Panel Paper: Do Tax Administrators Discriminate?: Evidence from a Correspondence Study

Friday, November 8, 2019
Plaza Building: Concourse Level, Plaza Court 6 (Sheraton Denver Downtown)

*Names in bold indicate Presenter

David Schwegman, Syracuse University

A central tenet of U.S. law and a core principle of public administration is the prohibition of systematic discrimination, based on race, gender, disability status, etc., by the government. Despite these decades-old laws prohibiting discrimination based on race or gender, there are few empirical studies that test if public servants discriminate against a citizen based on race, gender, or sexual orientation. Giulietti, Tonin, and Vlassopoulos (2017) find evidence of racial discrimination in the access to public services in the United States using an email correspondence study, and Butler and Broockman (2011) find that elected officials are also less likely to respond to black constituents.

In this study, I extend this work by investigating if tax administrators discriminate on the basis of race, gender, and sexual orientation in the administration of the property tax. I have collected the email addresses for tax assessors and administrators from 48 U.S. states. I will present the results of an email-correspondence study, in which I send emails inquiring about local assessments, local property tax exemptions, or assessment appeals to local tax officials. Emails will contain racialized and gendered names. I use the names from Bertrand and Mullainathan (2004). I send four emails to each assessor—one email containing a gendered black name, and one email containing a gendered white name, one email from a same-sex couple, and one email from an opposite-sex couple. I test for discrimination by comparing within-assessor response rates to the black/white and gay/heterosexual emails, respectively. I also compare the quality of the responses. Specifically, I examine if assessing officials provide more helpful or instructive responses (e.g., explaining the assessment appeal process) to white versus black citizens. Correspondence (or audit) studies have been used to investigate discrimination in a variety of settings, including employment (Bertrand and Mullainathan 2004), housing (Yinger, 1986; Ewens et al., 2014; Schwegman, 2019), and the sharing economy (Edelman, Luca, and Svirsky, 2017).

Based on a pilot study I conducted in January and February 2019, I find that 100 randomly-selected assessors were slightly more likely to respond to emails containing white names compared to emails containing black names. The responses sent by assessors to emails containing white names are also longer and more helpful (e.g., they are more specific about the particular issue raised in the inquiry). Property taxes and tax increases are politically contentious. If tax administrators do discriminate against specific groups, social trust in the property tax system may erode which may reduce the likelihood that citizens vote for socially optimal levels of taxation.