Version of 2/11/2006 1:20:00 AM Do Donors Care About Subsidy Type? An Experimental Study

An individual should be indifferent between a rebate subsidy of rate sr and a matching subsidy of rate sm = sr/(1 sr), and the total amount received by the charity should be the same regardless of subsidy type. Recent laboratory and field experiments contradict these straightforward predictions of standard economic theory: subjects consistently make decisions that result in larger amounts going to the charity under a matching subsidy than under an equivalent rebate subsidy. This paper tests whether this result is due to rebate-aversion – a preference by donors for a match over a rebate subsidy. Consistent with theory we find no significant preference for one or the other subsidy scheme. We do however find that, as in previous studies, participants selecting the matching subsidy made decisions that resulted in approximately twice the donations of participants selecting the rebate subsidy donated. DO DONORS CARE ABOUT SUBSIDY TYPE? AN EXPERIMENTAL TEST* Experimental and field studies of charitable giving conducted over the past several years consistently show that a rebate subsidy leads to lower levels of overall donations (the amount received by the charity) than an equivalent matching subsidy, despite the fact that both embody the same incentive structure for donating. A rebate subsidy refunds a portion of the contribution to the giver, thereby lowering the effective price of giving. The tax deductibility of charitable contributions under the U.S. Federal income tax system is an example. Matching subsidies are frequently offered by private employers, who agree to send a matching amount for every dollar contributed to a charitable organization by an employee. It is straightforward to construct equivalent subsidies -a rebate subsidy of rate sr and a matching subsidy of rate sm= sr/(1 sr) -that result in the same incentive structure for potential contributor. Nevertheless, the response of givers appears to vary with the form of the subsidy. This paper is one in a series that explores this somewhat puzzling result. In the initial laboratory study (Eckel and Grossman, 2003), we reported results from experiments that directly compared several different subsidy levels. Our data indicated that contributions were significantly higher with matching subsidies than with equivalent rebate subsidies. Participants in that study otherwise responded to the subsidies as expected, giving more to charity when the price of giving was lower and when their endowment in the experiment was larger. Indeed, estimated income and price elasticities for the experimental data are similar to those found in studies using field data. However, the total contributions received by the charity were 23% 86% larger under the matching subsidy than under the rebate, and this difference increased with the rate of subsidy.