Framing effects in consumer expectations surveys

Abstract

In a randomized survey experiment, I test how variations in question wording and format influence response behavior. Participants from a representative sample in Germany are divided into four groups, each receiving a different version of a question about expected inflation over the next 12 months. I compare two wordings used in leading consumer surveys: (i) the change in prices in general and (ii) the inflation rate. Additionally, I examine responses to a question about probabilistic beliefs regarding future inflation and a simpler one asking for the expected minimum, maximum, and most likely inflation rate. The findings show that response behavior varies significantly with framing. Simpler wording like ‘prices in general’ and less restrictive format produce higher mean expected inflation. While simpler wording leads to higher individual uncertainty, asking for the minimum, maximum, and mode yields lower uncertainty. The results suggest that framing in consumer expectations surveys can shape the elicited data, underscoring the importance of careful question design.