Working Paper: CEPR ID: DP14453
Authors: Dimitris Christelis; Dimitris Georgarakos; Tullio Jappelli; Luigi Pistaferri; Maarten van Rooij
Abstract: We measure wealth effects on consumption using a novel research design: responses to direct survey questions asking how much a household would change consumption in response to unexpected (positive and negative) shocks to own home value. The average wealth effect is in the 2-5% range, in line with econometric estimates that associate changes in housing wealth with consumption realizations. However, our analysis uncovers significant heterogeneity. Extensive margin responses are limited: more than 90% of the sample reports no consumption adjustment to wealth shocks. On the other hand, conditioning on adjusting, intensive margin responses are substantial. Finally, the consumption response to positive wealth shocks is greater than the response to negative shocks.
Keywords: wealth effect; housing; heterogeneity
JEL Codes: D12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| average wealth effect on consumption in response to positive shocks (E21) | consumption adjustment (E21) |
| average wealth effect on consumption in response to negative shocks (E21) | consumption adjustment (E21) |
| cash-on-hand (E41) | consumption adjustment (E21) |
| high loan-to-value (LTV) ratio households sensitive to negative shocks (G51) | consumption adjustment (E21) |
| increases in home values facilitate additional borrowing and spending (G51) | consumption adjustment (E21) |
| decreases in home values do not necessitate reductions in borrowing (G51) | consumption adjustment (E21) |