Unequal Burden: Subprime Lending in Atlanta
Income and Racial Disparities in Subprime Lending

View entire report with graphs (pdf file)

This study presents a preliminary analysis of mortgage originations in the Atlanta metropolitan area in 1998 using data reported under the Home Mortgage Disclosure Act (HMDA). Nationwide, the HMDA data demonstrate the rapid growth of subprime refinance lending during the 1990s and further, the disproportionate concentration of such lending in the nation's low-income and minority neighborhoods. These same conclusions hold in the Atlanta metropolitan area.

The growth in subprime lending over the last several years has been a beneficial development for borrowers with impaired or limited credit histories. Subprime lenders have allowed such borrowers to access credit that they could not otherwise obtain in the prime credit market. However, there is a growing body of anecdotal evidence that a subset of these subprime lenders, who generally operate outside the federal regulatory structure, engage in abusive lending practices that strip borrowers' home equity and place them at increased risk of foreclosure. For this reason, this report examines patterns in subprime lending to understand where the risk and impact of predatory practices may be highest.

AN OVERVIEW

In May 1988, The Atlanta Journal-Constitution published "The Color of Money", a series of Pulitzer Prize winning articles about the geographic disparity in mortgage lending between black and white neighborhoods in Atlanta. The Atlanta Journal-Constitution series revealed that in 1986, Atlanta lenders originated six times more home purchase loans per owner-occupied housing unit in predominantly white as in predominantly black neighborhoods. A recent study concluded that although there had been improvements during the 1990s, disparities in lending between black and white neighborhoods continue to persist in Atlanta.

These studies of mortgage lending in Atlanta focused mainly on geographic disparities in home purchase lending. HUD's analysis of home refinance lending in Atlanta reveals an equally troubling trend. Black neighborhoods disproportionately rely upon subprime lenders to refinance their mortgages.

By providing loans to borrowers who do not meet the credit standards for borrowers in the prime market, subprime lending can and does serve a critical role in urban areas such as Atlanta. Some borrowers may have blemishes in their credit record, insufficient credit history, or non-traditional credit sources. The subprime loan market offers these borrowers opportunities to obtain loans that they would be unable to realize in the prime loan market.

But there are two sides to this story. Since subprime lending often operates outside of the federal regulatory structure, it may be a fertile ground for predatory lending activities. Anecdotal evidence suggests that these practices may include imposing and financing excessive fees, bundling high-cost loans with lump-sum credit life insurance, and requiring prepayment penalties. Predatory lending can have disastrous consequences for less financially savvy borrowers. Equity may be stripped from their homes, and in more egregious cases, they may lose their homes altogether.

Some prime lenders have made significant progress in reaching underserved communities. A recent report for the Treasury Department showed that banks and thrifts increased the share of their mortgage originations to low-income borrowers and borrowers in low-income communities from 25 percent in 1993 to 28 percent in 1998. However, as the evidence in this report suggests, there are many Atlanta neighborhoods that could benefit from increased competition from prime lenders in the home refinancing market. Such increased competition would give borrowers in these communities alternative options to lenders that may engage in abusive lending practices.

The first step to ensuring that subprime lending enhances the economic health of the borrowers' families without exposing them to predatory practices is to learn more about how and where it operates in America. To further understand geographic disparities, HUD has analyzed the problem nationwide and has now taken a look at the data on subprime lending in Atlanta.

THE FINDINGS FOR ATLANTA

In general, the analysis shows that subprime lending is more prevalent in lower-income and minority neighborhoods than in higher-income and white neighborhoods. This likely indicates that because of their lower incomes, lenders may consider these borrowers to be a higher credit risk, and these borrowers may therefore be less likely to qualify for prime loans. However, a lack of competition from prime lenders in these markets to find creditworthy borrowers may increase the chances that borrowers are exposed to the predatory practices of a subset of subprime lenders. There is also evidence suggesting that after controlling for income, predominantly black neighborhoods may be comparatively underserved by prime lenders.

The importance of subprime lending to minorities and low-income Americans, which is documented in what follows, demonstrates how important it is to these communities that subprime lending not include any lenders engaging in predatory practices.

1. From 1993 to 1998, the number of subprime refinance loans originated in Atlanta increased by over 500 percent. The number of refinance mortgages reported under HMDA by lenders specializing in subprime lending in the Atlanta metropolitan area increased from 1,864 loans in 1993 to 11,408 in 1998.

2. Subprime loans are three times more likely in low-income neighborhoods in Atlanta than in upper-income neighborhoods. In low-income neighborhoods, subprime loans accounted for 21 percent of all refinance loans originated during 1998 - compared with only 11 percent in moderate-income neighborhoods and just 6 percent in upper-income neighborhoods. In the poorest communities, where median family income is 50 percent or less of the area median income, subprime refinances accounted for 43 percent of all refinance loans.

