MORTGAGE MONEY:WHO GETS IT?

A Case Study in Mortgage LendingDiscrimination in Hartford, Connecticut

U.S. Commission onCivil RightsClearinghouse Publication 48June 1974

U.S. COMMISSION ON CIVIL RIGHTS

The United States Commission on Civil Rightsis a temporary, independent, bipartisan agencyestablished by the Congress in 1957 to:

• Investigate complaints alleging denial ofthe right to vote by reason of race, color,religion, sex, or national origin, or byreason of fraudulent practices;

• Study and collect information concerninglegal developments constituting a denial ofequal protection of the laws under theConstitution because of race, colors-religion,sex, or national origin, or in the adminis-tration of justice;

• Appraise Federal laws and policies withrespect to the denial of equal protection ofthe laws because of race, color, religion,sex, or national origin, or in the adminis-tration of justice;

• Serve as a national clearinghouse forinformation concerning denials of equalprotection of the laws because of race,color, religion, sex, or national origin; and

• Submit reports, findings, and recommenda-tions to the President and the Congress.

MEMBERS OF THE COMMISSION

Arthur S. Flemming, ChairmanStephen Horn, Vice ChairmanFrankie M. FreemanRobert S. RankinManuel Ruiz, Jr.John A. Buggs, Staff Director

For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402

Page

CONTENTS1. Why Study Mortgage Lending? 32. Population and Housing in Hartford _ 5

Population 5Housing 6

3. Minority Rejection in Mortgage Lend-ing 9

The Role of Brokers 9The Role of Loan Officers 12Objective Criteria 13

4. Sex Discrimination in Mortgage Lend-ing 18

The Working Wife 20The Single Woman 26

5. Employment in Lending Institutions _ 316. Summary and Conclusions 33

TABLES

1. Racial and Ethnic Population of Hart-ford SMSA, 1960-1970 6

2. Owner-occupied Housing Units, byNumber and Value, Hartford SMSA,1970 7

3. Lender Policies on Broker Screeningof Applicants 11

4. Lender Policies on Ratio of Income toHousing Payment 14

5. Women's Earnings as Percentage of.Men's Earnings - 20

6. Median Income of Families with BothHusband and Wife Present, by WorkExperience and Race 20

7. Employment of Wives, by Race 218. Lender Policies on Wife's Income 229. Officials, Managers, and Professionals,

By Race and Sex, Four Hartford Mort-gage Lenders . 32

10. Office, Clerical, and Service Workers,By Race and Sex, Four Hartford Mort-gage Lenders 32

Homeownership is a goal most Americanfamilies aspire to, but which some find difficultor even impossible to achieve. In 1970, 65percent of all white families owned their homesbut only 42 percent of all black families and 44percent of all Spanish speaking families. In thatyear 68 percent of all families headed by menowned their homes contrasted to 48 percent ofall families headed by women.1 Discriminationby real estate brokers and mortgage lenders islargely responsible for this disparity.

In 1968, passage of the Federal Fair HousingLaw2 prohibited discrimination against minori-ties in obtaining mortgage financing. Yet realestate brokers and mortgage lenders still treatminority home buyers differently from whitepurchasers. White, male-dominated lendinginstitutions use imprecise, subjective criteria ingranting mortgages; and these criteria, such asmotivation and eligibility, are applied differen-tially.

Although most families have some difficultyin obtaining the financing necessary to purchasea home, minority families and women oftenencounter insurmountable obstacles when theyapproach the mortgage lending community.A white family headed by a male whose incomeis sufficient by itself to carry the cost of home

1 U.S., Department of Commerce, Bureau of theCensus, 1970 Census of Housing: Metropolitan HousingCharacteristics—Final Report, no. HC(2)-1 (Septem-ber 1972), pp. 9, 18. U.S., Department of Commerce,Bureau of the Census, 1970 Census of Housing: Hous-ing Characteristics by Household Composition—FinalReport, no. HC(7)-1 (February 1973), pp. 105, 114.See also, U.S., Department of Commerce, Bureau ofthe Census, 1970 Census of Housing: Structural Char-acteristics of the Housing Inventory—Subject Report,no. HC(7)-4 (June 1973), p. 2.

* 42 U.S.C. § 3601 et seq.

purchase conforms easily to the traditionalcriteria of the lending community and is readilyapproved. But minority families, familiesheaded by women, families in which both thewife's and the husband's incomes are necessary,and single persons do not fit as well into thetraditional perception of homeowners held bylending institutions. Their applications fre-quently are handled arbitrarily.

During a period of tight money, as at present,restrictive conditions worsen. As the pool ofmoney available for mortgages decreases, fewerfamilies are able to obtain financing. The accessof marginal families to money in this situation,including families headed by minorities orwomen, is even more restricted.

The purpose of this investigation has notbeen to uncover individual instances of dis-crimination by mortgage lenders, but to exam-ine the system of mortgage finance and its effecton homeownership opportunities for minoritiesand women. The Commission has examined anumber of the purportedly neutral criteria thatgovern mortgage loan decisions to determinethe extent to which they afford equal opportu-nity to minorities and women and assure thatdecisions will be made on the basis of objectivefactors, not personal or institutional bias.

Measuring the extent to which the mortgagefinance system results in discriminatory treat-ment of minorities and women is a difficult taskbecause data are unavailable, inadequate, ordifficult to obtain. The responsibility for requir-ing data collection of the Nation's lenders restssquarely with the Federal financial regulatoryagencies which have long resisted promulgatingsuch a regulation. Many statements and actionsof real estate brokers and mortgage lendersthat discourage minorities and women from

seeking to finance the purchase of a home neverbecome part of a written record. The fact thatloan inquiries and application procedures areinformal serves to emphasize that ampleopportunity exists for discrimination againstminorities and women. Although the practicesof mortgage lenders are more often covert thanovert, they nevertheless have the effect ofdenying many qualified families the opportunityof homeownership.

This report reflects the Commission's investi-gation of mortgage lending policies and prac-tices in a demonstrably typical American city—Hartford, Connecticut. A sizable minoritypopulation is clustered in the central city.Although there are both city and suburbanhomes priced within the income range ofminorities, the incidence of homeownership is

lower among blacks and the Spanish speakingthan among whites, and the suburbs remaininhabited almost entirely by whites. The centralcity is declining in population, mainly becauseof the exodus of white residents to rapidlygrowing suburbs. In short, the population shiftsoccurring in Hartford and many of the factorscontributing to these changes in choice ofresidence are characteristic of those found else-where. The practices and attitudes of Hartfordmortgage lenders also reflect those found incities throughout the Nation.

The Commission believes that the factsuncovered by this report are sufficiently alarm-ing to alert the community of mortgage lenders—and their regulatory agencies—to the needfor a reexamination of the policies and prac-tices under which they operate.

i

Chapter 2Population and Housing in Hartford

Hartford, the capital of Connecticut, is aconveniently located commercial center for theState. It is served by Bradley InternationalAirport and three major interstate highways.The insurance industry started in Hartford inthe early 19th century, and 39 companies haveheadquarters there. Like many other metro-politan areas in the country, the city of Hart-ford is very different from the surroundingtowns which make up the rest of the HartfordStandard Metropolitan Statistical Area(SMSA).3

POPULATIONDuring the decade of the 1960's, the popula-

tion of the Hartford metropolitan area grewsubstantially, from less than 550,000 to morethan 660,000. All of this growth occurred out-side of the central city, which declined frommore than 162,000 people in 1960 to about158,000 in 1970.4

The racial composition of the central citychanged significantly from 1960 to 1970: Theblack population nearly doubled, from close to25,000 to a little over 44,000, while the PuertoRican population almost quadrupled, from

3 An SMSA is a county or group of contiguouscounties containing at least one city of 50,000 or moreinhabitants, or "twin cities" with a combined popula-tion of at least 500,000. Contiguous counties areincluded in an SMSA if they meet criteria for socio-economic integration with the central city. In the NewEngland States, SMSA's consist of towns and citiesinstead of counties.

