Working Paper: NBER ID: w26643
Authors: Mark A. Aguiar; Mark Bils; Corina Boar
Abstract: Many households hold little wealth. In standard precautionary savings models these households should not only display higher marginal propensities to consume (MPCs), but also higher future consumption growth. In contrast, we see from the PSID that such “hand-to-mouth” households do not display higher growth in spending. They also exhibit greater volatility of spending and adjust their spending to a greater extent through the number of categories consumed. Consistent with a role for preference heterogeneity, the panel data show that it is persistent differences across households, not current assets, that predict low consumption growth and other spending differences for the hand-to-mouth households. To identify the extent of preference heterogeneity, we consider the model of Kaplan and Violante (2014) with both liquid and illiquid assets, but allow heterogeneity in preferences. To match the data, many poor hand-to-mouth must be relatively impatient and have a high inter-temporal elasticity of substitution (IES). The model shows that preferences predominantly explain the higher MPCs for low-asset households. Preference heterogeneity notably increases the spending impact of fiscal transfers, but only if targeted, while reducing that from interest rate cuts.
Keywords: household consumption; preference heterogeneity; hand-to-mouth households
JEL Codes: E21
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| h2m households exhibit greater volatility in spending (D12) | volatility in spending (E62) |
| spending impacts of fiscal transfers are significantly amplified when targeted towards low-asset households (H31) | spending impacts of fiscal transfers (E62) |
| hand-to-mouth (h2m) households do not exhibit higher consumption growth (D10) | consumption growth (E20) |
| persistent low wealth status (I32) | lower expected consumption growth (F62) |
| preference heterogeneity among h2m households explains their higher MPCs (D11) | higher MPCs (E49) |
| lower targeted asset levels (G19) | higher MPCs and lower consumption growth (E21) |