Voluntary contributions to a public good and endowments redistribution : An experimental study.

Summary Does income inequality affect the provision of public goods? Warr established a neutrality theorem in 1983: under certain conditions, a marginal redistribution of income between agents does not affect the quantity of public goods provided by their voluntary contributions. The generalizations of this result by Bergstrom et al (1986) have made it possible to better understand this phenomenon: neutral redistributions are of "small" magnitude, so that agents whose income has been cut always have the possibility of maintaining their expenditure on private goods, and adjustments to individual contributions leave the aggregate contribution to the public good unchanged. Itaya et al (1997) have investigated the welfare consequences of non-neutral redistribution. In the first two chapters of this thesis we test these predictions in the laboratory with a public good game with quadratic utilities. The first chapter considers a "small" redistribution that should not lead to a change in the quantity of the public good. In contrast, in chapter 2, the redistribution is of such magnitude that it affects the quantity of public good supplied and the welfare of society. Although we find some theoretical predictions in the laboratory, notably concerning the modification or not of the quantity of public good produced and of well-being, the predictions concerning individual behaviour and gains are rarely verified. In particular, we observe that, following the modification of their income, some players reduce or increase their contribution less than the theory predicts and that poor agents over-contribute. Finally, it seems that the emergence of inequalities does not affect behavior in the same way as when these inequalities are pre-existing, and thus that the direction of redistribution, depending on whether it creates or reduces inequalities, matters. In Chapter 3, we look specifically at an inequality-creating redistribution in a linear public good game and examine whether men and women respond to this change in the same way and what consequences this has for the quantity of public good provided. We show that when women benefit from the redistribution, the quantity of public good produced decreases. It also appears that behavior is modified when subjects know the gender of those who have become richer.