4 Unless specificially noted, population, housing, andincome figures are taken from the following U.S.Department of Commerce, Bureau of the Census, pub-lications: 1970 Census of Housing: MetropolitanHousing Characteristics—Final Report, no. HC(2)-1

about 2,300 to more than 8,500.5 In 1970, theblack and Puerto Rican population combinedrepresented about a third of the inner-citypopulation.6 Over the same 10 years, more than31,000 whites left the inner city, representingalmost 25 percent of the white population.

Although the black population in the suburbsof Hartford increased from nearly 4,000 to6,400 in the 1960's, blacks still representedbarely 1 percent of the suburban population in1970. Of the more than 9,000 Puerto Ricansresiding in the Hartford metropolitan area in1970, fewer than 700 lived in the suburbs, an

(September 1972). 1970 Census of Housing: HousingCharacteristics by Household Composition—FinalReport, no. HC(7)-1 (February 1973). 1970 Census ofHousing: Structural Characteristics of the HousingInventory—Subject Reports, no. HC(7)-4 (June 1973).1970 Census of Population and Housing: CensusTracts—Final Report, no. PHC(l)-87, Hartford,Conn., SMSA (May 1972). 1970 Census of Population:General Social and Economic Characteristics—FinalReport, no. PC(1)-C8, Connecticut (April 1972).1970 Census of Housing: Metropolitan Housing Char-acteristics—Final Report, no. HC(2)-89, Hartford,Conn., SMSA (May 1972). 1970 Census of Population:Puerto Ricans in the United States—Subject Reports,no. PC(2)-1E (June 1973). 1970 Census of Housing:Detailed Housing Characteristics—Final Report, no.HC(1)-B1, U.S. Summary (July 1972). 1970 Census ofPopulation: Earnings by Occupation and Education—Final Report, no. PC(2)-8B (January 1973).

5 Data on Puerto Ricans are limited in the 1970census because in some instances this ethnic group isclassified together with Cubans and Mexican Americansas persons of "Spanish language" origin. Wherepossible, data pertaining exclusively to Puerto Ricanshave been used.

6 The 1970 census may have substantially under-counted Spanish speaking persons in Hartford andblacks. Nevertheless, while the undercounting ofminority members distorts the actual numbers, thefigures at least give indications of trends.

increase of just 5 persons over the 1960 figure.(See table 1.)

The growth of the black and Puerto Ricanpopulation in the city of Hartford has notoccurred uniformly. Blacks and Puerto Ricansare concentrated largely in the northern sectionof the city, while the southern and westernsections have remained predominantly white.

HOUSINGHousing in the suburbs of Hartford differs

markedly from housing in the central city. It isnewer, less likely to be overcrowded, and muchmore likely to be owned by the occupant. Threeof every five suburban housing units have beenconstructed since 1950, while 80 percent of thehousing in the central city was built before1950. Fewer than 5 percent of the suburbanhouseholds are overcrowded, compared to morethan 10 percent of the city's households.

Of the more than 100,000 owner-occupiedunits in the Hartford metropolitan area, fewerthan 10 percent are in the central city. Eighty-five percent of all black families living in thecity of Hartford do not own their own homes;97 percent of all Puerto Rican families in thecity are renters. Of the city's female-headedfamilies, 85 percent do not own their ownhomes.7 By contrast, more than 70 percent ofthe housing in the suburbs is owned by theoccupants, including the few blacks and PuertoRicans who live there.

Earnings are one reason why so few blackand Puerto Rican families live in the Hartfordsuburbs or can be counted among the area'shomeowners. Median income for all families inthe city as of 1970 was $9,100 a year. But forblacks it was $7,000 and for Spanish speaking

? Ownership figures are for two or more personfemale-headed households only.

TABLE 1—RACIAL AND ETHNIC POPULATION OF HARTFORD SMSA, 1960-1970

HARTFORD1970

Number Percent1960

Number Percent

1960-1970Change

Number Percent

SMSAWhiteBlackPuerto Rican**CITYWhiteBlackPuerto RicanSUBURBS*** —WhiteBlackPuerto Rican

663,891600,80550,5189,236

158,017103,31944,0918,543

505,874496,486

6,427693

100.0*90.5

7.61.4

100.065.427.9

5.4100.098.11.30.1

549,249516,78428,8132,995

162,178134,72024,8552,307

387,071382,064

3,985688

100.094.1

5.20.5

100.083.115.31.4

100.098.7

1.00.2

114,64284,02121,7056,241

-4,161-31,401

19,2366,236

118,803114,422

2,4695

20.916.375.3

208.4-2.3

—23.377.4

270.330.732.562.40.7

Source: U.S., Department of Commerce, Bureau of the Census, 1970 Census of Population and Housing:Census Tracts—Final Report, no. PHC(l)-87, Hartford, Conn. SMSA (May 1972). pp. 1, 14. See alsoCity of Hartford, Conn., Commission on the City Plan, Analysis of 1970 Census Data (December1972), pp. 2-3, table I.

* Data for American Indians and Asian minorities are included in each total but are not shown separ-ately. Therefore, combined figures do not equal 100 percent in any category.

** Persons of Puerto Rican birth or parentage made up 59 percent of the Spanish speeking population inthe Hartford SMSA in 1970 and 72 percent of the Spanish speaking population in the city of Hartford.

*** Figures for the balance of the Hartford SMSA excluding the city of Hartford.

families, $5,250. Thirty percent of the minorityfamilies in the city of Hartford had incomesbelow poverty level. By contrast, in the suburbsmedian income for all families was $13,300 ayear. For blacks it was $12,600 and for Spanishspeaking families, $12,400.

Economics, however, by no means fully ex-plains the gross underrepresentation of blacksand Puerto Ricans in the suburbs of Hartfordor among the area's homeowners. Despite thedisparity between the annual income of minor-ity and majority group families, a substantialproportion of Hartford minority families earnenough to enable them to buy housing at marketprices. Some 20 percent of the Spanish speakingfamilies in the city and over 30 percent of the

black families in the city earn more than$10,000 a year and generally can afford housingcosting as much as $25,000. Nearly half of theowner-occupied housing in Hartford's suburbsand fully 75 percent of the owner-occupied unitsin the city are valued at less than $25,000.(See table 2.) Nonetheless, few black orSpanish speaking families are homeowners.

The more than 18,000 female^headed families,which are almost evenly distributed betweenthe central city and the suburbs, constitute 11percent of all families in the area. Nineteenpercent of these female-headed families areblack and 4 percent are Spanish speaking.

In both the city and the suburbs, female-headed families earned substantially less than

TABLE 2—OWNER-OCCUPIED HOUSING UNITS BY NUMBER AND VALUE,HARTFORD SMSA, 1970

HARTFORDSMSA

HARTFORDCITY

HARTFORDSUBURBS

Value of Owner-Occupied Units

less than $10,00010,000-14,99915,000-19,99920,000-24,99925,000-34,99935,000 or more

TOTAL*

less than $10,00010,000-14,99915,000-19,99920,000-24,99925,000-34,99935,000 or more

TOTAL*

less than $10,00010,000-14,99915,000-19,99920,000-24,99925,000-34,99935,000 or more

TOTAL*

Number ofUnits in

Area

1,0064,645

18,74228,38032,89220,580

106,245

75462

2,1091,9731,092

3666,077

9314,183

16,63326,40731,80020,214

100,168

Percentage ofTotal Units in

Area

0.94.4

17.626.731.019.4

100.0 ?o

1.27.6

34.732.518.06.0

100.0 %

0.94.1

16.626.431.720.2

Source: U.S., Department of Commerce, 1970 Census of Population and Housing: Census Tracts—FinalReport, no. PHC(l)-87, Hartford, Conn., SMSA (May 1972), table H-l.

