|
Written by the Consumers Union Southwest Regional Office.
on behalf of the Texas Community Reinvestment Coalition
"I was tired of moving and the kids were torn apart because of the moving. Each time we moved their grades went down. They never grew up with anybody. They don't have a friend they can say 'I grew up with them;. Now, we've been here for two years. Each one has their own room, and it feels better knowing that we don't have to move anymore. Now I know how the pilgrims felt when they settled." |
 |
The American dream to own a home creates an opportunity for lenders to serve their communities and earn a good profit at the same time. But in Texas, reality curbs both the dream and the opportunity.
About one third of all bank failures nationwide during the 1980s occurred in Texas1 and low income neighborhoods were disproportionately affected by the closures. According to a Texas A&M study, low income and minority areas all experienced a decline in the availability of bank branches during this time.2 At the same time, the home mortgage loan market is increasingly dominated by mortgage companies that are not subject to the nation's major law requiring an equitable distribution of home loans to underserved areas, the Community Reinvestment Act (CRA).
Consumers Union's new analysis of home mortgage lending data suggests that many mainstream lending institutions do not serve borrowers in low income and minority areas with the same energy that they devote to upper income and low minority areas.
Consumers Union (CU) analyzed data on every home loan application submitted by borrowers in Texas in 1996 and found that several major lenders have little or no lending activity in low income and minority areas. Further, many institutions deny minority applicants at a higher rate than they deny White applicants, even at the same income levels.
However, the study also finds that some lenders can and do find a market among low income and minority borrowers. Some banks subject to CRA make loans in low income and minority census tracts. But, CU finds that several of the lenders who consistently target low income and minority areas are manufactured housing lenders that may charge substantially higher rates than conventional mortgage loans. While these disparities continue, a viable and profitable loan market in low income and minority neighborhoods remains - largely ignored by mainstream lenders.
Researchers, the press, and community groups have used Home Mortgage Disclosure Act (HMDA) data to produce several notable studies on discrimination in home mortgage lending. The Wall Street Journal studied the impact of race on mortgage lending by examining more than three million mortgage applications submitted by Blacks, Hispanics, Native Americans, Asians, and Whites to lenders in 1992.3 Of the groups, the findings were most alarming among Black applicants, who were twice as likely to have been denied a loan as compared to White applicants-the Black denial rate was 34 percent, while the White denial rate was 17 percent.4
Using a similar methodology, the Texas Community Reinvestment Coalition (TCRC) released a study, "Racial/Ethnic Disparities in Home Mortgage Lending in Texas," in May 1994, based on data from 186,000 home mortgage loan applications. On a statewide average, lenders rejected Blacks and Hispanics 2.2 and 1.9 times, respectively, as often as Whites for home mortgage loans. Among those approved for loans, White applicants disproportionately received approximately 81 percent of the lending, although Whites made up 61 percent of the state population.5 "These preliminary results show quite substantial and disturbing disparities in home mortgage lending by race/ethnicity," said the author.6
The Federal Reserve Bank of Boston, in October, 1992, 7 released the most significant and controversial HMDA study to date. To supplement the applicant and loan characteristics data collected under HMDA, the Federal Reserve Bank of Boston gathered additional financial, employment, and geographic information identified by mortgage lenders operating in the Boston Metropolitan Statistical Area (MSA) relevant to the mortgage lending decision, including census information on neighborhood characteristic, individual applicant credit histories, and loan-to-value ratios.8
The study found that while minorities, on average, had weaker credit histories and higher loan-to-value ratios (the loan amount is high compared to the value of the property), these differences accounted for only a portion of the difference between White and minority denial rates. The study primarily attributed the remaining statistically significant gap to the "considerable discretion" lenders have over mortgage lending decisions. "[F]or the same imperfections Whites seem to enjoy a general presumption of creditworthiness that Black and Hispanic applicants do not, and that lenders seem to be more willing to overlook flaws for White applicants than for minority applicants."9 This result provoked numerous reassessments of the Boston Federal Reserve data, some confirming the conclusions and others disputing them.10
HMDA data provides researchers with substantial information on lending activity by race, income, and geographic area. Because it indicates nothing about the credit history of a borrower or the terms of a home mortgage loan, it is difficult to conclusively prove discrimination solely based on HMDA information. But, HMDA does provide a useful snapshot of a lender's mortgage activity - where loans are made and where they are not. It identifies clearly underserved communities and the lenders who are making a good faith effort to serve these areas.
