Who gets to Vote? New Evidence of Discrimination among Local Election Officials

• In Event: Field Experimental Studies of Registration, Turnout, and Vote Choice

Abstract

As the 2016 general election illustrated, few questions are more central to contemporary American democracy than how group identity affects political participation. A large body of scholarship contends that racial identity shapes who votes, which candidates they choose, and how campaigns seek to mobilize them. In a recent paper, White Nathan and Faller (2015), make an important contribution to our understanding of race in American politics by providing evidence of bias in an unlikely, and largely unexplored arena -- among the so-called ``street-level'' bureaucrats responsible for fielding electronic inquiries about voting requirements. These local officials were significantly less responsive to requests from putative voters with Latino surnames than putative voters with ``white’’ surnames. In this paper, we successfully replicate the main finding of White et al. Using an experimental design that largely follows the original, we find that Latinos received responses from local officials at a rate 3.1 percentage points lower than whites. Our estimate of this parameter is similar to the original in magnitude and dispersion. Moreover, we extend existing findings by experimentally testing for the presence of bias toward Arab and African Americans. In line with the theoretical expectations that arise from the literature on discretionary bureaucrats, Arab American names received responses at a rate roughly 11 percentage points lower than whites. Surprisingly, constituents with African American names received responses at a rate indistinguishable from that of white voters. These new results provide compelling evidence that bias among local election officials varies in unexpected ways across racial groups.

Authors

• Micah Gell-Redman, University of Georgia

• Douglas Alexander Hughes, University of California, Berkeley