An Unreliable Housing Audit

In May, Portland got smacked in the face with embarrassing news: Nearly two of every three times they sought a place to rent, African-American and Latino renters found evidence of discrimination from the city's landlords.

The news came from an audit paid for by the city of Portland and conducted by the nonprofit Fair Housing Council of Oregon. The audit claimed minority testers ran across evidence of potential discrimination 64 percent of the time they talked to landlords or rental agents.

The audit set off howls of protest and frowning editorials about the city's shameful treatment of minorities looking for a place to live. 

But interviews and newly released records from the state Bureau of Labor and Industries show the audit is marred by errors and sloppy work, and that the Fair Housing Council withheld evidence that would have weakened claims that it had found discrimination. Although city officials believed Fair Housing Council testers did all their surveys in person, for example, at least some of their evidence of supposed discrimination against Latino testers was gathered over the phone.

As a result, the audit exaggerates the degree of housing discrimination it found. That raises questions about the reliability of the Fair Housing Council, a watchdog group that contracts with cities across Oregon to do audits.

City officials have refused to renew their contract with the Fair Housing Council and suspended negotiations for a new one Aug. 9 after discovering the audit was rife with problems.

"We have concluded the audit is not reliable," City Commissioner Nick Fish, who oversees the Housing Bureau, tells WW. "There have been failures of communication and breakdowns of protocol. The test they gave us has been demonstrated to be something we cannot rely on."

Fish says problems with the audit shouldn't take away from the city's plan to end discrimination in housing, which he says is supported by other evidence of housing bias in Portland. But the Fair Housing Council audit has proven a political headache for Fish and threatens to frustrate the city's work on equal housing issues.

Moloy Good, the Fair Housing Council's executive director, acknowledges the audit contains errors, but he stands by its basic conclusions.

"We think our report is still reliable," he says. "We reported when we found different treatment of our testers. A lot of what we are looking at here is very subtle different treatment; it's not as blatant as it used to be."

Housing audit testing is done across the country, but this is the first time Portland has sponsored it. The tests work like this: Two people pose as potential renters and separately go in as "testers" to see how each is treated by the apartment manager or rental agency reps. One is white. The other is African-American or Latino. Each tester is supposed to have a nearly identical profile in terms of job, income and personal situation.

The minority tester goes in first and asks a series of questions. The white tester follows. Then each writes a summary of what happened, and Fair Housing Council officials compare the reports.

The Fair Housing Council did 25 tests each using African-American and Latino testers. The audit claimed that the African-American testers found evidence of discrimination 15 times; the Latino testers, 17.

On Tuesday, the state Bureau of Labor and Industries released files of its investigation into the Fair Housing Council's claims. BOLI dismissed one complaint based on the audit; its review of 13 other examples found inaccuracies and discrepancies in the original report.

In some cases, for example, the audit claimed the white tester was offered financial incentives to rent an apartment and the minority testers were not. BOLI documents show this wasn't true in at least two cases—the minority testers were, in fact, offered price deals. 

Good acknowledges the errors but says the testers still faced different treatment.

In these audits, different testers typically talk to the same rental agent to guarantee an apples-to-apples comparison. But BOLI found cases in which testers spoke to different agents—and the audit still claimed to find potential evidence of discrimination. As the BOLI report said of one case, "A review of the actual tester narratives is inconclusive, primarily because the testers met and spoke with different agents."

In at least one case, the testers didn't portray similar background situations, which can muddy the results—something the audit failed to note.

In June, WW reported on potential problems with the Fair Housing Council's methodology. One expert told WW the 64 percent rate of alleged discrimination was beyond credibility and cast doubt on the rest of the results.

Fish says the city's painful lesson in this case won't stop future testing; he says the city plans to do another round this fall. "The public has to have confidence in these tests and the results," Fish says. "Going forward, we need to fix the problem.” 

WWeek 2015