* Total may not add to 100 percent due to rounding.

all families. Within the city of Hartford thesefamilies had a mean income of $6,206 comparedwith $10,011 for all families. Their meanincome in the suburbs was $9,562 compared to$14,595 for all families. In addition, the dis-parity between male and female median earn-ings in the metropolitan area was equally severefor all races and ethnicities.8

Approximately 5 percent of all male- andfemale-headed families were living in povertyin 1970 in Hartford. Yet, of the female-headedfamilies alone, 14 percent of the white families,71 percent of the black families, and nearly57 percent of the Spanish speaking familieswere living in poverty. These women are notpotential homeowners. Still, a substantialnumber—about 15 percent—of those womencurrently renting their homes earn incomesover $10,000 and can afford to purchase inner-city or suburban housing.

Undoubtedly a variety of factors contributesto the gross underrepresentation of minorityand female-headed families among Hartfordhomeowners. What role does mortgage financeplay in denying these people homeownershipopportunities?

8 Data are not available on median family income byrace and sex in Hartford. Median earnings of personsby race and sex in the Hartford metropolitan area are:all males, $8,918; all females, $4,225; black males,$6,237; black females $3,548; Spanish speaking males,$6,131; Spanish speaking females, $3,861. U.S., Depart-ment of Commerce, Bureau of the Census, 1970 Censusof Population: General Social and Economic Character-istics, no. PC(1) C-8, Connecticut (April 1972), tables89, 93, and 99.

Chapter 3Minority Rejection in Mortgage Lending

In Hartford, as in other communities, theprocess by which mortgage loans are made iscomplex.9 At each stage opportunities exist fordenying mortgage loans to qualified minorityfamilies on the basis of nothing more than thepersonal prejudice of individuals in positions todecide. Few complaints are made by minorityfamilies who have been rejected for mortgageloans because there is little objective basison which they can complain. Usually therejection is informal, and the minority familiesrarely know the precise reasons why they arerejected. Indeed, lending institution officialsthemselves frequently cannot explain thesereasons by reference to objective credit factors.

Typically, most American families must suc-cessfully run a three-stage screening gauntletbefore obtaining their first mortgage loan.They are screened initially by a real estatebroker, next by a loan officer, and finally by theinstitution's loan committee. Minority families,however, are much more closely scrutinizedthan majority families, and each stage isfraught with the possibility of discriminatoryrejection. The criteria considered in determin-ing approval or rejection include several factorswhich are so subjective as to permit decisionson the basis of personal prejudice. Even thosecriteria which appear to be objective andsusceptible to precise measure turn out, on closeexamination, to be inconsistent and open todifferential application.

9 This report is based on data obtained through inter-views of real estate brokers, lenders, home buyers,public interest groups, and Federal and city housingspecialists. Interviews were conducted from July 1972through September 1973. The original written materialsare on file and available for review at the U.S. Com-mission on Civil Rights in Washington, B.C.

THE ROLE OF THE BROKERSThe real estate broker serves several func-

tions in a real estate transaction. Generally,brokers represent sellers and their principalobligation is to arrange for a sale on the mostadvantageous terms to their clients. Brokersalso have a direct relationship with buyers inthat buyers usually deal with them whenlooking for a house.

Studies have documented discriminatoryrejection by brokers of minority families whoare ready, willing, and able to meet the termsof sale offered by sellers.10 White brokers haveattempted to dissuade minority homeseekersfrom making an offer on a house they desire;they have gone to great lengths to avoid havingto deal at all with minority homeseekers; andthey have flatly refused to tender good-faithoffers by minority families. These actions—subtle and covert—usually have been con-sciously discriminatory.

Often, brokers also have a direct relationshipwith lenders. Through continuing contactswith one or more lenders, they can help arrangefor expeditious financing on favorable terms.It behooves brokers to remain in the goodgraces of lending institutions to assure afriendly reception for those whom they referfor financing.

Although brokers may handle sales every-where in the Hartford metropolitan area,11 the

10 See Rose Helper, Racial Policies and Practices ofReal Estate Brokers (Minneapolis: University ofMinnesota Press, 1969). Bibliography included.

11 Commission staff interviewed members of 16 realestate firms from July 1972 through September 1973.Seven of the firms were black owned; one of the personsinterviewed was a black woman. This sample is approxi-mately 10 percent of the total number of real estate

minority brokers are located and work pri-marily in the minority area of the city—BlueHills and the suburb of Bloomfield. There aremany more white firms and they are generallylarger, some having branch offices dispersedthroughout the city.

The broker's treatment of the home buyerdiffers somewhat according to the home buyer'srace. Both black and white brokers are con-cerned that the homeseeker is financially ableto buy a home. But several acknowledged thatthey expend more energy ascertaining thepurchasing ability of the potential black homebuyer, making the assumption that the whitefamily will be credit-worthy and the blackfamily unqualified in the eyes of the lender.

A black homeseeker reported that he hadcontacted a white broker to purchase a house ina white suburb. The broker carefully reviewedhis credit report, which reflected currentstability but showed a pattern of late payments3 years previous owing to illness. In apparentresponse to the report, the broker discouragedthe homeseeker from pursuing further hisdesire to buy in the suburbs and steered himtoward a "more suitable" section of town(a minority neighborhood),12 which the brokerfelt would be more acceptable to the lender.The black homeseeker subsequently changedbrokers, found a suburban house, and obtaineda mortgage.

Another crucial variable in the broker-buyerrelationship is the race of the broker. Minoritybrokers serving minority families who arelower middle-income buyers view their role asone of counselor, confidant, and benefactor.They emphasize the importance of developingrapport with homeseekers and offering advice,based on their knowledge of lending institutionpolicies and practices on mortgage qualifica-tions. This advice constitutes a screening ofwould-be mortgage applicants before the lender

firms in Hartford and 80 percent of the black-ownedfirms.

12 The U.S. Department of Justice charged sevenHartford real estate companies in May 1974 withpromoting resegregation by steering black home buyersinto integrated areas and white home buyers awayfrom them. The areas are Blue Hills and the suburb ofBloomfield. All of the firms are white owned andoperated.

is in any way involved.13 The brokers encouragecomplete disclosure of social and financial data;then, if they judge the homeseekers "respon-sible," they make an appointment with a lender,coach the buyers in preparation for the inter-view, and accompany them to the lender'soffice.

White brokers, who generally serve moreaffluent and knowledgeable homeseekers,develop a much different relationship. Theymay advise the buyer to some degree, but it isthe applicants' responsibility to canvas themortgage market and arrange financing. Themost realtors will do is provide buyers withthe names of several institutions in the areaand give some indication as to the kinds ofterms they can expect to be offered.

By contrast, less affluent purchasers, a dis-proportionate number of whom are minorities,frequently need help in obtaining the necessaryfinancing. This is especially true for first-timepurchasers, who tend to be unfamiliar with thecomplexities of mortgage finance. In seekingthis help, they turn to real estate brokers.

If it appears that a minority family isfinancially able to buy a home, then the brokerhelps them to obtain a mortgage. A numberof the brokers interviewed in Hartford reportedthat they arrange mortgages for up to 90percent of their sales. While this practice hasbeen construed as a service to the families, theprequalifying process may be detrimental tohomeseekers. The brokers admitted, forexample, that they automatically run creditchecks on minority families, but not on whitefamilies unless special circumstances appearto require it.

