Evidence and perceptions of discrimination in restaurants

Abstract

We present a natural field experiment designed to examine price discrimination in retail markets. This is done by examining the portion sizes served in British Carvery Restaurants. Carvery restaurants serve traditional, fixed-price roast dinners, and are characterised by the manner in which customers are served: a single chef serves every customer individually and, under observation, cuts them a portion of meat from a roasted joint. We employed 147 testers to pose as diners. We find systematic variations in served meat quantities that correlate with the testers’ gender, with men receiving significantly more meat than women. The gender disparity in portion sizes is robust to controlling for a range of appearance and physical characteristics, and cannot be fully explained by women taking more vegetables or wasting more food than men in a complementary lab-in-the-field experiment. Responses to an online survey conducted via Prolific using a representative sample from the UK point towards a widely held belief that women eat less meat than men, suggesting that the observed discrepancies could stem from statistical discrimination. Evidence from a complementary framed-field experiment highlights how both women and men are negatively affected by this gender disparity. Furthermore, the Prolific survey reveals that neither gender believes serving less meat to women than men when they pay the same price is socially acceptable.