A Dollar a Day Keeps the Pessimism Away: Ambiguity, Expectations and the Income Distribution.
This paper presents three key findings from consumer survey data. First, it provides evidence of systematic forecasting errors across a range of economic outcomes. These errors suggest that households tend to form pessimistic expectations, which increase as household income decreases. Second, expectations react asymmetrically to news, particularly among lower-income households. Third, consumer sentiment is shown to account for cross-sectional differences in expectations, and this too is strongly related to income. The paper links these patterns to ambiguity aversion, suggesting that lower-income households behave as though more ambiguity-averse. To explain these findings, the paper proposes a theory of expectation formation in which agents incur costs to learn about the informativeness of ambiguous news. The model predicts that large shocks can lead to persistent pessimism, creating the potential for poverty-pessimism cycles. When calibrated to the US economy, the model shows that unexpected inflation shocks have significant long-term welfare effects, larger than those predicted by standard models. A simple transfer policy, tested in the context of the 2021-2022 inflationary episode, is found to mitigate these effects at a relatively low cost. I Link to paper.