Testing Near-Rationality using Detailed Survey Data

This paper considers the evidence of "near-rationality," as described by Akerlof, Dickens, and Perry (2000). Using detailed surveys of household inflation expectations for the United States and Sweden, we find that the data are generally unsupportive of the near-rationality hypothesis. However, we document that household inflation expectations tend to settle around discrete and largely fixed "focal points," suggesting that both U.S. and Swedish households gauge inflation prospects in rather broad, qualitative terms. Moreover, the combination of a low-inflation environment and an inflation target in Sweden has been accompanied by a disproportionately high proportion of Swedish households expecting no inflation. However, a similar low-inflation trend in the United States, which does not have an explicit inflation target, reveals no such rise in the proportion of households expecting no inflation. This observation suggests that the way the central bank communicates its inflation objective may influence inflation expectations independently of the inflation trend it actually pursues.