If brokers inaccurately assess the financialstatus of homeseekers, they may use any one ofa number of ways to dismiss and discouragethem. Homeseekers may be led to believe thatthey are ineligible when, in fact, they have beenvictims of deception, discrimination, or both.A case in point is that of a black couple whocontacted a white broker in response to anadvertisement for a house. The broker asked

!3 Although not all were asked, 11 of the Hartfordbrokers interviewed by Commission staff assertedthat mortgage lenders expect them to screen would-beapplicants.

10

about the husband's income and employmentand then informed them that they did not meetthe "income standards of the community."He did not state the asking price of the houseor ask how much money the couple had for adownpayment. (They had $5,000.) He simplylooked through his listings and told them thathe had nothing he "felt was suitable" for them.Only later did the couple think that they mighthave been discriminated against, but in theprocess they were prevented from buying thehouse they wanted.

Families deemed unqualified by a broker forthe mortgage they seek are so advised infor-mally and the matter usually ends there. Nowritten record is kept, nor is a lender involvedand put in the position of having to reject theapplicant. Consequently, acquiring documenta-tion on the extent of discriminatory rejectionof minorities by brokers is nearly impossible.

Just as minority home buyers are dependentupon brokers to obtain a mortgage, minoritybrokers in Hartford have provisional relation-ships with white lenders. Compared with whitefirms, minority firms are smaller, with lowersales volume and fewer mortgages to arrange.Thus, they deal with lenders on fewer occa-sions ; and, when they do, they feel they mustpresent only well-qualified families becausethey do not have the sales volume to use asleverage in favor of marginal applicants.

Black brokers are forced to be more cautious

than white brokers in dealing with lenders.If they feel that a lender is discriminating,they are reluctant to make an issue of it forfear of offending the lender. One black brokerreported that she applied for a mortgage forherself to a lender with whom she dealtregularly. Her income was $13,000 and shewas ready to put $2,000 down on a $20,500house. Her credit report was favorable. Thelender, however, was unusually slow in process-ing the application and excessively demandingin the documentation required. For example,he stipulated that cash for the closing inaddition to the downpayment had to be in thebank prior to approval of the loan. She thenapplied to another lender and swiftly obtaineda mortgage. Although this is not an overt caseof discrimination, the broker believes that shewas discriminated against because of her raceand/or her sex. But she would not press theissue because it might threaten her workingrelationship with the lender.

Hartford brokers contend that they judgethe ability of a potential mortgage applicant onthe same basis as the lending institution. Theyargue that because they arrange mortgages sofrequently and are in continuing contact withloan officers, they are thoroughly familiar withthe criteria used by the institutions. In fact,it is virtually impossible for brokers to knowprecisely the policies of lending institutionsbecause lending criteria vary widely among

TABLE 3—LENDER POLICIES ON BROKER SCREENING OF APPLICANTS

LendingInstitution

1 (S&L)

2 (S&L)

3 (Bank)

Source: Commission staff interviews.

OfficialInterviewed

President &vice presidentBranch managerVice presidentBranch managerBranch managerLoan officerVice presidentBranch managerBranch manager

Do you expect thebrokers to have

screened the applicant?

YesYesYesYesNoNoNoYesNo

Location ofOffice

Main officeNonminority areaMain officeMinority areaNonminority areaMain officeMain officeMinority areaMinority area

11

lending institutions in Hartford and also varyto a large degree among different offices of thesame institution.

Lending institutions also differ on whetheror not brokers are expected to screen potentialapplicants. As table 3 shows, only one of threeinstitutions surveyed by Commission staff hasa consistent policy on broker screening. Screen-ing requirements in the other two are deter-mined by individual loan officers. Nonetheless,a number of Hartford brokers act on theassumption that they are expected to conductscreening.

The absence of consistent objective lendingcriteria, coupled with the brokers' desire toremain in the good graces of lending institu-tions, tends to make brokers overcautious.They, thus, screen out families whom lendersmight well approve. Several brokers concededthat, unless a family fits the traditional charac-teristics of acceptable mortgage loan applicants,they will discourage the family because theyfeel that the loan officer will not approve aloan.

In short, if brokers are to err in theirevaluation of an applicant's qualifications, theywould rather err on the side of conservatism.One lending official reported that, on severaloccasions when applicants referred by brokershave been rejected, the brokers have felt com-pelled to apologize for having inadequatelyscreened the applicant. The same official alsoexpressed the view that brokers could beconsiderably more liberal in their approval.

The process of broker screening also opensup broad possibilities for rejection on discrimi-natory grounds. Brokers may screen outqualified minority applicants on the basis oftheir personal bias or their perception, rightor wrong, of the lender's discriminatory poli-cies. Rejection at this stage is informal. Nospecific reason need be given; discriminationis, therefore, difficult to detect.

THE ROLE OF LOAN OFFICERSIf minority applicants successfully clear the

hurdle of broker screening, they still must passanother test before reaching the stage atwhich their formal application for a mortgageloan is considered and decided upon by the

lending institution's loan committee. This is asecond screening process, this time by a loanofficer. As in the case of broker screening, thisstage is informal and based, at least in part, onsubjective criteria. The loan officer can rejectapplicants for any reason, including personalbias, without having to explain why.

The basic concern of lenders in determiningwhether to approve a mortgage loan is securityagainst loss. One consideration is the value ofthe property on which the mortgage is held.Another is the credit-worthiness of the bor-rower. For the latter purpose, lenders inquireinto such seemingly objective factors as income,occupation, length of employment, age, andcredit rating. They also inquire into less tan-gible factors that clearly call for subjectivejudgments. For example, one standard text onmortgage credit risks states:

In judging a borrower's reasons for request-ing a loan, the lender should consider thestrength of his [her] attachment to theproperty and his [her] probable futureattitudes toward it.14

The text goes on to assert:A borrower's relationship to his [her]family and friends is a significant element ofrisk although it is difficult to rate. Evaluatorsusually consider whether a borrower has anestablished reputation, a harmonious homelife, associates with good reputations, and ifhe [she] is active in civic affairs or whetherhe [she] has been dishonest and untruthful inthe past, has a troubled family life, and asso-ciates of doubtful reputation.15

Compounding the problem caused by thesubjective nature of the factors considered bythe loan officer is the fact that, at this stage ofthe mortgage application process, decisionsare made informally. The same text advises:

14 U.S., Department of Housing and Urban Develop-ment, Mortgage Credit Risk Analysis and Servicingof Delinquent Mortgages, by Anthony D. Grezzo(Washington, D.C.: Government Printing Office, 1972),p. 14.

15 Ibid. See also Robert H. Pease, ed., MortgageBanking (New York: McGraw-Hill, 1965), p. 216, whostates that some lenders visit the applicant's home todetermine personal habits that indicate "thoseimportant intangibles . .. such as pride of ownership,general housekeeping standard, and reputation withinthe immediate community."

12

An analysis of the credit risk should includean informal interview between a representa-tive of the lending agency and the borrowerand his wife [her husband].... The results[including a verification of employment,bank account, and a check against publicrecords] should enable the lending agency todecide if the borrower should submit a formalapplication for a loan, or if he [she] shouldbe told that his [her] application would beaccepted.16

Commission staff investigations in Hartfordindicate that lending institutions in that cityadhere closely to the above standards andprocedures. Savings and loan officers makedecisions on the suitability of filing an applica-tion on a purely impressionistic basis. In fact,one savings and loan association presidentreported that only "eligible" persons areallowed to apply for a mortgage. When askedto elaborate concerning what he meant by theterm eligible, he said that it was based onhis "feeling about the applicant."

