Earlier this year, Ta-Nehisi Coates published a cover story in The Atlantic called “The Case for Reparations.” Much of the article is a chronicle of U.S. housing policy, and for good reason: America has a nasty history of denying homeownership to black citizens.

For decades, restrictive covenants kept black people out of white neighborhoods, while redlining prevented black people from getting mortgages in black neighborhoods. These stories bear repeating. Government housing policies in the 20th century helped white America buy homes and build wealth, forming an escalator to the middle class. Many black Americans were never allowed a ride.

Civil rights legislation was supposed to fix these issues, but racism persisted. In 1992, the Boston Fed released a study showing continued discrimination among mortgage lenders. “[E]ven after controlling for financial, employment, and neighborhood characteristics,” the authors wrote, “black and Hispanic mortgage applicants in the Boston metropolitan area are roughly 60 percent more likely to be turned down than whites.”

The Boston Fed paper launched a flock of Justice Department investigations. Legislation was revised to make minority access to credit a priority. In recent decades so much attention has been paid to fair housing and fair mortgage lending that it gave birth to a myth that lax loans to minorities caused the housing crash. That argument has largely been debunked, but it speaks to a perception that the government has tried hard to wring racism out of a system that ran on it for such a long time.

The research shows, though, that racial differences remain in America’s mortgage market.

A recent working paper from the National Bureau of Economic Research found that between 2004-2008, African Americans were 54 percent more likely to be charged high interest rates on their mortgages. Previous studies have pointed out these differences, but this paper is particularly notable for the range and the richness of its dataset.

The major obstacle to any analysis is that minority homebuyers are poorer, on average. They have lower credit scores and  patchier credit histories. So to make a fair comparison, all of these characteristics need to be taken into account.

The authors of this study — Patrick Bayer of Duke; Fernando Ferreira of the University of Pennsylvania; and Stephen Ross of the University of Connecticut — combined federal mortgage data with public housing records and data from a credit agency. They assembled information on individual credit scores, incomes, age, home values and a slew of other underwriting factors.

After controlling for a much as they had data for, they found that people in essentially the same financial situations got different mortgages depending on the color of their skin.

The authors calculate the likelihood that someone received a high-rate mortgage, which is a federal definition that typically means that the mortgage carries an interest rate a couple of percentage points above the going Treasury rate. For instance, a first mortgage would be considered high-cost in this study if its interest rate was three or more percentage points higher.  In this sample of eight major metro areas, an average of 14.2 percent of Americans had mortgages with high interest rates.


The authors estimate that, all else being equal, black Americans were 7.7 percentage points more likely to have a high-interest mortgage. That’s a 54 percent increased risk.

Latino Americans also faced significant disadvantages. They were 6.2 points (45 percent) more likely to have a high-interest mortgage.

Asian Americans were 1 percentage point (7 percent) more likely to have a high-interest mortgage.

Most of these people, the authors write, were concentrated in high-poverty neighborhoods, and especially in places where the residents were less educated.

Was it racism or not?

The economists hit on something surprising when they tried controlling for one more thing: the lending organization.

Here’s what they didn’t find. There wasn’t much evidence of what we would consider traditional racism, like the kind reported in the 1992 Boston Fed paper. Individual lending companies appeared to treat everyone who came in the door more or less the same. (There was still a statistically significant but small difference.)

“A huge amount of the differences in high-cost loans is not whites and blacks going to the same lender and blacks being given a much higher rate,” said Ross, one of the study’s authors. “Rather it’s the fact that there are big differences in the lenders that black and Hispanics are doing business with.”

The economists say these results suggest that during the height of the housing boom, some lenders seemed to consistently steer customers toward costlier mortgages. These lenders perhaps also concentrated their business in minority neighborhoods.

Housing activists have long warned of predatory lending, a problem that reached a peak during the frenzied days of the housing boom.Bank of America’s Countrywide Financial unit was the target of a Justice Department investigation that accused the lender of upcharging black and Latino customers between 2004 and 2008. In 2011, Bank of America agreed to pay $335 million, the largest settlement of its kind.

But that case relied on proving statistically that Countrywide’s brokers treated white clients differently than black or Latino clients. This paper argues that overt discrimination is only a small part of the story.

“This is not something that is going to be remediated with more strict anti-discrimination laws,” Ross said. “It’s more about how the mortgage market is structured.” In short, black and Latino borrowers don’t have access to the kinds of lenders that are giving their white counterparts a sweet rate.

This is what racism looks like in the 21st century. You might not face discrimination at the bank office, but the banks in your black neighborhood are all offering the same raw deal. That deal might be couched in race-neutral terms. These are riskier borrowers, so interest rates need to be higher. Perhaps overt racism is dead, but racial disparities — and the policies that prolong them — persist.

In January, the Consumer Financial Protection Bureau enacted new rules to make mortgages safer. Lenders are discouraged from steering their customers into higher-cost mortgages, and have to make an effort to show that their customers are able to repay their loans. These are steps toward preventing what happened during the housing boom — what happened disproportionately to minority borrowers — from happening again.