3. Subprime loans are almost five times more likely in black neighborhoods in Atlanta than in white neighborhoods. In predominantly black neighborhoods in Atlanta, subprime lending accounted for 33 percent of home refinance loans originated during 1998 - compared with only 7 percent in predominantly white neighborhoods.

4. Homeowners in moderate-income black neighborhoods in Atlanta are almost twice as likely as homeowners in low-income white neighborhoods to have subprime loans. In 1998, only 9 percent of homeowners in moderate-income white neighborhoods have subprime refinance loans while 27 percent of homeowners in moderate-income black neighborhoods have subprime loans, which is almost double the 14 percent of homeowners in low-income white neighborhoods who have subprime loans.

In addition, a recent study by Abt Associates of foreclosures in Atlanta found that foreclosures of mortgages originated by subprime lenders have substantially increased since 1996 -- while the overall volume of foreclosures in Atlanta declined by 7 percent between 1996 and 1999, the volume of foreclosures started by subprime lenders grew by 232 percent. As noted earlier, this increase in foreclosures in the subprime market occurred amidst a trend of rapidly increasing subprime lending. Among lenders that report to HMDA, the subprime share of total foreclosures had grown to16 percent by 1999, making it larger than the subprime share of total originations (9-12 percent between 1996 and 1998). Abt also finds that, like originations, the subprime share of foreclosures is highest in lower-income and predominantly minority neighborhoods. The Abt study is reviewed below.

THE ANALYSIS

Subprime mortgage lending provides credit to borrowers with past credit problems, often at a higher cost or less favorable terms than loans available in the conventional prime market. In most cases, these lenders offer credit to borrowers who would not qualify for a loan in the prime market, thus expanding access to credit and helping more families to own their own homes. The higher costs of these loans may serve to offset the increased risk that these lenders assume in lending to these borrowers.

In some cases, however, subprime lenders engage in abusive lending practices known as "predatory lending", which hits homeowners with excessive mortgage fees, interest rates, penalties and insurance charges that raise the cost of refinancing by thousands of dollars for individual families.

HUD's study of subprime lending focuses mainly on subprime refinance lending, which accounts for nearly 80 percent of total (home purchase and refinance loans combined) subprime mortgage lending nationwide and for 71 percent of subprime lending in Atlanta. HUD's study of subprime loans in the Atlanta metropolitan area found that:

1. From 1993 to 1998, the number of subprime refinance loans increased by over 500 percent.

In 1993, there were 1,864 subprime refinance loans in Atlanta reported under HMDA. By 1998, this number had increased by over 500 percent to 11,408. (See Figure 1.)

The magnitude and speed of the increase in subprime lending creates a critical need for greater scrutiny. The rapid growth of subprime lending may help expand credit access for more borrowers; however, some portion of subprime lending may be occurring with borrowers whose credit would qualify them for conventional loans. Subprime lending may expose borrowers to higher up-front fees and interest rates than they would bear if they had obtained prime loans.

2. Subprime loans are three times more likely in low-income neighborhoods in Atlanta than in upper-income neighborhoods.

HUD's analysis reveals that subprime lending is being provided increasingly to low- and very low-income families and their communities. In Atlanta, 11 percent of all refinance mortgages in 1998 were subprime, but in low-income neighborhoods, the percentage of refinances in the subprime market was nearly double -- 21 percent. (See Figure 2). In the poorest communities, where families make 50 percent or less of the area median income, subprime refinances accounted for 43 percent of all refinance loans. In upper-income neighborhoods, only 6 percent of refinancing families relied on a subprime loan. In 1993, only 4 percent of refinance mortgages in low-income neighborhoods and 2 percent each in moderate- and upper-income neighborhoods were subprime.

Broken down by individuals instead of entire neighborhoods, the impact is similar. Low-income borrowers throughout the Atlanta metropolitan area are almost three times as likely as upper-income borrowers to rely upon subprime refinancing (19 percent of low-income borrowers versus 7 percent of upper-income borrowers.) (See Figure 5).

3. Subprime loans are almost five times more likely in black neighborhoods in Atlanta than in white neighborhoods.

In predominantly black neighborhoods in Atlanta, subprime lending accounted for 33 percent of home refinance loans in 1998 - compared with only 7 percent in predominantly white areas. While subprime refinance mortgages accounted for only one in fourteen refinance loans originated in predominantly white neighborhoods, they accounted for one-third of the refinance loans originated in predominantly black neighborhoods. (See Figure 3.) Comparable 1993 figures were 8 percent in black neighborhoods and 2 percent in white neighborhoods.

Broken down by individual borrowers instead of neighborhoods, blacks in the Atlanta metropolitan area are still carrying the greater proportion of subprime: in 1998, 20 percent of refinance mortgages for black borrowers were subprime, compared with only 6 percent for white borrowers. (See Figure 5).