This study incorporates two basic measures of lender performance: a "disparity ratio" based on the rate at which different types of applicants are denied mortgage loans; and a "market share ratio" comparing each company's mortgage loan market share in underserved communities to its overall mortgage loan market share in an MSA.
We collected 1996 data on more than 400,000 Texas home loan applicants.11 In order to focus on first mortgages, we reduced the statewide lender data for Texas from the total of 846,724 applications to only home purchase loans for a one-to-four family home (homes with one to four units)..12
A denial rate is the ratio of the number of loans denied to the total completed applications (which includes applications resulting in a loan, those approved but not accepted by the applicant, and those denied). We calculated individual company statewide denial rates for Whites, Hispanics, and Blacks.13
We also calculated statewide and major MSA aggregate denial rates for these racial and ethnic groups by income level, to determine whether higher denial rates among minorities could be attributed to lower income levels.
To describe the difference between denial rates for White and minority applicants, we created a "disparity ratio" (used in all charts). The "disparity ratio" is the denial rate for Blacks or Hispanics divided into the denial rate for Whites.
If there were no disparities among these groups of loan applicants in the rate of loan denial, the disparity ratio for each would be 1.0. Where the disparity ratio is substantially higher for minority applicants than for White, it may indicate unfair lending practices.14
In addition to data on the approval and denial rates for different racial and ethnic groups, HMDA data allows researchers to quantify geographically where lenders make loans and where they do not. For Austin, Dallas, Ft. Worth, Houston, and San Antonio, we compared lending patterns in minority and low income census tracts to lending in the city as a whole.15 For each large lender studied we compared its own share of lending in minority and low income tracts to its share of lending across the entire city.
We grouped census tracts in two ways: by their minority concentration (into low minority, below citywide average, and above city average minority areas) and income (low, median, above median, and upper income).16 We examined distribution of loans by census tract minority concentration to determine if lenders treat census tracts with a high minority concentration differently from other tracts. We also analyzed loan distribution by income category because this is one way regulators may judge a lender's performance under CRA regulations.17
For each city, we counted all one-to-four family loans and all applications18 for all lenders in each grouping of tracts (deleting only those loans where demographic information is not available from the census--the total number of loans analyzed is somewhat less than the total loan volume listed on the graphs). The total number of loans in each group of census tracts is the existing "market." This is a conservative approach, since it assumes that there are no additional credit needs unserved by some lender, when in fact, there are certainly unmet capital needs in low income and minority urban areas. CU believes that company performance would appear worse if measured against a reasonable "target," one that included an estimate of unmet demand.
In each of these urban areas we then reviewed the lending patterns of only the largest five mortgage companies, banks, and savings and loans (largest by loans originated) to ensure adequate loan volume for study.19 For each of these large mortgage lenders, we counted their loan originations and their total loan applications in each of the census tract groupings. We calculated a market share by dividing each company's total originations and total applications in the census tract grouping into the total activity by all lenders in that grouping. We calculated an overall market share by dividing each company's total loans and total applications into the totals for the MSA as a whole.
In order to offset differences in size and market share among companies, we then divided each company's market share in low income and minority tracts into its own overall market share. If the company made loans equitably, the low income and minority "market share ratio" would be 1.0 - the company's share of the existing loans in low income and minority tracts is about equal to its share of all loans throughout the MSA. A ratio below 1.0 indicates that the company makes relatively few loans in those census tracts.
Finally, we selected the lenders whose performance in low income and minority census tracts was substantially below their performance in the market as a whole (ratio less than 0.5), and consolidated their loan results with lending by any affiliates or subsidiaries to determine if another affiliated lender for the same parent in fact made the corporation's loans in low income or minority areas.
While lenders continue to deny Black and Hispanic loan applicants at higher rates than they deny White applicants at the same income level, analysis at the individual city level shows that certain lenders are lending in low income and minority census tracts. In some cases, some depository institutions covered by CRA or their affiliates-Frost National Bank in San Antonio or NationsBanc Mortgage, subsidiary of NationsBank, in Dallas, for example-now have a higher market share in these neighborhoods than their market share in the MSA as a whole.
Other banks and savings and loans continue to have a poor lending record in low income and minority areas in their own communities, and major mortgage companies, by far the largest lenders in the mortgage market, also largely ignore these neighborhoods. Typically, only mortgage companies in the manufactured home market-like Green Tree Financial and Bank of America FSB-have a strong presence in low income areas and also deny White and minority applicants at about the same rate.
