Working Paper: NBER ID: w22977
Authors: Partha Deb; Christian A. Gregory
Abstract: We study the effects of SNAP participation on food insecurity and food spending using finite mixture models that allow for a priori unspecified heterogeneous effects. We identify a low food security subgroup comprising a third of the population for whom SNAP participation increases the probability of high food security by 20-30 percentage points. There is no affect of SNAP on the remaining two-thirds of the population. SNAP increases food spending in the previous week by $50-$65 for a low modal spending subgroup comprising two-thirds of the population, with no effect for the remaining third of the population.
Keywords: SNAP; food security; food spending; heterogeneous effects; finite mixture models
JEL Codes: D12; I38
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| SNAP participation (H53) | high food security (Q18) |
| SNAP participation (H53) | food spending (D12) |
| SNAP participation (H53) | food security (Q18) |