Study shows African Americans discriminated against in access to US local public services

The study finds that email queries coming from senders with distinctively African American names are four per cent less likely to receive an answer than identical emails signed by 'white-sounding' names.

The difference in response was most evident in correspondence to sheriffs' offices, with 'black-sounding' names seven per cent less likely to receive a response than 'white-sounding' names.

Responses to 'black-sounding' senders were also less likely to have a 'cordial' tone, that is, respondents were less likely to address the sender by name or with a salutation (such as "Dear" or "Hello").

Co-author of the study Dr Corrado Giulietti, from the Institute for the Study of Labor (IZA), comments: "Despite the fact that prohibition of racial discrimination by the government is a central tenet of US law, our finding shows that not all citizens are treated equally by local public service providers.

"Local services constitute the majority of interactions between government institutions and citizens and perform central functions, for instance in education. The discriminatory attitude that our study uncovers could be one of the factors behind the disadvantaged position of black people in American society and could be a major obstacle towards addressing racial inequality."

The researchers conducted what is known as a correspondence study, a well-established approach of detecting discrimination that has previously been used in contexts like job applications and the housing market.

Using this strategy, the researchers sent emails soliciting information relevant to access a public service, such as office opening hours or documentation needed for school enrollment, from 19,079 local public offices around the country. Targeted services include school districts, local libraries, sheriff offices, county clerks, county treasurers and job centres in every US state. Four correspondent names (two to represent each ethnicity) were chosen as most distinctively recognizable to each group, based upon information from previous studies.

While emails signed by 'white-sounding' names received a response in 72 per cent of the cases, identical emails signed by 'black-sounding' names received a response 68 per cent of the time -- a four-percentage point difference. The difference was the largest for sheriff offices (seven percentage points), while small and statistically insignificant for county clerks and job centres.

There was also a difference in the tone of the response; 72 per cent of responses to people with 'white-sounding' names addressed the sender by name or with a salutation, as opposed to 66 per cent of responses to people with 'black sounding' names.

While discrimination is often thought of as being stronger in different regions of the country, the gap in the response rate is not concentrated in a specific area of the US.

Co-author Professor Mirco Tonin, from the University of Southampton, explains: "We find similar levels of discrimination in each of the four regions defined by the Census Bureau (North-East, Mid-West, South and West). We do find a stronger racial gap in rural rather than urban counties. Moreover, it appears that discrimination is not solely due to the perceived lower socio-economic background of black senders. We obtain very similar results when we indicate the very same profession (real estate agent) in the signature of black and white senders."

Regarding possible interventions to address the problem, Dr Michael Vlassopoulos, also of Southampton, comments: "When trying to identify the race of the respondent, we find suggestive evidence that black respondents are less likely to ignore emails from black senders than white respondents. This suggests that increasing diversity among the public sector workforce, particularly in the services where we detect higher discriminatory attitudes, could be an effective way of addressing discrimination."

The paper, "Racial Discrimination in Local Public Services: A Field Experiment in the US," was published by the Institute for the Study of Labor (IZA) in Bonn, Germany, as IZA Discussion Paper No. 9290. It can be found online at: http://ftp.iza.org/dp9290.pdf