Social choice theory casts serious doubts on the viability and meaning of democracy. As an example, imagine a group of three friends who want to go on a joint vacation. However, they disagree on three possible destinations: a beach vacation, a city trip, or to go hiking in the mountains. Anne prefers the beach, followed by the city trip and then the mountains. For Paul, a city trip would be the best option. Hiking is his second-best alternative and the beach is his least-preferred destination. Emily favors the mountains followed by the beach and a city trip. Of course, the decision problem of Anne, Paul, and Emily resembles the well-known Condorcet paradox. If they vote on their common destination, every alternative will beat one of the other two alternatives in a pairwise majority vote and loose against the remaining alternative. In this case, voting will not result in a meaningful collective decision, because a majority favors the beach over a city trip, the city beats the mountains, and the mountains are majority-preferred in comparison to the beach.
Several classical social choice theorems like Arrow’s Impossibility theorem have demonstrated the generic instability of majority rule. Hence, in almost all situations, majority decisions exhibit voting cycles like in the example above. In the resulting absence of an unambiguous voting equilibrium any alternative in the policy space can be reached, given the appropriate agenda. From the perspective of social choice theory, democracy could result in highly arbitrary decisions with frequent radical policy changes. Hence, the important question arises how to explain the apparent stability of real democratic decisions.
While traditional explanations highlight the importance of institutions for inducing stability in majority decisions, more recent contributions analyze how social preferences might help to overcome democracy´s theoretical instability problem. Individuals motivated by social preferences such as fairness or reciprocity maximize not only their own well-being, but also take the well-being of other actors into account. Consequently, if group members are motivated by social preferences, alternatives offering a fair distribution of benefits among group members become more popular and provide stable equilibria in otherwise unstable democratic decisions. However, there is little research on the influence of social preferences in majority decision-making.
In my article in Political Behavior I present results from laboratory experiments examining whether social preferences influence majority decisions in committees. In the experiment, five-member groups have to select points from a two-dimensional policy space. Treatments of the experiment systematically vary the fairness properties of the alternatives in the policy space by manipulating subjects’ payout functions. The experiment is designed as a hard test for the influence of social preferences. For one, participants play for money in the experiment. Group members have ideal points in the policy space and earn more the smaller the distance between their ideal point and the outcome chosen by the committee. Secondly, subjects interact anonymously via a computer network preventing direct face-to-face communication among committee members. Thirdly, all subjects are students, most of them students of economics or related fields. All three design features should promote the influence of egoistic motivations and weaken the influence of social preferences in the decisions. However, the results clearly show that distributional concerns have an important influence on majority decisions. Committees systematically choose fair points during the experiment. Hence, my main conclusion from the experiment is that democratic theory in general and social choice theory in particular could profit considerably from systematically incorporating social preferences in their analyses.