A vice president of another Hartford mort-gage loan institution told Commission staffthat he placed great stock in whether theapplicants were "highly motivated" in decidingwhether to accept their application. Thesecriteria obviously open up broad possibilitiesfor decisions on the basis of prejudice or otherirrational factors. Nonetheless, they arewidely used.

The subjective latitude exercised by a loanofficer in judging the merits of individualapplications is reflected in the case of a blackcouple who were rejected because they had"a lot of debts." The couple's combined incomewas $18,700, which normally would havequalified them to purchase the $32,000 house oftheir choice. They had on hand the 20 percentdownpayment required. Yet because of a longterm debt of $3,000, their application wasrefused by a white loan officer who argued thatthe debt was too much for them to handle inaddition to the mortgage payments. The sameloan officer counsels all minority families tosave $15,000 before attempting to purchase ahouse. The couple subsequently obtained amortgage at another institution.

Officials at five lending institutions reported16 Mortgage Credit Analysis, p. 9.

that, following informal interviews at whichvarious subjective factors are considered, theyare expected to discourage applications frompeople they consider "ineligible." No reasonsneed be offered to the rejected applicants,nor is any written record kept to provide abasis for complaint. Applicants are merelyinformed that they should not make formalapplication because it will be rejected, reveal-ing another opportunity for discriminatoryrejection of minority applicants.

Beyond this, loan officers tend to be over-cautious in accepting loan applications andforwarding them to the loan committee.Several loan officials explained that theirfuture careers are determined in part by thedefault rate on applications they recommendfor approval to the loan committee. Thus, theytry to avoid accepting applications which theyfeel might be rejected by the loan committee,since this would reflect adversely on theirjudgment and might hinder their careeradvancement.

These policies inevitably result in informalrejection of applicants who might well qualifyfor a mortgage and never default on it. At onesavings and loan institution, two branchmanagers reported that 80 to 85 percent of theapplications they forward to the loan com-mittee are approved. Two branch managers ata savings bank similarly indicated that nearlyall of the applications they accept are approvedby the loan committee. At both institutions,submission of an application is tantamountto approval. One can only speculate, however,how many applications would have beenapproved by the loan committee had they notbeen informally rejected at an earlier stageof the process.

OBJECTIVE CRITERIAAlthough lending institution officials make

substantial use of subjective criteria such asmotivation and character in determining anapplicant's qualifications for a mortgage loan,they still place heavy reliance on criteria whichappear to be objective and susceptible to fairlyprecise measure. These criteria seeminglywould not permit the latitude for decisionsbased on personal bias.

13

One such criterion is the income of theapplicants in relation to the monthly paymentsthey will be required to pay under the mort-gage. The lending institution is legitimatelyconcerned that the applicant's income be suffi-cient to afford the monthly payments, so as toassure against default and possible financialloss to the lender.

The credit-worthiness of the applicant isanother legitimate concern of the lendinginstitution. An applicant with an unstablefinancial past or a history of failing to satisfydebts is a questionable risk and the lender isobliged to be as sure as possible that the appli-cant, if approved, can be expected to meet thelong term substantial payments that themortgage involves.

Another legitimate concern of the lendinginstitution is the value of the house beingmortgaged. It is the lender's obligation to setthe amount of the mortgage loan -in reasonableproportion to the value of the property, afterconducting an appraisal of the property todetermine its fair market value, which maydiffer from the sales price.

Although these three criteria are traditional

mortgage credit standards, in Hartford theyare applied inconsistently and offer ampleopportunities for decisions based on subjectivejudgments, including personal bias.

RATIO OF MONTHLY PAYMENT TO INCOME

A common formula lenders use is that oneweek's income should be equal to or greaterthan the monthly mortgage payment. Hartfordlenders, however, use a variety of otherformulas as well. (See table 4.)

Two institutions surveyed by Commissionstaff have no official policy on the ratio ofmonthly payment to income and leave thedecision on an acceptable ratio to the individualloan officer. Although most loan officers are ina position to judge an acceptable ratio based onexperience and training, without official policiesor guidelines it is difficult to hold them account-able. At these institutions the apparentlyobjective criterion of the ratio of monthlypayment to income can easily foe appliedsubjectively and discriminatorily.

Seven institutions were surveyed that do haveofficial policies on the maximum ratio of

TABLE 4—LENDER POLICIES ON RATIO OF INCOME TO HOUSING PAYMENTInstitution

(S&L)

(S&L)

(Bank)

(S&L)

Official InterviewedPresident and vice

presidentBranch managerVice presidentBranch managerBranch managerVice presidentBranch manager

President

Vice presidentVice presidentVice presidentVice presidentVice president

PolicyPITI* = 25% of groas or net income

PITI = 25 % of gross incomePITI = 25% of net incomePITI = 25% of net incomeHousing expenses** = 30% of gross incomeHousing expenses = 35 % of net incomePITI plus fixed debts = 38 to 40% of net

incomeHousing price = twice the gross annual

incomePITI = 25% of gross or net incomeHousing expenses = 35 % of gross incomeNo policyHousing expenses = 25 % of gross incomeNo policy

Source: Commission staff interviews* PITI is the sum of the Principal and Interest payments on the mortgage, Taxes and property

Insurance costs.** Housing- expenses are the sum of PITI, utilities, heat, and maintenance costs.

14

monthly payment to income. The two institu-tions with the most liberal ratios (a bank anda savings and loan association) require that oneweek's gross income equal the principal,interest, taxes, and insurance (PITI). Threeinstitutions use differing ratios which arebased not only on PITI cost but on total hous-ing expenses, including heat, utilities, andmaintenance, in relation to gross income. Twoother institutions use ratios .of net income tohousing payments in assessing applicants.

Thus, there is little consistency among thenine lending institutions on the permissibleratio of monthly payment to monthly income.Moreover, there is also little consistency in thisregard among the different offices of the sameinstitution. At one savings and loan association,for example, while the vice president in chargeof residential mortgage lending and one branchmanager were in accord that the ratio should be25 percent of net income to equal the PITI cost,another branch manager operated under theformula that 30 percent of gross income shouldequal total housing expenses.

Even on a day-to-day basis a bank may alterits preferred ratio of housing payment toincome, permitting ample opportunity fordiscrimination. For example, a black couplesought a loan to purchase a condominium in anew development. They went to the lender thathad financed a number of the new condo-miniums and were able to satisfy the institu-tion's standard ratio requirement. The lenderturned them down, however, on the groundsthat they could not meet a newly imposedhigher ratio. Subsequently, they obtained a loanfrom a different institution and were the firstblacks to move into the development.

In addition to inconsistent application of themonthly payment to income ratio amongHartford lenders, there also is inconsistency inthe definition of the basic term, income. Gen-erally, income can include salary earned 'by theprimary borrower and secondary borrower,along with commissions, overtime pay, andbonuses. All Hartford lenders agree that theincome of the primary borrower should be

stable, reliable, and have reasonable prospectsfor continuation. These same criteria apply tothe income of the secondary borrower.17 Theconditions under which overtime pay andincome from second jobs are counted, however,are not uniform among the institutions.

Three institutions allow overtime and secondjob income as effective income, provided thatthe same criteria as applied to the primary jobare met. Two institutions do not necessarilyallow such income to be counted even if thecriteria are met. Each case is treated individ-ually, and the basic criterion of income becomesless susceptible to objective definition.

The failure of Hartford lenders to acceptincome from overtime pay and second jobs ona consistent basis necessarily has a discrimi-natory effect on minority homeseekers, whooften rely on these sources of income.18 TheFederal Housing Administration, however,accepts overtime pay as part of income whensuch pay is characteristic of the job. FHA alsoaccepts income from a second job if the appli-cant has held it over a substantial period oftime and is expected to continue in it duringthe early period of the mortgage.