4. Homeowners in moderate-income black neighborhoods in Atlanta are almost twice as likely as homeowners in low-income white neighborhoods to have subprime loans.

Notably, even after controlling for differences in neighborhood income, homeowners in black communities are three times as likely as homeowners in white communities to refinance in the subprime market. (See Figure 4.) Subprime loans accounted for 41 percent of all refinance loans in low-income black neighborhoods, compared with only 14 percent in low-income white neighborhoods.

Among homeowners living in the moderate-income white neighborhoods, only 9 percent turn to subprime lenders. But 27 percent of homeowners living in moderate-income black neighborhoods have subprime refinancing. In fact, this subprime share (27 percent) for moderate-income black neighborhoods is almost twice the subprime share (14 percent) for low-income white neighborhoods.

The maps of the Atlanta metropolitan area summarize the concentration of subprime refinance loans in Atlanta. In 1998, subprime mortgages accounted for at least 25 percent of all refinance mortgages in 101 (or 21 percent) of the 475 census tracts in the Atlanta metropolitan area. Predominantly black census tracts accounted for 74 of these 101 census tracts with high subprime activity. (See Figure 6.) Subprime refinances accounted for at least 25 percent of all refinance mortgages in an additional 20 tracts where Blacks comprised between 30 and 75 percent of the population. (See Figure 7.)

Analyzing the data by the income and racial characteristics of the borrower (rather than the neighborhood) provides similar patterns to those discussed above. There remains a large disparity between black and white borrowers with similar incomes. For example, subprime loans accounted for 31 percent of refinance loans originated for low-income black borrowers, compared with only 10 percent for low-income white borrowers. Similarly, subprime loans accounted for 19 percent of refinance loans for upper-income black borrowers, compared with only 4 percent for upper-income white borrowers. As these comparisons show, upper-income black borrowers are almost twice as likely as low-income white borrowers to rely on the subprime market (19 percent versus 10 percent market shares).

ABT STUDY

Abt Associates recently conducted an analysis of mortgage foreclosures in 12 of the 16 counties in the Atlanta metropolitan area. Abt examined mortgages that were entering the foreclosure process between 1996 and 1999, and compared foreclosures of loans by subprime and non-subprime lenders; Abt identified subprime lenders using HUD's list of lenders that primarily originate subprime loans. The Abt study showed the following:

• The overall volume of foreclosures in Atlanta declined by 7 percent between 1996 and 1999, but the volume of foreclosures started by subprime lenders grew by 232 percent. (See Figure 8.) Among lenders that report to HMDA, the overall share of foreclosures attributable to subprime lending increased from 5 percent in 1996 to 16 percent in 1999. The subprime share of foreclosures (16 percent in 1999) is now larger than the subprime share of originations (10 percent in 1996, 12 percent in 1997 and 9 percent in 1998).

• Over the 1996-99 period, foreclosures with high interest-rate spreads (more than four percentage points over 30-year Treasury) represented 44 percent of the subprime loans entering foreclosure, compared with only 8 percent of non-subprime loans entering foreclosure.

• The median age of loans entering foreclosure is only two years for subprime loans, compared to 4 years for non-subprime lenders.

• The share of foreclosures accounted for by subprime lenders has been growing rapidly in Atlanta neighborhoods of all income levels. But like subprime lender originations, the foreclosure share of subprime lenders was highest in very-low income neighborhoods. Considering only foreclosures by HMDA reporters, subprime lenders accounted for 36 percent of all foreclosures in very-low income neighborhoods during 1999, compared to subprime origination shares of 31 percent in 1997, 39 percent in 1997, and 34 percent in 1998.

• The share of foreclosures by subprime lenders has grown rapidly in neighborhoods of all racial categories. The subprime share of foreclosures, like originations, is largest in neighborhoods where minorities comprise a majority of the residents. Considering only foreclosures by HMDA reporters, subprime lenders accounted for 36 percent of all foreclosures in high-minority neighborhoods during 1999, compared to subprime origination shares of 28 percent in 1997, 31 percent in 1997, and 26 percent in 1998.

CONCLUSIONS

HUD's analysis of refinance mortgages originated in the Atlanta metropolitan area during 1998 clearly demonstrates the exponential growth in subprime lending and its disproportionate growth for lower-income and, particularly, minority homeowners and communities. While this growth in subprime lending has expanded access to credit for many borrowers with impaired or limited credit histories, these borrowers may also be vulnerable to predatory lending practices.

Despite the progress made by prime lenders in reaching these markets, the growth of subprime lending in both lower-income and minority communities strongly suggests that much more can be done by both primary and secondary market participants to expand access to the prime lending market.