Working Paper: NBER ID: w26106
Authors: Olivier Coibion; Dimitris Georgarakos; Yuriy Gorodnichenko; Maarten van Rooij
Abstract: We implement a survey of Dutch households in which random subsets of respondents receive information about inflation. The resulting exogenously generated variation in inflation expectations is used to assess how expectations affect subsequent monthly consumption decisions relative to those in a control group. The causal effects of elevated inflation expectations on non-durable spending are imprecisely estimated but there is a sharp negative effect on durable spending. We provide evidence that this is likely driven by the fact that Dutch households seem to become more pessimistic about their real income as well as aggregate spending when they increase their inflation expectations. There is little evidence to support the idea that the degree to which respondents change their beliefs or their spending in response to information treatments depends on their level of cognitive or financial constraints.
Keywords: inflation expectations; household consumption; randomized control trial
JEL Codes: C83; D84; E31
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Inflation expectations (E31) | Durable goods spending (L68) |
| Inflation expectations (E31) | Non-durable goods spending (E20) |
| Inflation expectations (E31) | Overall household spending (D19) |
| Inflation expectations (E31) | Anticipation of decline in real income (D84) |
| Anticipation of decline in real income (D84) | Durable goods spending (L68) |