CREDIT REPORTS

Reports on the credit history of mortgageapplicants are necessary information for thelender to have in deciding .whether to approvean application. Again, Hartford lenders are notconsistent in their use of these reports. Somelenders limit their examination of an appli-cant's credit history to the previous 2 years,reasoning that this is sufficient to determine thefinancial responsibility of the applicant. Others,however, extend their examination to theprevious 5 or even 7 years. By the same token,while some lenders do not accept at face value

17 Issues relating to the secondary borrower, who isoften the wife of the primary borrower, are discussedin ch. 4.

18 This argument is supported by guidelines releasedby the Federal Home Loan Bank Board in December1973. "Automatically discounting . . . income frombonuses, overtime, or part time employment will causesome applicants to be denied financing. ... Sincestatistics show that minority group members and lowand moderate-income families rely more often on suchsupplemental income, the practice may be raciallydiscriminatory in effect, as well as artificially restrictiveof opportunities for home financing." 38 Fed. Reg.34653 § 531.8(c).

15

adverse reports on a family's credit but lookfor possible mitigating circumstances or otherclarifying information, other lenders assertthat any adverse information in the reportautomatically disqualifies the applicant. Noinvestigation is conducted to determine thecircumstances surrounding the adverse reportor even to determine whether the adverse infor-mation is accurate. The discriminatory effectof this policy of automatic disqualification onminority applicants is apparent, owing to thediscriminatory judgments made by creditbureaus in assigning credit ratings.19

A 1971 survey of savings and loan associa-tions by the Federal Home Loan Bank Board(FHLBB) revealed that 57 percent of theassociations looked at credit reports for theprevious 2 years or less; 34 percent consideredthe previous 3 to 5 years and 9 percent lookedat the previous 6 to 7 years.20 The survey alsorevealed that 26 percent of the associationssought information on arrestjcecords, and 12percent automatically disqualified applicantswho had ever been arrested, whereas 14 per-cent disqualified applicants if they had beenconvicted. Because minority persons arearrested in disproportionate percentages, thedisqualification on arrest records alone has adiscriminatory effect.21

Similarly, 73 percent of the associations re-quire information on previous homeownership,and 23 percent (automatically disqualify appli-cants with "unsatisfactory answer (s)."Because minority persons are overwhelminglynot homeowners in metropolitan areas, this

19 See S.N. Sesser, "Big Brother Keeps Tabs onInsurance Buyers," New Republic, April 27, 1968,pp. 11-12.

20 Federal Home Loan Bank Board Survey (releasedMarch 1972). The FHLBB considered the results ofthe survey inconclusive, since it included only 74 ofthe 5,000 federally-supervised savings and loanassociations.

21 Figures taken from the Uniform Crime Reportpublished by the Federal Bureau of Investigation in1973 and interpolated with national populationstatistics show that the arrest record of blacks wasmore than three times that of whites in 1972 inproportion to their percentage in the population;Spanish speaking persons are included in the category"white."

imposes an additional, undue burden on them.22

APPRAISALS

If a mortgage is needed by most families tobuy a house, it also is true that the mortgagemust be large enough. If not, families musteither obtain a second mortgage, which can beexpensive, or abandon their efforts to purchasethe house they desire. The amount of themortgage a lender will offer to an applicant isbased largely on the appraisal, which deter-mines the worth and marketability of thehouse. One traditional way of discriminatingagainst minority homeseekers is through under-appraisal of the houses they wish to purchase.23

Commission staff did not attempt to conducfan in-depth investigation of the extent to whichdiscriminatory underappraisal was prevalent inHartford. Instead, knowledgeable local lendersand borrowers were asked for their expertopinion. As expected, most stated that discrimi-natory underappraisal was not a commonpractice, that most homes were appraised atthe sales price.

Officials of one lending institution and threebrokers, however, reported that appraisalvalue was lower than the sales price in asubstantial number of cases. In their view, thishappened most often when the purchaser was aminority group member or when the house waslocated in an area of minority concentration.24

Transitional neighborhoods—those in the

22 Only 34 percent of black families living in centralcities and just 14 percent of the Nation's urbanPuerto Rican families were homeowners in 1970. U.S.,Department of Commerce, Bureau of the Census, 1970Census of Housing: Detailed Housing Characteristics—Final Report, no. HC(1)-B1, U.S. Summary (July1972), p. 292. Also, U.S., Department of Commerce,Bureau of the Census, 1970 Census of Population:Puerto Ricans in the United States—Subject Reports,no. PC(2)-1E (June 1973), p. 94.

23 Underappraisal may be discriminatory in effectwhen it results in a lowered loan amount and a higherdownpayment, thus forcing the homeseeker out of themarket.

24 According to the FHLBB survey, only 4 percentof the savings and loans use minority appraisers. Thus,the appraisers are overwhelmingly white and undoubt-edly reflect the views of the associations: 30 percentof the associations "disqualify some neighborhoods

16

process of integrating—very likely fall into thiscategory. They added that often the under-appraisal was made without discriminatoryintent, but rather on the basis of the traditionalappraiser view that property values decline inminority and transitional neighborhoods.25

They conceded, however, that sometimes theappraisals were the result of nothing more thanthe personal bias of the appraisers.

from lending because they are low-income or minority-group areas" and 78 percent feel loans in such neigh-borhoods are "more risky than other loans." Addi-tionally, 11 percent of the associations' appraisers use"different methods or factors" for such neighborhoods.For purchases in such areas, over a quarter of theassociations require higher downpayments; 11 percentlevy interest rates one-half of one percent higher, and

almost a third give terms 7% years shorter than loansfor homes in other areas.In May 1972 the Department of Housing and UrbanDevelopment (HUD), Office of Equal Opportunity,issued an initial report on a private lending institutionsquestionnaire. It showed that 18 percent of the associa-tions surveyed refuse to make loans in areas with highminority concentrations.The new FHLBB guidelines prohibit "redlining," therefusal to lend in a neighborhood solely because ofminority concentration. "The racial composition of theneighborhood where the loan is to be made is alwaysan improper underwriting consideration." 88 Fed. Reg.34653 § 531.8(c).

25 This traditional appraiser perception is prohibitedby the new FHLBB guidelines: "Refusal to lend in aparticular area solely because of the age of the homesor the income level in a neighborhood may be discrimi-natory in effect since minority group persons are morelikely to purchase used housing and to live in low-income neighborhoods. 38 Fed. Reg. 34653 § 531.8 (c).

17

Chapter 4Sex Discrimination in Mortgage Lending

Unequal access to mortgage money for Hart-ford's minorities occurs largely by virtue ofmortgage procedures and criteria which permitand even facilitate decisions based on personalbias and other factors not related to objectivelending criteria. Whether because discrimina-tion in mortgage lending is prohibited by bothConnecticut and Federal law or for otherreasons, lenders in Hartford do not generallyadmit that they reject applicants on the basisof their race or national origin. In fact, thecriteria which govern whether mortgage ap-plicants will be approved or disapproved are,at least on the surface, nondiscrimiriatory. Tothe extent discrimination does occur, it issubtle, often unconsciously practiced, anddifficult to detect.

Discrimination on the basis of sex is adifferent matter.26 Here, the major problem isnot that mortgage procedures or criteria permitopportunities for decisions on the basis ofdiscrimination. Rather, traditional mortgagelending criteria followed by Hartford mortgagelenders virtually require sex discrimination.27

Under these criteria, women are automaticallyconsidered suspect risks.

26 Sex discrimination in mortgage lending currentlyis not prohibited under Federal law. The Federal HomeLoan Bank Board, which regulates Federal savings andloan associations, issued guidelines in December 1973on nondiscriminatory lending practices with respectto age, sex, and marital status. 38 Fed. Reg. 34653.The Federal Deposit Insurance Corporation, whichsupervises most of the Nation's commercial banks, hashad under consideration for well over a year, but hasnot acted on, the issuance of a regulation prohibitingsex discrimination in lending.

