By George Galster (Wayne State University), Heather MacDonald (University of Technology Sydney) and Jacqueline Nelson (University of Technology Sydney)
In Sydney’s highly competitive rental market, we were hearing anecdotal reports of rent seekers being treated differently according to their ethnic background. We designed an experiment to test whether these anecdotal reports reflected systematic differences in treatment. Using a method widely used in the US and elsewhere, in late 2013 we conducted a ‘paired testing’ experiment, which involved sending renters of Anglo, Indian and Muslim Middle Eastern backgrounds to rental properties advertised on a large real estate website. The experiment worked by pairing an Anglo or White rent seeker with an Indian or Muslim Middle Eastern rent seeker and comparing their experiences in the rental inspection process. The pairs were the same gender, approximate age, and the researchers allocated each member of these pairs equivalent occupations, family makeup and income. One such scenario might involve two 26-year-old single females with similar occupations and incomes inquiring about the same one bedroom apartment in the inner city by phone and, where possible, then inspecting the apartment in person. The only meaningful difference between the two renters in each ‘test’ was their ethnic background – one of the renters was always an Anglo Australian and the other renter was of an Indian or Muslim Middle Eastern background. The paired testing experiment allowed us to compare the treatment of Anglo, Indian and Muslim Middle Eastern rent seekers in the Sydney rental market.
Our research shows that Indian and Muslim Middle Eastern renters may be systematically disadvantaged when they look for rental accommodation in Sydney (MacDonald et al. 2016). We found several meaningful differences in how Anglo renters were treated compared to Indian or Muslim Middle Eastern renters; Anglos were more likely to be offered an individual appointment to inspect a property, and the agent was more likely to discuss the Anglo renters’ housing needs and tell them of other housing. Anglo renters were also more likely to be given extra information about applying for properties and to receive a follow up call or email from the agent after they inspected a property.
In on our most recent study (published in UAR) we ask why Anglo renters were favored over Indian and Muslim Middle Eastern renters. Is the personal prejudice of agents a convincing explanation for these differences? Alternatively, are the reasons to be found in agents’ perceptions of the discriminatory attitudes of their other customers (customer prejudice theory), or their preconceptions about the types of dwellings and neighborhoods most likely to appeal to particular sorts of renters (statistical discrimination theory).
Our most general and robust result was that rental agents in Sydney, regardless of their ethnicity, were more likely to favor Anglos in neighborhoods that were more ‘desirable’ – those with lower crime rates and a higher proportion of owner-occupied dwellings.
Surprisingly, non-Anglo real estate agents favored Anglo renters to a greater extent than Anglo agents. It may be that non-Anglo agents are concentrated in a niche segment of the Sydney rental market, typically serving a distinctive, overwhelmingly minority clientele, and Anglo agents largely have Anglo customers. If this were the case, it could mean that non-Anglo and Anglo agents see renters from different ethnic backgrounds differently.
There was little evidence that discrimination was driven by personal prejudice or customer prejudice. If personal prejudice was driving discrimination, we would expect that it would largely be Anglo real estate agents discriminating against non-Anglo home seekers, and they would treat renters differently regardless of the location or type of property. This did not happen. Similarly, the customer prejudice theory was not supported by our study. The customer prejudice theory predicts there would be more discrimination in Anglo neighborhoods, but we find little variation in discrimination associated with the ethnic makeup of a neighborhood.
Our analysis supports the statistical discrimination theory. Our findings are consistent with the claim that both Anglo and non-Anglo agents in general are motivated by perceptions of their clients’ willingness to pay for particular properties. Real estate agents consistently varied their behaviors according to the safety and tenure features of the neighborhood a property was in. That is, our results are consistent with the idea that agents assumed that Anglo renters are more willing to pay for rental properties in areas of lower crime and with a higher proportion of owner-occupied properties. Given this assumption, it would make economic sense for all agents to encourage clients differentially in such contexts, and this is what we found. When looking at an expensive rental property agents may have anticipated that prospective Indian and Muslim Middle Eastern rent seekers would be less likely to quickly sign a lease, make timely rental payments and/or maintain the property, and agents therefore give these rent seekers less encouragement than their Anglo counterpart, even when they have the same educational and work backgrounds.
If discrimination in seen as profitable by agents working in the Sydney rental market this must be considered in Australian fair housing policy. To be effective, there must be a strong deterrent to discrimination, where agents perceive that the costs of being tried, convicted and penalized for discrimination are greater than the, probably small, profits earned by discriminating. As far as we know, all anti-discrimination enforcement mechanisms in Australia rely on victims recognizing they have been discriminated against, making a formal complaint and then having the case adjudicated. Unfortunately, most victims never realize that they may have faced discrimination, because they are unaware how other similar home seekers have been treated and agents’ behaviors usually do not reveal bias explicitly. Indeed, some agents may exercise their personal prejudice through more subtle forms of social interaction or even inaction. This means that there is unlikely to be a broadly credible deterrent established unless NGOs and/or governmental agencies are empowered to conduct regular paired testing, the method piloted in this study. Results from the current study offer practical guidance about where enforcement testing is most likely to uncover discriminatory behavior in the rental market. Focusing tests on rental properties in neighborhoods with lower rates of crime and high proportions of owner-occupied dwellings are most likely to reveal favouritism toward Anglo home seekers, and thus would maximize the effectiveness of resources available for ensuring fair housing.
George Galster earned his Ph.D. in Economics from M.I.T. and now serves as Clarence Hilberry Professor of Urban Affairs at Wayne State University. He has published 145 peer-reviewed articles, 8 books and 35 book chapters on a wide range of urban topics. The Urban Affairs Association placed him on their “Service Honor Roll” in 2014 and awarded him the “Contributions to the Field of Urban Affairs” prize in 2016.
Heather MacDonald earned her PhD in Urban Policy and Planning from Rutgers University. She is Professor of Planning and Head of the School of Built Environment at University of Technology Sydney. She has published widely on housing policy and finance, spatial analysis, and urban governance.
Jacqueline Nelson is a Chancellors Postdoctoral Fellow at the University of Technology Sydney. She is interested in how racism manifests and exploring how we can respond to racism, both as individuals and by challenging cultures and practices that reproduce racism and inequality. Her areas of interest include racism in families, local or place-based responses to racism and denial of racism.