June 5, 2017

Collecting data to investigate discrimination in the NYC Rental Market


In “Can the Government Deter Discrimination?” our Moore-Sloan postdoctoral researcher, Andrew Guess, along with Albert Fang (Yale) and Macartan Humphreys (Columbia), conducted a 20 month experiment from April 2012 to December 2013 in partnership with the NYC municipal government to explore the extent of racial discrimination in the city’s rental market.

The researchers began by scouring daily rental ads on Craigslist, the primary advertising platform for NYC rental listings, and randomly selected the ads where potential tenants were invited to inquire by phone.

Then, the experiment employed Black, Hispanic, and White testers to make individual appointments for viewing advertised units, to interact with the associated landlord (or broker), and to record interactions before, during, and after the viewing. When a Black, Hispanic, and White tester of the experiment all made an appointment to view the same housing unit with the same landlord, that landlord was admitted into the field experiment.

652 were admitted in total, and randomly assigned to one of three treatment conditions. The first was the control condition: no governmental messages were sent to landlords and brokers in this group. The second was the monitoring condition. A treatment script informed the landlord that the call was from the NYC Commission on Human Rights, and expressed a friendly reminder of their responsibility to comply with fair housing laws. Lastly, the final condition began with the same treatment script as the monitoring conditionbut with the addition of potential legal ramifications should they not comply.

The control condition found that Hispanics are 28% less likely than Blacks and Whites to receive a callback or an apartment offer from landlords and brokers, and 49% less likely to receive an offer for an apartment than Whites. There was no difference between monitoring and punitive messaging relative to the control, but the collected data did suggest that punitive messaging has some effect on reducing discrimination towards Hispanics.  

While their data did not produce statistically significant results to all aspects of their study, these preliminary results do challenge the US Department of Housing and Urban Development’s 2012 audit report, which found that there was a lack of racial discrimination in the New York City rental market—a conclusion that, as Guess and his researchers have shown, may need to be revised.

by Nayla Al-Mamlouk