27 In Connecticut, in June 1973, a State law waspassed which prohibits discrimination on the basis ofsex or marital status in credit transactions including

If married and working, women's incomesare discounted for purposes of determining thefamily's eligibility for a mortgage. No matterhow important their income is to the familybudget, it is considered "secondary" for mort-gage lending purposes, and the family's chancesfor a mortgage loan are decided largely on thebasis of the husband's financial status.

If unmarried, women are viewed with greatskepticism under traditional mortgage lendingcriteria. Regardless of their professional back-ground or work experience, their status asunmarried women renders them suspect creditrisks. Female heads of household who areseparated or divorced also face unfavorabletreatment by lenders. Separated women are inan ambiguous legal status in terms of debtliability, while both separation and divorcebear a traditional social stigma.

This chapter examines the barriers to home-ownership only of those women who areeconomically capable of owning a home. Theability to purchase a house depends directly onhousehold income, and on this point, womenwho are black, Spanish speaking, and white areat a severe disadvantage compared to men.28

(See table 5.) Moreover, the impact of discrimi-nation is felt doubly by black and Spanishspeaking women who are penalized 'because ofrace and ethnicity as well as sex.

Sex discrimination in mortgage lending isnot nearly as difficult to detect as discrimination

mortgage lending transactions. Conn. Public Act73-573. Evidence for this chapter was gathered priorto passage of this law; its impact has not been analyzedhere.

28 Median earnings of women in the Hartford SMSAin 1970 were only 47 percent of men's earnings.

18

TABLE 5—WOMEN'S EARNINGS AS PERCENTAGE OF MEN'S EARNINGS, 1969

Men*White $8,601Black 5,809Spanish speaking 6,606

Women*$4,0843,2803,621

Women's Earnings as Percentageof Men's Earnings

5655

Source: U.S., Department of Commerce, Bureau of the Census, 1970 Census of Population: Earnings byOccupation and Education—Final Report, no. PC(2)-8B (January 1973), tables 1, 2, 7, 8.

*Ages 25-64

Source: U.S., Department of Commerce, Bureau of the Census, The Social and Economic Status of theBlack Population in the United States, 1971, P-23, no. 42 (July 1972), p. 34.

* Heads of household under 35 years of age only.

on the basis of race or national origin. Much ofit is based on what lenders consider prudentand objective criteria. In Hartford, as else-where, sex discrimination is part and parcel ofofficial bank policy.

This chapter will detail the various formsthat sex discrimination in mortgage lendingtakes in Hartford. Women of different maritalstatus—married, unmarried, widowed, sep-arated, and divorced—are all viewed somewhatdifferently by the mortgage lending community.No group of women, however, has equal accessto mortgage money.

THE WORKING WIFEMarried women are a substantial part of theNation's labor force. The stereotyped patternof women giving up their jobs once married tospend full time caring for the house andchildren clearly is no longer true. As of 1970,in two of every five families with husband and

wife both present, both the wife and thehusband worked. Increasingly, the workingwife's income is being relied upon as a substan-tial and continuing part of a family's assets.

Important as a working wife's income is tofamilies generally, it often is essential tominority families. As of 1970, in more thantwo-thirds of the Nation's black families withhusband and wife both present, both worked.Table 6 demonstrates the importance of thewife's income to the economic well-being ofblack families.29 Among families in which onlythe husband worked, iblack family income wasless than two-thirds of that for white families.By contrast, among families in which bothhusband and wife worked, black family incomewas nearly 90 percent of the income for whites.

29 Work experience and income data on two-earnerSpanish speaking families are not available. However,median household income variation among the threegroups would suggest that a similar pattern occurs intwo-earner Spanish speaking households.

20

TABLE 6—MEDIAN INCOME OF FAMILIES WITH BOTH HUSBAND AND WIFEPRESENT, BY WORK EXPERIENCE AND RACE

Work DollarExperience Amount

Black, total $3,534Only husband worked 3,025Husband and wife worked 3,845

White, total 5,658Only husband worked 5,233Husband and wife worked 6,013

Black asPercentage

of White62%5864

DollarAmount$8,032

5,9659,2679,7969,065

10,396

Black asPercentage

of White82%6689

19701959

47 ?&

TABLE 7—EMPLOYMENT OF WIVES, BY RACE

White

Number PercentageSample 34Total wives employed 14

Part-timeFull-time

1526

BlackNumber Percentage

38224 11

18 47

TotalNumber Percentage

72369

27

50%1238

Source: Commission staff interviews with home purchasers, Hartford, Conn., September 1972.

This nationwide pattern is reflected in Hart-ford. In 45 percent of white husband-wifehouseholds in the Hartford metropolitan area,and 60 percent of black husband-wife house-holds, the wife works.

The importance of the working wife's incomein achieving the goal of homeownership isshown by data from a sample of mortgagetransactions in Hartford, for the years 1971-72.30 As table 7 shows, in half of the 72 familiesin the siample, the wife worked. This was truefor 41 percent of the white families and 58percent of the black families.

These statistics strongly indicate that, to theextent mortgage lenders discount some or allof the working wife's income in determiningwhether to approve a mortgage loan applica-tion, all families necessarily are penalized. Forminority families the penalty may be doublysevere because this form of sex discrimination,coupled with discrimination on the basis of raceor national origin, effectively places minoritywomen and their families in double jeopardy.

Traditionally, lenders have ignored theworking wife's income in assessing a family'sfinancial status. Recently, however, because oftheir growing recognition of the fact thatmarried women are a (substantial part of thelabor force and because of pressure from vari-ous public interest groups,31 they have come to

30 Data gathered through interviews in 1972 withhome purchasers. In only one instance was the femalethe head of the household. She was a nurse whoseincome was fully considered by the lenders.

si For example, a statement in opposition to theFederal National Mortgage Association's proposedguidelines restricting mortgage credit was issued in

count some portion of it. Nonetheless, the poli-cies and practices of mortgage lenders still fallfar short of fully accepting the income ofworking wives.

The 1971 survey by the Federal Home LoanBank Board on practices of savings and loanassociations showed that 25 percent of therespondents would not count any of the incomeof a wife, age 25, with two school children,who held a f ull-time secretarial position. Morethan half of the mortgage lending institutionswould limit credit to 50 percent or less of hersalary. Only 22 percent would count all of it.

The results of the survey not only show theoverly conservative view of mortgage lenderstoward a wife's income, but also demonstratethe lack of any uniform policy. This is reflectedin Hartford. As table 8 illustrates, not only arethere inconsistencies in policy among the nineHartford lending institutions surveyed by Com-mission staff; but, even within the same insti-tution, officers differ markedly in their viewof a wife's income.

The refusal to credit the working wife's in-come is based on the unsupported assumptionthat to do so would increase the risk of defaultand subsequent foreclosure. In fact, to date nostudies have controlled for this crucial vari-able. One study found, however, that as thepercentage of family income earned by thehusband decreased, the chance of a loan'sbeing delinquent actually decreased

1971 by 30 public interest groups including the NationalOrganization for Women, National Association for theAdvancement of Colored People, Center for NationalPolicy Review, and Nonprofit Housing Center.

21

TABLE 8--LENDER POLICIES ON WIFE'S INCOME

Institution1 (S&L)

2 (S&L)

3 (Bank)

4 (S&L)5 (Bank)6 (Bank)

7 (S&L)8 (Bank)9 (Bank)

Official InterviewedPresident and vice

presidentVice president

Branch manager

Branch manager

Vice presidentBranch manager

Branch manager

PresidentVice presidentVice presidentBranch manager

Vice presidentVice presidentVice president

Policy100% allowance

100% if over 35 and employed 2 or 3 years;if under 35, would take a "hard look";might require a "'baby letter."*

100% if 29 or older, with no children and ina professional occupation.

