A Parsimonious Experimental Test for Distributional Concerns

Evidence collected by psychologists and experimental economists suggests that in economic decisions many humans are motivated by items beyond material self-interest. Typical examples for such items are other people`s well-being (altruism, inequality aversion, spite, envy), others` intentions (reciprocity and similar motives), others` expectations (guilt aversion), or others` types. The proposed research project focuses on the first mentioned subclass, i.e. on `distributional preferences`, where beyond one`s own material payoff, the (material) well-being of others affects an agent`s "utility". Distributional preferences have been shown to be behaviorally relevant in important market and non-market environments - there is a vast body of literature on this. This research project adds to this literature in several dimensions. First and most importantly (in Contribution 1) it introduces and experimentally implements a new allocation-choice test that allows to identify the type and the intensity of distributional concerns at the individual level. Major advantages of the test over previous ones are that it is (i) simple and short as the subjects` task is merely to make 6- 10 binary decisions without feedback; (ii) parsimonious as it relies on a small set of plausible assumptions; (iii) general as it directly tests the core features of different types of distributional preferences rather than concrete models or functional forms; (iv) flexible as test size and test design can easily be fine-tuned to the research question of interest; (v) precise as it identifies the archetypical distributional types with unprecedented precision and allows to measure the intensity of distributional concerns at any desired accuracy; and (vi) minimizing experimenter demand effects as subjects are asked to make binary decisions in a neutral frame and do not have the option to do nothing. Other aims of the proposed research project are (in Contribution 2) to investigate how stable distributional preferences are - an issue still largely unresolved in the literature so far; (in Contribution 3) to extend and refine the test in several dimensions, where each extension aims at answering a different research question; (in Contribution 4) to apply the test to detect relations of distributional preferences to other forms of other-regarding preferences (like reciprocity or guilt aversion); (in Contribution 5) to apply the proposed test (together with tests for risk and time preferences) in experiments with large demographic variation (age, gender, income, education) to detect patterns and correlations; and (in Contribution 6) to implement the test in a between-subject design using alternative protocols to shed light on important methodological questions (e.g. the effects of single-blind vs. double-blind protocol; the effects of fixed-role vs. random-role vs. double-role assignment; etc). The current project has important implications for the field and beyond: In experimental economics the proposed test has the potential to become a standard tool to control for subject pool effects and to help to interpret data from other (unrelated) experiments. In economics, the proposed project might help to design better institutions. And beyond economics, it might be useful to address important research questions in biology and psychology as, for instance, `What determines human altruism (or spite)?`, `What distinguishes human altruism form animal altruism?`, or `What drives altruistic punishments and rewards?`

Evidence collected by psychologists and experimental economists suggests that in economic decisions many humans are motivated by items beyond material self-interest. A factor that shapes the decisions of many actors is the consequence of the decision for other people's well-being. If the (material) well-being of others affects an agent's "utility" then we say that the decision maker has distributional preferences. This research project contributes to the vast body of literature on the impact of distributional preferences on economic decisions in several important ways. A central contribution is the introduction of a geometric delineation of distributional preference types (selfish, altruistic, inequality averse, etc.) and a nonparametric approach for their identification at the individual level. The proposed procedure has many advantages over previous ones and therefore might have the potential to become a standard tool in experimental economics to disentangle the impact of distributional preferences from that of other factors thereby helping to interpret data from other (unrelated) experiments. An important follow-up study proposes and implements an experimental design that allows identifying the impact of distributional preferences on the provision behaviour of experts in a credence goods market. A core characteristic of credence goods (like medical treatments and car repairs) is that experts are better informed about the appropriate quality than customers, which leaves scope for fraud. An important insight of that study is that instead of choosing doctors, mechanics or computer specialists exclusively according to their training, customers on markets for credence goods should worry more about the attitudes of these experts towards their customers. Another important study combines theory and lab experiments in an investigation of the importance of different motives for trustworthiness. Specifically, the paper introduces formal definitions for the motives that could induce trustworthiness in a trust situation and then uses economic experiments to identify the most important drivers for trustworthiness. An important insight from that study is that distributional preferences and vulnerability-responsiveness are important for trustworthiness while reciprocity and efficiency concerns are less important. A further study first shows theoretically that convex distributional preferences generate peer effects in choices under uncertainty even when there are no material externalities, no stochastic dependence of lotteries, and no information on outcomes; it then confirms the theoretical findings experimentally. This study is important because it provides a new explanation for the empirical observation that people tend to follow their peers in making risky choices. A further study compares experimentally the revealed distributional preferences of individuals and teams in allocation tasks. It shows that teams are significantly more benevolent than individuals in the domain of disadvantageous inequality while the benevolence in the domain of advantageous inequality is similar across decision makers. A further study experimentally investigates the relationship between distributional preferences and lying aversion at the individual level. A major finding of this study is that altruists lie less when lying hurts others but there is no evidence in the data that supports the hypothesis that altruists are simply upright persons who do not only care about the well-being of others but are also more averse to lying than others