Date Published: February 10, 2017
Publisher: Public Library of Science
Author(s): Jesse Marczyk, Cheng-Yi Xia.
Previous theorizing about punishment has suggested that humans desire to punish inequality per se. However, the research supporting such an interpretation contains important methodological confounds. The main objective of the current experiment was to remove those confounds in order to test whether generating inequality per se is punished. Participants were recruited from an online market to take part in a wealth-alteration game with an ostensible second player. The participants were given an option to deduct from the other player’s payment as punishment for their behavior during the game. The results suggest that human punishment does not appear to be motivated by inequality per se, as inequality that was generated without inflicting costs on others was not reliably punished. Instead, punishment seems to respond primarily to the infliction of costs, with inequality only becoming relevant as a secondary input for punishment decisions. The theoretical significance of this finding is discussed in the context of its possible adaptive value.
Consider the following scenario: a tourist is vacationing in a poor, but scenic part of the world. During a trip through the local town the tourist is pickpocketed by a poor man, who quickly absconds with his wallet. While the tourist could chase down the pickpocket himself or report the theft to the local police, he decides against it, viewing the theft as permissible on the grounds that the thief was likely still worse off than he subsequently was. Upon his returning home, the tourist is assaulted by a local man simply for being better off than his assailant.
The present research used a similar design to that found in Raihani & McAuliffe , but with a few important alterations. First, a free option for punishment was provided, as compared with the pay-to-punish methods frequently employed by past research. Participants were given the choice between either (a) deducting a set number of experimental points from another player’s payment for free or (b) not deducting. This modification was made in order to help disambiguate the desire to punish others (or avoid inequality) from the desire to avoid paying the costs of punishment; it is reasonable to suspect that people who would otherwise want to inflict costs on another would refrain from doing so if the costs of doing so were prohibitively high The free punishment option also allows us to more directly assess the outputs of particular preferences for inequality, rather than people’s abilities to instantiate those outputs in the world. For example, a weak individual may be unable to inflict punishment on a stronger one, even if the weaker one desires to do so. In much the same vein, researchers will often use ratings of physical attractiveness rather than actual dating behavior when assessing psychological attraction to others, as one might be unable to successfully mate with others he finds desirable.
Previous research has claimed that human punishment behavior is motivated directly by inequality itself. However, that interpretation was drawn in the presence of important confounding factors, chief among which is that the possibility of experiencing losses (or interpreting the behavior or another as inflicting costs) was always present. When that confound was removed, punishment did not appear to be driven by an aversion to inequality per se; instead, punishment seemed to be driven primarily by the experience of losses.