Inequality Aversion and Stochastic Decision-making: Experimental Evidence from Zimbabwean Villages after Land Reform

Inequality considerations are a motive for making positive offers in the Ultimatum Game and rejecting small ones, but decision error could have the same effect. I find evidence for both of these considerations and a different relative importance amongst Zimbabwean villagers, of whom some resettled after a government organized land reform during the 1980s. Resettled villagers have higher inequality aversion and lower decision error than those who live in traditional villages but, after accounting for different levels of inequality aversion, the difference in decision error between both groups of villagers is no longer significant. There are no gender differences in preferences. The model estimated was first used by De Bruyn and Bolton (2004) on a large set of bargaining data but the best fit of 64 percent overall coincidence of observed and predicted behavior is achieved for a different `symmetric` specification of inequality aversion in the model. As the use of field data is a recent development in experimental economics, I reestimate the model applied to the Zimbabwean data on the laboratory Ultimatum Game data of Roth et al. (1991) and further field data from Henrich et al. (2005). Estimates are compared comprehensively.