Some lenders with poor performance in low income areas appear to focus their sales efforts in upper income areas, where they control a much higher share of the market (as measured both by applications taken and loans originated) than their share of the citywide market as a whole. By contrast, the manufactured housing lenders do not have a strong presence in upper income or non-minority areas, a sign of increased fragmentation of the loan market into competitively priced "premium" loans for suburban and upper income buyers and high priced manufactured housing loans for lower income and minority areas.
CU began its analysis by comparing the rate at which Black and Hispanic applicants are denied loans with the rate Whites are denied.20
Statewide, lenders continue to deny loans to Black and Hispanic applicants at higher rates than they deny loans to White applicants. This is also true at the city level. The highest disparities appear in Abilene, Austin, San Angelo, Lubbock, and Victoria.
Interestingly, CU found little disparity on a statewide basis between denials of White, Hispanic or Black applicants who earn less than $50,000 per year. However, lenders reject high earning Black applicants more than twice as often as they reject high earning Whites, and reject high earning Hispanic applicants 70 percent more often. This pattern also recurs at the MSA level, for the five large MSAs selected for more detailed analysis.21
CU examined statewide denial rates for the 15 largest mortgage companies, the 15 largest banks, and the 10 largest savings and loans.
Seven of the 10 mortgage companies with the least disparity between denial rates to Black and White applicants (the best disparity ratios) are major manufactured housing lenders - Green Tree Financial, Access Financial, The CIT Group/Sales Finance, Ford Consumer Finance, Oakwood Acceptance, Vanderbilt Mortgage and Bank of America FSB.22 These lenders also have low Hispanic disparity ratios as well, and many accept a far larger number of minority applications than lenders with the worst disparity ratios.
The lender showing highest rejection rates for Blacks relative to Whites is CTX Mortgage, the subsidiary of home builder Centex (rejecting Blacks more than five times as often as White applicants), while Aegis Mortgage denied Hispanic applicants at ten times the rate they denied Whites. Ten of the 40 lenders we reviewed denied Blacks at three times the rate of Whites or more. Two lenders denied Hispanic applicants at more than three times the rate of Whites.
In response to TCRC's 1994 study focusing on denial rates, Stan Liebowitz of the University of Texas at Dallas offered this justification for these disparities: "the typical minority loan applicant has, on average, a smaller down payment, a shakier credit history, fewer assets and a higher likelihood of reporting information that cannot be verified"23 and therefore, understandably, will be denied at a higher rate. This justification does not address differences at higher income levels, nor to they explain how some lenders find a market where others do not.
Because some lenders consider denial rates a controversial tool for assessing equitable lending, CU performed a more detailed analysis in five major urban areas with significant loan volume (Dallas, Ft. Worth, Houston, San Antonio, Austin).
Lenders have argued that they receive few if any applications from low income or high minority areas.24 However, we found that while most lenders currently make fewer loans in these areas than in upper income and low minority areas, a market certainly exists.
If a particular lender pursued loans equally across parts of a city, we would expect its market share of the existing loan market in low income and high minority areas to be similar to its market share in high income and low minority areas.
While we found no lenders in any city with loans equally spread across all census tract groups, we found several where the differences in lending among tracts (low/high income, low/high minority) were nominal.
Some banks and savings and loans (or their mortgage company affiliates) performed better in minority and low income neighborhoods than in their citywide market share. Frost Bank in San Antonio demonstrates a strong presence in both minority and low income tracts, particularly in the number of loans originated.
NationsBanc Mortgage in Dallas performed well in minority tracts, as measured by both applications collected and loans originated. While NationsBanc Mortgage in Austin accepted fewer than its citywide share of applicants in minority and low income areas, it originated loans at about the same level as Austin as a whole. Also performing well was Summit Mortgage and Bank United, Houston.
The majority of lenders studied by CU consistently show little application or loan activity in these tracts, compared to their marketshare elsewhere - except for the manufactured home lenders, which also showed low "disparity ratios," Green Tree Financial (Ft. Worth, Dallas, San Antonio) and Bank of America FSB (Ft. Worth, San Antonio). But, these lenders may be making higher cost loans.
Manufactured Home Lenders
Green Tree Financial and Bank of America FSB both serve low income census tracts. And Green Tree has very low marketshare in higher income areas (see Appendix, Table 9).
