeds University)
Date & time:
November 5, 2012 4:00 pm - May 11, 2012 4:30 pm
External seminars home