Is There a Two-tiered Mortgage Market Based on Race?

Red-lining has long been a topic of concern in the housing industry.  I have known people who actually saw the maps in bank offices, with the big red line surrounding minority neighborhoods on it, or the predominantly minority Census Tracts colored red.

Red-lining was a very real and pervasive problem until Congress stepped in with the Home Mortgage Disclosure Act (HMDA) and the Community Reinvestment Act (CRA), and bank regulators began to impose sanctions on mortgage lenders who demonstrated bias in their lending practices.

No doubt, there are still banks and mortgage brokers who have not gotten the message, and there is still bias in out society, in home mortgages as well as other areas.  But I was surprised to see the news articles about a recent study that seemed to show that not only was racial bias still practiced, but that Black and Latino borrowers were being broadly discriminated against in major American cities.  This study, produced by the California Reinvestment Coalition (CRC), Empire Justice Center, Massachusetts Affordable Housing Alliance (MAHA), Neighborhood Economic Development Advocacy Project (NEDAP), Ohio Fair Lending Coalition, Reinvestment Partners, and Woodstock Institute, reviewed HMDA data for 2010 and compared conventional and government-backed prime mortgage lending in seven cities, based on borrowers’ race and ethnicity, and the racial and ethnic composition of neighborhoods.

Then I read the report itself.  

I was very disappointed to see major news organizations reporting bias and racial steering when the facts reported don’t really support those conclusions very well.

What the data reported in Paying More for the American Dream VI: Racial Disparities in FHA/VA Lending do show is that loans insured by the Federal Housing Administration (FHA) or guaranteed by the Department of Veterans Affairs (VA) are more prevalent in neighborhoods where the residents are predominantly Black, Latino or when borrowers were Black or Latino.

First, some basics.  FHA and VA loans are guaranteed by the United State government, so they have the lowest rates available for the same length and size of loan.  Also, VA and FHA loans almost always provide the borrower with the opportunity to buy a home with a smaller down payment than a conventional lender would require.  In fact, VA and FHA loans were created as part of the drive to make home ownership available to families who would not qualify for loans from conventional lenders.

So, receiving a VA or FHA loan is not necessarily an indicator of discrimination.  It may be an indicator that the borrowers were not otherwise qualified to buy a home or did not choose to make a sizable down payment.  (Disclosure:  I married a woman who bought her first home with a VA loan, and she and I bought our next home with an FHA loan.  Neither of us is Black or Latino.)

Second, the study only looked at the correlation of the race of the borrowers or their neighborhoods, and the prevalence of FHA and VA loans.  It did not look at other factors that are relevant.

What factors other than race and ethnicity would affect the ability to get a mortgage and buy a home?  Income, assets, job history, and prior payment history are all important.  Together these contribute to credit scores.

Right or wrong, in today’s world, credit scores control access to borrowed money.  There have been many criticisms of how we came to the point where credit scores, calculated primarily by one company, FICO, previously known as the Fair Isaac Corporation, could determine whether we could get a loan, and what the terms of that loan would be. That discussion is way beyond the scope of this comment, but that is the reality today.  Other credit reporting agencies, like Experian, TransUnion and Equifax, also produce consumer credit scores, and some major lenders have developed their own proprietary scoring models, but for real estate lending, FICO seems to be the standard.

(One factor, which stops me from including anything about VA loans, is that you must be a military veteran in order to qualify for a VA loan.  There is no reference about the different rates of military service among the different populations in this report.  To imply that high use of VA loans is somehow important is unsupported.  See the table below to see the differences between veteran status and groups.)

What this study did not take into account was the different income, asset or credit score information about these borrowers.  It simply ignored the credit worthiness of the borrowers.  And that is fatal to the purpose of the study.

A presentation by Harriet Newburger, of the Federal Reserve Bank of Philadelphia, reviewed FHA Lending in the 2000s.  The data showed that during the period 2000 to 2009 about 4/5ths of FHA borrowers were first-time home buyers, 40-50% of FHA borrowers had less than 80% of the area median income in their communities, while only 15-30% had more than 120% of area median income, and that less than 40% of FHA borrowers were minorities, dropping from a high of about 38% in 2000 to about 25% in 2009.  The data also showed that FHA borrowers had relatively low FICO scores, but that Non-Hispanic Whites made up over half of the FHA borrowers with FICO scores below 580 and below 620.

The data also showed that FHA borrowers were distributed among race and ethnicity in about the same proportion as the population, as shown in the 2010 Census.  The American Housing Survey shows that minority families made up a much smaller proportion of all home buyers than one would expect, based on their share of the population.

Racial/Ethnic Group

Share of Population

Share of Veterans

Share of All Home Buyers

Share of FHA Buyers

Black

12.1

11.6

8

14.0

Hispanic

12.5

6.0

9

17.5

Non-Hispanic White

69.1

78.7

76

64.7

In fact, it may have been the availability of the FHA program that allowed these households to become home owners, not because of discrimination by conventional lenders based on race or ethnicity, but because of the differences in their incomes and credit scores.  This possibility is reinforced by a study released by the Federal Reserve Board in 2007.  In the Report to the Congress on Credit Scoring and Its Effects on the Availability and Affordability of Credit  if was reported that Black and Hispanic credit scores were lower on average and more heavily distributed toward the lower end of the credit score scale (shown in deciles) than Non-Hispanic White and Asian scores (see the chart below).  The same results were reported by the Center for Responsible Lending.

The result of my brief review is a feeling that the California Reinvestment Coalition and the other authors wasted their time in publishing this report, not because their data is wrong, but because they didn’t go far enough in their analysis to identify the real problem.

Their data and analysis show a correlation of minority borrowing and FHA and VA loans, but says nothing useful about whether or not that is a good thing or a bad thing.  And it doesn’t tell us anything about why there is a difference in FHA and VA borrowing rates among different racial and ethnic groups.

I hope that they will follow this up with a multiple regression analysis of all of the factors that might affect borrowing patterns by different racial and ethnic groups.  If, as I suspect, that shows that the major factors are income, credit histories, and wealth, then we can determine whether the FHA program is accomplishing what it was intended to do––allow lower income households with fewer assets to become home owners.  And that will help inform the debate about whether that is good public policy or not.

1 reply

Trackbacks & Pingbacks

  1. […] income, asset or credit score information about these borrowers. … Continue reading here: Is There a Two-tiered Mortgage Market Based on Race? ← Do You Have a Better Chance of Getting a Mortgage From a … Closing Costs When […]

Comments are closed.