While manufactured home lenders loan in low income and minority neighborhoods, the loans may come at a higher cost. Under the Credit Code, manufactured lenders are permitted to charge 13.32 percent for a manufactured home loan,25 or lenders may use an alternative rate as high as 18 percent.26 In some cases, state laws on interest rates may be preempted by federal law if the loan is backed by a federal program or the guarantor is a federal agency.27
Currently, the national average interest rate for a 30-year fixed rate loan is 7.08 percent.28 We contacted a few manufactured housing lenders and were quoted interest rates ranging from 10 percent to 13.5 percent on a new, single-wide manufactured home.29 All of the rates we were quoted were substantially higher than the national average for conventional loans. According to the most recent Consumer Reports, interest rates on mobile home loans typically run 2 -3 percentage points higher than those for conventional mortgages.30 The 13.5 percent rate charged by some in the manufactured home industry is nearly double the national average rate for a fixed-rate home loan.31
Banks
Banks and their subsidiaries generally have a poor track record in low income lending, although the Community Reinvestment Act requires them to make the loans.
Norwest, a bank with branches in Houston and San Antonio and a major lender (through its mortgage company) in all five cities we studied, has particularly poor low income and minority lending performance in Dallas and Houston. It did not have a strong presence in low income or minority census tracts, or a presence on par with its overall market share in any of the areas examined. This is particularly disturbing because over the past three years Norwest has increased its market share in central Texas through the rapid acquisition of local banks, increasing its reinvestment responsibility to Texas borrowers.32 When we combine lending by Norwest Mortgage with lending by its affiliates the picture remains about the same (see Graphs 4-10: Disparate Lenders).
Bank One Texas NA is a major Dallas-Fort Worth lender and a poor performer in low income and minority neighborhoods in both Dallas and Ft. Worth (see Graphs 4 and 7, Performance). However, when we combine the lending of Bank One Texas NA with its affiliates Banc One Mortgage and Banc One Financial Services, the company as a whole exhibits a more equitable lending pattern.
Some lenders appear to have developed their low income and minority lending services in some cities to a far greater extent than in others. Sterling Capital Mortgage, affiliate of Sterling Bank, is a major lender in Austin and Houston. In Austin, the company has a much stronger presence in minority areas than its presence in the city as a whole (particularly when measured in loans originated). However, in Houston where Sterling Bank has more than a dozen branches, Sterling Capital Mortgage is a poor performer in both low income and minority areas. When we combine the mortgage company loans with loans from Sterling Bank the results remain essentially unchanged (see Graph 10: Houston Disparate Lenders).
Savings and Loan Associations
We examined the five largest savings and loans in each urban area because home mortgage lending has historically been the backbone of this industry. Despite the massive closures after the Texas S&L crisis of the 1980s, several savings and loans remain major players in the home loan market. Like the banks, some appear to have found markets in low income and minority communities, while others have not.
Bank United, a Houston based thrift with branches in San Antonio has more than twice the market share in Houston minority areas than it has in the city as a whole. On the other hand, Guaranty Federal Bank FSB and Temple-Inland Mortgage, subsidiaries of the Texas based lumber giant, are major lenders in four of five MSAs. Guaranty Federal Bank FSB is a depository institution in Dallas with branches in most major Texas MSAs, and a poor performer in Houston, Austin, and Dallas. When CU combined the affiliates and examined Temple-Inland's lending patterns as a whole in each of these cities, the picture improved only slightly.
Other poor performing savings and loans include World Savings and Loan, with branches in Dallas and Houston and particularly poor performance in Dallas, Colonial Savings in Dallas and Home Savings of America in San Antonio and Ft. Worth.
Home-Builder Affiliates
Mortgage companies held by new home builders in our study had a poor lending record in low income and minority areas. CTX Mortgage is one of the largest mortgage companies (by loan volume) in four of the five cities studied. In 1973, the Texas based home builder Centex, established CTX Mortgage to provide mortgage loans for its own developments. But, in 1985 CTX expanded its lending to the broader home mortgage market, and today it makes more than three quarters of its loans outside its own developments.33 However, this expansion into the general market was not universal across all census tracts.
CTX's performance is especially poor in three of the four urban markets it dominates. In Austin, for example, it has only 18 percent of its own citywide marketshare (based on total applications) in minority census tracts, and only 22 percent of its own citywide market share in low income tracts. By contrast, its market share in upper income tracts is 113 percent of its own citywide market share.
Other home builders also dominate home loan originations in Texas cities. CH Mortgage, subsidiary of Continental Homes, is Austin's single largest lender. According to Continental's Annual Report, the lender funds 80 percent of Continental's Texas homes.34 Although Continental builds homes for the entry level home buyer (its homes for the entry level single family market range from $58,950 to $132,000)35 its position in the low income home market is marginal. Continental has less than a quarter of its own citywide market share in low income tracts, focusing instead on median income tracts where it has more than 160 percent of its own citywide market share.
