Name matters! The cost of having a foreign-sounding name in the Swedish private housing market

Published: 11 April 2022| Version 1 | DOI: 10.17632/hkvv87yzm7.1
Contributor:
Elina Lampi

Description

In this internet-based field experiment, we investigated whether there exists discrimination in the Swedish private rental housing market based on the names of apartment seekers. We used a correspondent test by randomly submitting equivalent applications from four fictitious, highly educated, and seemingly “well-behaved” male applicants in response to a number of randomly selected private housing ads. Each advertising landlord received applications from two applicants with names signaling Swedish, Arab/Muslim, Eastern European, or East Asian ethnicity. Our results show that the person with a name associated with the dominant ethnic group received most callbacks from the landlords, while the persons with Eastern European- and East Asian sounding names, and especially the Arab/Muslim-sounding name, yielded significantly lower callback rates. Moreover, each applicant’s callback rates are about the same regardless of whom he was paired with, reinforcing our result that a person’s name clearly matters when applying for an apartment. The comparisons with previous discrimination research focusing on the Swedish housing market show that the situation for a male person with an Arabic/Muslim-sounding name has at least not improved in Sweden in the past decade. The individual data set is expanded so an applicant is the unit we investigate. In the second data set (Pairwise data set) an advertisement is the unit. This since we sent 2 email applications to each landlord/advertisement.

Files

Steps to reproduce

The data was gathered by replying rental housing ads on the Swedish buy-and-sell website Blocket.se. We have used STATA in our analyses. The economtric model was binary probit model where the standard deviations are clustered at advertisement (landlord) level since we sent 2 applications to each landlord. The on average marginal effects are shown in the article.

Institutions

Goteborgs Universitet

Categories

Behavioral Economics

Licence