How Peer Groups Influence Economic Perceptions

We use survey data to demonstrate that individuals extrapolate from their peer groups when forming estimates about aggregate economic outcomes (e.g. the aggregate homeownership rate). In a first approach, we follow the previous literature and construct hypothetical peer groups using information on the location, age and education of respondents. We confirm that the observed homeownership rates in these "constructed peer groups" affect respondents' subjective estimates of national homeownership rates. In a second approach, we extend the previous literature and utilize direct information provided by survey participants about the characteristics of their peer groups. We show that these "self-assessed peer groups" are even better predictors for how survey respondents assess aggregate economic outcomes. We show that the same mechanisms are at work for estimates of the national unemployment rate and exploit variation in respondents' interest in economic matters to demonstrate that agents rely more on peer group information if they are less knowledgeable. Finally, we provide evidence that (biased) subjective perceptions affect individuals' investment intentions.