In San Antonio, home builder Kaufman and Broad's affiliate lender, Texas Homestead Mortgage, made no loans in minority or low income census tracts in 1996. The Wall Street Journal in 1995 named Kaufman and Broad one of the "Institutions With Most-UnBalanced Lending Patterns," nationwide.36
Mortgage Companies
A number of independent mortgage companies in our study also showed poor performance in low income and minority areas, sometimes in spite of their existing affordable loan programs. Countrywide Home Loans, Houston's largest mortgage lender (with more than 2700 loan originations in 1996), has an affordable home loan program for low and moderate income borrowers called House America. Nationally, originations through this program dropped from $1.3 billion in 1995 to $0.6 billion in 1996 (FY ending February 1997).37 In Houston in 1996, Countrywide had twice the market share in low minority areas that it had in high minority areas and similarly enjoyed more than twice the market share in high income areas that it had in low income areas (see Appendix: Table 8).
After that [denial], I...gave up because if my own bank wasn't going to give me a loan, I figured what other bank would.--San Antonio resident denied by Bank of America. |
|
Lenders with poor performance in low income and minority tracts also tend to have a much stronger presence in high income and low minority tracts than their market share as a whole. As illustrated by the "disparate lender" charts on the following pages, many of these companies appear to make an extraordinary effort to acquire new loans in high income areas. Their market share as measured by applications accepted is even higher than their market share in loans made in these high income tracts. At the same time, in low income tracts the market share as measured by applications falls below the market share in loans. We believe this indicates lenders that are not making the same marketing investment in low income areas that they make in high income areas.
Overall, the patterns in lending described by this report demonstrate a fragmentation of the home loan market into specialty "niches:" the manufactured housing loan market; the new home loan market with its builder affiliated lenders; and the upper income home market. The new home lenders do not offer loans in many low income and minority areas, largely because they specialize in suburban development. Many banks and major mortgage companies show a strong tendency to compete for this same suburban market. If such specialization continues, many low income and minority borrowers may find that a loan from a manufactured home lender is the only loan available. These lending patterns push low income and minority borrowers into manufactured homes financed with substantially more expensive loans.
By grouping together all census tracts with certain demographic characteristics, this study may fail to distinguish significant differences among underserved areas. For example, lenders in the manufactured home market may concentrate their lending only in a small subset of low income areas. Alternatively, a bank with a poor record lending in minority census tracts may, in fact, serve one minority area well but not others.
In order to identify those neighborhoods where loans are made and where they are not made, further analysis, including mapping underserved areas, would help identify where lenders are not serving communities well. CU plans to include maps as we conduct additional analysis.
In addition, this study does not distinguish among borrowers in low income and minority census tracts. We performed no analysis to determine if lenders with a strong presence in low income and minority areas were making loans to the low income and minority residents of those communities. Though we expect to see a correlation between a lender's presence in these areas and the number of loans made to low income and minority residents, we intend to do further study to confirm this.
Finally, as we have seen in other lending markets - including the home equity and home improvement markets - often low income and minority communities have fewer options for borrowing, and those options are generally more expensive. In this study, low income and minority areas had loans from manufactured home lenders and certain mortgage companies, but far fewer from home builder-lenders, some banks, and some mortgage companies. This growing fragmentation of the home lending market also warrants further study.
In the 1970s, community groups began to challenge redlining - the practice of refusing to lend or provide services in selected minority and low income urban areas. Advocates believed this practice exacerbated the growing disinvestment of urban communities. Groups waged their battle against the lenders in the halls of Congress, and Congress responded. 38
The creation of CRA represented confirmation by Congress that banks and other depository institutions have an affirmative obligation to meet the credit needs in the local communities in which they operate. According to the National Community Reinvestment Coalition, since its inception in 1977, "banks have pledged $200 billion... to fund loan programs for low income and minority consumers." 39
The results of this study indicate that the affirmative lending requirements of CRA have helped to move some banks into lower income and minority communities. But overall these areas remain underserved by most of the major lenders in the market.
First, mortgage companies do not fall under CRA,40 although they are now the major source of home mortgage lending. According to the CU analysis, independent mortgage companies made 49 percent of Texas home loans in 1996, regulated banks made 36 percent, savings and loans 13 percent, and credit unions only one percent of home loans.
Nationally, the banking industry's share of the mortgage market has plummeted. As reported in the American Banker, banks originated only 43 percent of mortgages in 1996 compared to 81 percent in 1977. Nonbanks, by contrast, have seen their share rise to 57 percent from 19 percent in 1977.41
For low income and minority borrowers, this shift from lenders subject to CRA to lending by non-CRA mortgage companies significantly alters the financial landscape. Mortgage companies compete directly with banks and savings and loans for loan customers, but do not shoulder the same kind of responsibility to the communities they serve.
