Directed Giving: Evidence from an Interhousehold Transfer Experiment

Working Paper: NBER ID: w20605

Authors: Catia Batista; Dan Silverman; Dean Yang

Abstract: We investigate the determinants of giving in a lab-in-the-field experiment with large stakes. Study participants in urban Mozambique play dictator games where their counterpart is the closest person to them outside their household. When given the option, dictators do a large fraction of giving in kind (in the form of goods) rather than cash. In addition, they share more in total when they have the option of giving in kind, compared to giving that can only be in cash. Qualitative post-experiment responses suggest that this effect is driven by a desire to control how recipients use gifted resources. Standard economic determinants such as the rate of return to giving and the size of the endowment also affect giving, but the effects of even large changes in these determinants are significantly smaller than the effect of the in-kind option. Our results support theories of giving where the utility of givers depends on the composition (not just the level) of gift-recipient expenditures, and givers thus seek control over transferred resources.

Keywords: interhousehold transfers; in-kind giving; dictator games; Mozambique

JEL Codes: C92; C93; D01; D03; D64; O17


Causal Claims Network Graph

Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.


Causal Claims

CauseEffect
in-kind option (M52)influence on sharing behavior (D16)
desire for control (D91)preference for in-kind giving (D64)
in-kind option (M52)total giving (D64)
cash-only option (E42)total giving (D64)
in-kind option (M52)cash giving (E41)

Back to index