Abstract |
Elected officials discriminate against African Americans. Is this behavior a departure from or an alignment with what their constituents want and would, in fact, do themselves when given the chance? In this article, we present the results from a novel large-scale field experimental technique designed to benchmark levels of racial bias among elected officials to the constituents they serve. We conducted the first audit study on the public---sending correspondence to 250,000 randomly-drawn citizens---and paired that with the largest audit study of public officials to date. Our within-subjects experimental design tested the public's and their elected officials' responsiveness to simple informational requests from either an ostensibly black or an ostensibly white sender. We show clearly that the public systematically discriminates against African Americans. This discrimination is rampant and suggests that the public is just as racially biased as their elected representatives. This suggests that improved democratic representation that more closely aligns elected officials with the preferences of their constituents may actually incentivize---rather than discourage, as many have assumed---elected officials to discriminate on the basis of race. Our results provide an instance that shows that \emph{even when} elected officials are aligned with what their constituents want, and would do themselves, democracy may realize outcomes that perpetuate long-lasting social inequalities. Our results also provide a window into the discrimination that African Americans in this country face in day-to-day interactions with their fellow citizens. |