Under CRA regulations, lenders undergo periodic examinations by federal regulatory agencies to monitor the level of lending, investments, and service in their "assessment area." If an agency determines that a lender failed to meet the credit needs of the low and moderate income communities within this area, the lender receives an unfavorable rating, and can potentially be delayed or denied in future merger, acquisition, or expansion approvals. But several provisions of the new regulations (effective July 1, 1997) give lenders great flexibility in this examination.
A wide variety of investments, loans, and services other than direct home lending can be used to fulfill CRA requirements: construction and permanent financing loans for multifamily residential property serving low and moderate income persons; loans for community facilities; credit counseling; home buyer counseling; school savings programs; and others.
Banks with mortgage company subsidiaries may determine at their option whether the subsidiary's lending should be reviewed as part of the bank's CRA compliance. During the comment period to the 1994 final regulations, community and consumer groups recommended that regulators examine the lending patterns of subsidiaries, but the regulatory agencies rejected this proposal.42 Instead, banks may elect to ask examiners to review loans by bank affiliates. If a bank chooses to have examiners review loans by affiliates, they must provide data for all affiliates.43
Although the current regulation generally puts greater weight on direct loan performance indicators than earlier regulations (performance rather than process),44 the basis for evaluating loan performance is limited. While, the lending test considers the distribution of a bank's loans among borrowers of different income levels, it does not require distribution among different racial and ethnic groups. And, the lending test considers the number and amount of loans in low income areas and the geographic distribution of loans, but does not set clear targets using a market share approach like the one used in this analysis.
1 Kenneth J. Robinson, "The Performance of Eleventh District Financial Institutions in the 1980s: A Broader Perspective," Financial Industry Studies, Federal Reserve Bank of Dallas (May 1990) pp. 20-21. More accurately, 36 percent of closures occurred in the Eleventh Federal Reserve District, which is primarily Texas but also includes portions of New Mexico and Louisiana.
2 Bierman, Leonard, Donald R. Fraser, Javier Gimeno, and Lucio Fuentelsaz, "Regulatory Change and the Availability of Banking Facilities in Low-Income Areas: A Texas Empirical Study," SMU Law Review (Vol. 49, No. 5, July-August 1996), pp. 1438-1439.
3 Albert R. Karr, "Loan-Denial Rate is Still High for Blacks," Wall Street Journal, (Dec. 21, 1993), p. A2. (hereinafter Wall Street Journal). The study analyzed conventional and federally-backed home purchase loans only-not refinancings.
5 Wilson, Robert H., Racial/Ethnic Disparities in Home Mortgage Lending in Texas, (May 1994), pp. 5-6.
7 Munnell, Alicia H. et al. Mortgage Lending in Boston: Interpreting HMDA Data, Federal Reserve Bank of Boston, (Oct. 1992) (hereinafter Mortgage Lending in Boston); see also Closing the Gap: A Guide to Equal Opportunity Lending (April 1993).
8 Federal Reserve Bank of Boston, Mortgage Lending in Boston, pp. 1, 13-18.
10 For criticism of the methodology and analysis in the Boston Fed study, see Stan Liebowitz, "A Study That Deserves No Credit," Wall Street Journal, (Sept. 1, 1993) at A14; Mark Zandi, "Boston Fed's Bias Study Was Deeply Flawed," American Banker (Aug. 19, 1993) at p. 13; see also Mitchell B. Rachlis and Anthony M. J. Yezer, "Serious Flaws in Statistical Tests for Discrimination in Mortgage Markets," Journal of Housing Research, (Vol. 4, Issue 2, 1993), at p. 317 (concluding that there is "no adequate statistical test for discrimination in mortgage markets."). But see Lynn Elaine Browne and Geoffrey M.B. Tootell, "Mortgage Lending in Boston - A Response to the Critics," New England Economic Review (Sept. 19, 1995), at p. 53 (finding the study and its conclusions sound and solid). For a re-examination of the Boston Fed Study, see James H. Carr and Isaac F. Megbolugbe, "The Federal Reserve Bank of Boston Study on Mortgage Lending Revisted," Journal of Housing Research, (Vol. 4, Issue 2, 1993), at p.277-313; Peter Passell, "Of Race, Mortgages and Statistical Studies," New York Times, (May 10, 1996).
