Discrimination in a deprived neighbourhood: An artefactual field experiment

•We present results from an artefactual field experiment.•We use a previously unstudied subject pool: the poorest people in England.•We collect choices in dictator and other allocation games.•We find discrimination towards receivers with Muslim surnames.•We structurally model these group contingent social preferences.•We perform counterfactual simulations.

We present a field experiment designed to examine the discriminatory motives of an understudied demographic: the poorest people in England. Subjects are first asked to divide £10 between two strangers, and then play a £10 dictator game with another stranger. We subtly vary the ethnicity of the receivers by providing subjects with surnames randomly drawn from the electoral register, including treatments that allow us to parse behaviour into either in-group favouritism or out-group negativity, an important behavioural distinction that is typically overlooked in the discrimination literature. Our results suggest that the observed discriminatory attitudes are the result of out-group negativity rather than in-group favouritism. We advance the literature on discrimination through the estimation of a structural model of group-contingent social preferences, which we exploit to perform counterfactual simulations. Our results provide insights into the behaviour of this unique demographic and provide a rationale for why they may support discriminatory policies in their voting behaviour.