We experimentally study the impact of inequality on the effectiveness of contests for funding public goods in a development context. We observe that the typical result of a lottery funding mechanism leading to greater funding for the public good than predicted by theory extends to groups with inequality. However, while theory suggests that increased inequality should lower total contributions to a lottery funded public good, we observe the opposite pattern. This result differs from prior results for the standard voluntary contribution mechanism where increased inequality has been found to reduce public good provision. Moreover, we find that the poor do not contribute a greater share of their endowment to the public good than do the wealthy. Thus, overall our study demonstrates the potential for community development projects, when funded with a lottery mechanism, to be highly successful even in the presence of inequality and may facilitate a progressive redistribution of wealth.

This paper takes the advantage of two sets of bidding data from eBay on charity and non-charity of otherwise identical products and nonparametrically identifies bidders’ public goods values within an independent private values paradigm. Specifically, in the charity auction set, we develop structural altruistic estimators that could be used to back out each bidders latent private value, a joint value of the auction item and the linked charity/public good. Using the non-charity auction set, we identify each bidder’s latent private value on auction items. Under the assumption that both charity and non-charity auction sets have the same underlying distribution of private values on the auction item, the difference between the latent private values from charity and non-charity auction sets is the value of linked charity. Through this approach, we find that people’s valuations on charity are driven by both pure altruism and warm glow.

Given the threshold role of the unit elastic in determining whether the current tax policy is effective in terms of encouraging charitable giving, clear conclusion on the relationship between the estimated price elasticity and unit elastic is required. However, studies on this topic report mixed results in terms of the magnitude of priceelasticity. To investigate the heterogeneity between studies on estimating the price elasticity of charitable giving, we identify and synthesize 81 studies with 113 observations. By doing so, we updated the robust estimation of the elasticitie. We also contribute a better understanding the current mixed results on this topic by exploring the potential moderators. These moderators in turn shed light on the future research direction.