11 Federal Financial Institutions Examinations Council, A Guide to HMDA Reporting, Getting it Right!, Jan. 1996, pp. 2-5. Under HMDA, most mortgage lenders regulated by either the Office of the Comptroller of the Currency (OCC), the Federal Reserve (FRB), the Federal Deposit Insurance Corporation (FDIC), the National Credit Union Association (NCUA) or the Department of Housing and Urban Development (HUD) must report their mortgage loan applications by census tract to the Federal Financial Institutions Examination Council (FFIEC). A lender must include a range of demographic information about its applicants and specific information about each loan (its purpose, amount and type). Only the following institutions are exempt from reporting: independent mortgage companies that make fewer than100 home purchase and/or refinancing loans, mortgage companies where home loans represent less than 10% of total loan originations, depository institutions without a home office in any MSA, and non-federally insured or guaranteed depository institutions making uninsured loans.
12 We eliminated all home improvement loans, refinancing loans, multifamily dwelling loans and all records of mortgage loans purchased by a lender from another lender. We also excluded all records that contained validity or quality edit failure codes. To avoid multiple counting and focus on loan originations, we eliminated loans included in HMDA data that were purchased by one lender from another lender.
13 Because far fewer Asians or Native Americans applied for loans than Blacks or Hispanics, and even fewer of these were denied loans, we decided not to analyze these results to minimize the effect of a single random loan decision on the overall results. We limited the statewide study to the fifteen largest mortgage companies and banks, and the ten largest savings and loans to ensure adequate population for analysis. The smallest lender included in this portion of the analysis made more than 1200 Texas home loans. No lender had fewer than 50 completed applications from a single category of applicant.
14 Lenders argue that denial rates do not provide any information about the credit risk of the various applicants and encourage researchers to review the reasons for denial provided as part of each FFIEC report. However, several problems with this data prevent effective analysis. First, only lenders regulated by the OCC and the FDIC code a reason for denial with any consistency. These lenders represent less than 35 percent of loan originations. Second, but perhaps more important, lenders choose from among eight listed reasons, including credit history, employment history or debt-to-income ratio. Rather than explain lending patterns, a reason code may merely mask discriminatory lending.
15 In this study, all references to "city" refer to the entire Metropolitan Statistical Area (MSA) as defined by the census.
16 We grouped together all tracts where the tract income was less than 80% of the MSA median income (lower income), all tracts where tract income ranged from 80-119% of MSA median (median income), all tracts at 120-149% of MSA median (above median) and all tracts at 150% of median or higher (upper income). We also aggregated and measured separately only those tracts where the income is less than 50% of MSA median.
For each MSA we also grouped tracts together where the minority population is 0-49% of the citywide minority population (low minority areas), 50-99% (below average minority areas), 100% and higher (minority areas). The minority population for each city is the 1990 census population figure published in the 1997 Statistical Abstract of the United States, Table 46.
We eliminated from the analysis all applications where census demographic data was not listed by FFIEC (NA) and loans where the ratio of tract income to MSA income was listed as 0000.00.
17 12 C.F.R. pt. 25, Sec. 25.22; see also 12 C.F.R. pts 228, 345, 563(e) and 203.
18 Applications include originations, approvals not accepted by the applicant, denials, withdrawn applications and incomplete applications. Lenders may group as home purchase loans all loans for properties housing no more than four families (duplexes, condos etc.).
19 We defined all lenders regulated by HUD as independent "mortgage companies." We defined "banks" as all lenders regulated by either the OCC (regulating nationally chartered banks), the FRB (regulating state chartered, Federal Reserve member banks and bank holding companies) or the FDIC (regulating state chartered, non-member banks). We defined "savings and loans" as all lenders regulated by the Office of Thrift Supervision.
20 This analysis is similar to TCRC's 1994 study, except that this analysis uses the Metropolitan Statistical Area (MSA) as its basic unit rather than the county.
21 Mark Mensheha, "S.A. Home-Loan Figures Show Racial Disparity: Findings Are Reflected Across Industry Sectors," San Antonio Business Journal, (Vol. 10, No. 4, Feb. 9, 1996), p. 1; Earl Golz and Jeff South, "Minorities Get More Home Loans But More Rejections, Too; Loan Applications, Rejections Up For All Texans," Austin American-Statesman, (Sept. 15, 1996), p. A1. These findings affirm similar findings by the San Antonio Business Journal and the Austin American-Statesman studying the San Antonio and Austin lending markets in 1996 (using 1994 HMDA data).
