I was happy to hear today that a paper of mine, “Directed Giving: Evidence from an Inter-Household Transfer Experiment,” has been accepted at the Journal of Economic Behavior and Organization. My co-authors on the paper are Catia Batista (Nova University, Lisbon) and my former colleague Dan Silverman (Arizona State). The working paper version is here.
We ran a lab-in-the-field experiment in urban Maputo, Mozambique, among clients of a microfinance bank. We had each study participant designate a “counterpart,” the closest person to them outside their own household. (Most people named a sibling, friend, cousin, child, or parent.) Then we had participants play “dictator games,” a common type of lab experiment in economics where the respondent splits a certain “pot” of money between themselves and someone else (for keeps!). The amounts that the participants had to split ranged from $10-$40 (substantial sums in Mozambique.) We found that participants shared more of their “pot” with their counterpart when they had the option of sharing the resources in-kind (in the form of items that they chose) instead of just in cash. In other words, when participants had the option of making a gift to their counterpart in the form of goods, they shared more overall than if they could only share in cash.
To be specific, participants shared about 40% of the pot when they could only share in the form of cash with the counterpart. But when participants could share in the form of goods as well as cash, they shared 45% of the pot (and the difference is statistically significant).
Participants could choose to share whatever goods they wanted with counterparts (we would do the shopping for them and deliver the goods), and were most interested in sharing building materials…
What do we take away from these results? When people think about making gifts or transfers to others, they care about what specifically gift-recipients are consuming (the composition of their consumption), not just about their level of consumption. Gift-givers may get more utility from giving when they know recipients are going to consume certain specific things.
An implication is that having an ability to control what gift-recipients consume may lead to more giving. I consider this issue a lot in other research of mine among migrant populations, where I look at the impact of giving migrants more control over how remittances are used in recipient households. See, for example, recent papers of mine on channeling remittances to education in the Philippines, or getting remittance-recipient households to save more.