Stability and Explanatory Power of Inequality Aversion – An Investigation of the House Money Effect

Theories of fairness preferences have gained remarkable attention throughout much of recent economic literature. Formal models have been proposed which are able to explain behaviour that is yet unexplained by the classical model of the strictly egocentric economic man ("homo oeconomicus"). These models have been tested in laboratory experiments. In another, seemingly unrelated, strand of experimental literature the focus lies on the influence of the origin of money, disposed of by subjects in economic experiments on the subjects' behaviour. It has been frequently found that it makes a difference whether the money comes in form of windfall gains, granted by the experimenter ("house money"), or if it is a form of compensation for "real efforts", exerted by the subjects during the experiment itself. An important open question is how these two strands of research fit together. How does the house money effect influence fairness preferences revealed in the lab, and their explanatory power for individual behaviour in other games? If there is a significant influence to be found, the origin of the money clearly deserves to be a part of modern theories of individual preferences. This paper is dedicated to answering the above question. For this purpose, we experimentally elicit subjects' fairness preferences controlling for the origin of the money and test the theoretical predictions for individual behaviour in a social dilemma situation. As a representative for theories of fairness preferences, we chose the model of inequity aversion by Fehr and Schmidt (1999). Our results indicate that individual inequality aversion is not generally robust to the way endowments emerge. Overall, we observe a low predictive power of the theoretical model which is significantly affected by the way the endowment in the preference elicitation games emerges. In particular, the theoretical model has only predictive power for individual behaviour in selected cases when the endowment is house money. As soon as the endowment for preference elicitation has to be earned, the predictive power disappears. Therefore, future experimental research into fairness preferences and their relevance for individual behaviour in many economic areas has to consider the origin of the monetary endowment