22 Companies describe their manufactured home lending business in their filings to the Securities and Exchange Commission. Access Financial Lending Corp, Form 424B2 (Prospectus), filed 10/30/97. Bank of America, Form 10K405, filed 3/14/97 for FY ending 12/31/96. The CIT Group, Form 10K, filed 3/6/97 for FY ending 12/31/96. Clayton Homes, Form 10K, filed 9/24/97 for FY ending 6/30/97. Ford Motor Co, Form 10K405, filed 3/18/97 for FY ending 12/31/96. Green Tree Financial, Form 10K405, filed 3/14/97 for FY ending 12/31/97. Oakwood Homes Corp. Form 10K, filed 12/29/97 for FY ending 9/30/97.
23 Stan Lebowitz, "Study Author Misses Basket: Limited Research Reaches Flawed Conclusions on Lender Bias," Texas Banking, (July 1994), p. 5.
24 Dee Gill, "Loan Rejections Raise A Tangle of Racial Issues," Houston Chronicle, (Nov. 10, 1991), p. 1.; Rickie Windle, "Frost Bank Under Fire: Lending Questioned By Minority Group," Austin Business Journal, (Vol. 13, No. 45, Sec. 1, Jan. 3, 1994), p. 1.
25 Tex. Rev. Civ. Stat. Ann. art. 5069-6A.03 (Vernon 1993).
26 Tex. Rev. Civ. Stat. Ann. art. 1.04 (Vernon 1991).
27 Interview with Rod Nelson, Legal Aid of Central Texas (Feb. 3, 1998); see also Green Tree Financial, Form 10K, filed 3/14/97 for FY ending 12/31/96.
28 Bank Rate Monitor National Averages (visited Feb. 3, 1998) http://www.bankrate.com/newbrm/avg_natl.htm?product=mtg.>
29 Telephone interview with customer service representatives with the following companies on the current rate of a manufactured home loan as of this writing: Oakwood Acceptance - the highest rate lender will charge is 13.5 percent; Bank of America FSB - the highest rate quoted at the minimum 5 percent down payment is 11.25 percent. Bank of America's rate quoted for used manufactured homes is 13.25 percent.
30 "Dream Home.or Nightmare?" Consumer Reports, (Feb. 1998), p. 33.
31The difference between 7.08 percent - the national average rate for a 30-year fixed mortgage - and 10 percent, the lowest manufactured housing rate quoted to us, is 41.24 percent. The difference was 90.68 percent when we calculated the highest manufactured housing rate quoted,13.5 percent.
32 Norwest Corp., Form 10K, filed on 3/4/97 for FY ending 12/31/96. In 1996 alone, Norwest acquired The Bank of Robstown, Robstown Texas, Henrietta Bancshares, Henrietta Texas, Victoria Bankshares, Victoria Texas, Union Texas Bancorporation, Laredo Texas, Texas Bancorporation, Odessa Texas, West Columbia National Bank, West Columbia Texas, as well as a mortgage company, The Prudential Home Mortgage Company, doing business in Texas.
33 Centex Corporation, Form 10K405, filed on 6/27/97 for FY ending 3/31/97.
34 Continental Homes, Form 10K405, filed on 8/15/97 for FY ending 5/31/97.
36 Albert R. Karr, "Race and Mortgage Lending in America," Wall Street Journal, (Feb. 7, 1995), p. A1.
37 Countrywide Home Loan, Form 10K, filed on 5/21/97 for FY ending 5/31/97.
38 Bradford, Calvin and Gale Cincotta, "The Legacy, the Promise, and the Unfinished Agenda", in From Redlining to Reinvestment: Community Responses to Urban Disinvestment, Gregory D. Squires, ed., Philadelphia: Temple University Press, (1992).
39 Seiberg, Jaret, "20 Years Later, CRA Controversy Going Strong," American Banker, (Oct. 21, 1997), p. 4.
40 Starobind, Paul. "Make 'Em Pay," National Journal, (July 24, 1993), p. 1856.
41 Seiberg, "20 Years Later," p. 4.
42 12 C.F.R. pts 25, 228, 345, 563e, 203. Joint final rule, Response to Comments, The Final Rule, The Lending Test.
43 12 CFR pt 25 Section 25.22(c). Banks may not use affiliate loan data to comply with the performance criteria related to the proportion of the bank's lending in the assessment area.
44 Grady, Francis, The New CRA: A Practical Guide to Compliance, Irwin Professional Publishing, 1997, p. 14
[ Health ] [ Finance ] [ Food ] [ Product ] [ Other ]
[ About CU ] [ News ] [ Tips ]
[ Home ]
Please contact us at: http://www.consunion.org/contact.htm
All information ©1998 Consumers Union