Requires professional occupation; would takea "hard look" at younger than 30.

100% allowance50 % allowance unless in a professional occu-

pation, with no children and a "babyletter."

75% allowance fora professional occupationif under age 30; less allowed for others.

50% allowanceDepends upon the case100% allowanceDoes not count income from nonprofessional

occupations.Depends upon the caseDepends upon the case100% allowance

Source: Commission staff interviews with banking and savings and loan o _cials in Hartford, January-February 1973.

* The "baby letter" is a physician's statement attesting to sterility of husband or wife, their use ofapproved birth control methods, or their willingness to terminate pregnancy.

slightly.32 Another more recent study, whichrelated borrower characteristics to delinquencyand foreclosure rates, found that none of thecontributing factors related to the income ofthe working wife.33 In short, there is no empiri-cal evidence to support the widely practicedlender policy of discounting the working wife'sincome.

POLICIES OF CONVENTIONAL LENDERS

Age and Children.—A primary considerationin mortgage loan underwriting is the female

32 Leon Kendall, Anatomy of the ResidentialMortgage Market (Washington, B.C.: U.S. Savingsand Loan League, 1964).

33 John Herzog and James Barley, Home MortgageDelinquency and Foreclosure (New York: NationalBureau of Economic Research, 1970).

wage earner's age. This is directly related tothe probability of childjbearing. For example,a married woman in her twenties generallywould not have more than 50 percent of herincome counted, owing ,to the likelihood thatshe will bear children and, it is assumed, leavethe labor force.34 By contrast, 75 to 100 percentof the income of a married woman in her latethirties would qualify, 'according to Hartfordlenders.

In addition to the wife's age, the number and

34 Nationally, 38 percent of all women in the laborforce have at least one child under 18 years of age.Forty-three percent of all ever-married women withchildren under 18 are in the labor force. U.S., Depart-ment of Labor, Bureau of Labor Statistics, Maritaland Family Characteristics of Workers: March 1972(April 1973), p. A-18.

22

ages of children living at home are also con-sidered. Lenders assume that families withyoung children will have additional childrenand discount the working wife's income accord-ingly. There are, however, exceptions. Amarried woman, even in her twenties, with nochildren may have all of her income counted,with certain documentation. This documenta-tion, known in the lending trade as a "babyletter," consists of a physician's statementwhich attests to her 'or her husband's sterility,their use of approved birth control methods, ortheir willingness to terminate pregnancy.

Because of the absence of uniform institu-tional policy, the requirement of a 'baby letterdepends entirely on the attitudes of individualloan officers. Branch managers in two lendinginstitutions stated to Commission .staff thatthey required the baby letter as a preconditionto crediting all the income of a young wife.However, at the central offices .of these sameinstitutions opinion differed as to the need forthese letters.35 The lack of consistent policy andresultant confusion is illustrated by the follow-ing cases.

A young Hartford couple applied for a 90percent loan on a $16,000 home. The husbandearned approximately $10,000 a year; his wife,$9,500. Their only major long term obligationwas a monthly $200 car payment. At first thesavings and loan the couple applied to for amortgage loan was reluctant to approve themortgage based on the husband's income alone.In order to count the wife's income, the institu-tion asked for a baby letter. The couple refused.Ultimately, the loan was approved, but thecouple was never informed what part, if any, ofthe wife's income was counted.

In another case a married woman was askedby the broker-builder-seller of the property tofile a baby letter, which she did. Subsequently,however, the lending institution assured herthat no letter was on file. The institution alsoinformed her that her income had not beencounted. In this case the broker assumed that

the wife's income was necessary and that ababy letter had to be presented. Neitherassumption was correct.

Working wives with preschool age childrenare the least likely of any female subgroup tohave their income fully counted towardsmaximum mortgage allowance. Lenders raisequestions and make determinations relying ontheir own assumptions about the likelihood ofmore children, the costs of child care, and thelength of time spent away from the job afterchildbirth. A branch manager at one savingsand loan told Commission staff that his deci-sions on crediting a working wife's income arebased, in part, on his observation that a womanusually does not return to work until a yearafter the birth of a child. During that time herincome will not be available for mortgage pay-ments. This attitude is common among Hart-ford lenders. They are reluctant to count morethan 50 percent, if that, of a young mother'sincome, predicting that she will again becomepregnant and drop out of the work force for atleast a year during the early mortgage period.

The case of one young Hartford coupleclearly illustrates the effect this policy has onthe ability to purchase a home. The husband,23 years of age, was a fifth grade teacher whoearned an annual salary of $8,574. His wife, 22years of age, was a secretary earning $5,600.They had two children, a 5 year old and aninfant.

The couple first contacted a savings and loanin October 1972 and applied for a $16,150, 8percent, 30-year mortgage on a house priced at$17,000. The wife was not then employed.Although the application was approved by theMortgage Guaranty Insurance Corporation(MGIC),36 it was subsequently rejected by thelender because the husband had cosigned anautomobile loan for his brother.37

Two months later after the husband hadremoved his name as cosigner on the auto loan,

35 At one lending institution, one vice presidentwould require a baby letter while another would not.In general, executives at the central level in the lendingestablishments favored full inclusion of the wife'sincome, unlike more conservative branch managers.

36 MGIC is one of a number of State-licensed privatemortgage insurance companies which insure lendinginstitutions against loss, generally on the top 20 percentof mortgage loans.

37 The rejection and reason therefor were communi-cated orally to the applicants by their broker.

23

the couple reapplied to the savings and loan.By this time the wife had assumed a positionas a full-time secretary with the University ofHartford. Several weeks later they received asecond notice of rejection from the lending in-stitution. This time the reason was that thewife was young, in her child-bearing years,and, therefore, likely to become pregnant anddrop out of the work force.

This case illustrates the perplexing, some-times erratic, behavior of mortgage lenders inprocessing applications from married couples.The fact that the application initially was ap-proved by MGIC, although subsequently re-jected by the lender, suggests that the couplewas, at worst, a marginal risk for the mortgagethey sought, even though the husband wassubject to liability as cosigner on an auto loanand the wife was not then working. Twomonths later, the husband's liability as co-signer had been eliminated and the wife wasworking. Given these two new sets of condi-tions, the couple had good reason for optimismand resubmitted their application. Nonetheless,it was rejected. The fact that the wife now helda responsible, full-time position and earned asubstantial income apparently counted forlittle in the lender's judgment.

Occupation.—Another consideration whichlenders take into account in crediting the work-ing wife's income is the type of job she holds.The income of women categorized as "pro-fessional" by lenders is counted more readilythan that of women whose jobs are considered"nonprof essional." Although some central officelending officials interviewed by Commissionstaff insisted that type of occupation is not adetermining factor in income allowance, mostbranch managers interviewed stated that theyconsistently differentiate according to profes-sional and nonprofessional categories. A womanwho is a store clerk or bank teller, for example,would not have as high a percentage of herincome counted as a woman who is a businessexecutive, teacher, or nurse. The presumptionsare that professional jobs are stable, as arethe women who hold them, whereas nonprofes-sional jobs are short term and unstable.

Type of occupation does not represent anindependent factor considered by itself in

determining whether and how much to creditthe working wife's income. The wife's age andthe number and ages of the children also figurein, and the various factors are weighed in animprecise way that defies objective analysis.One lending institution official said that hisrule of thumb in crediting wife's income wasthat, if the woman is in her child-bearing years,she must hold a professional position and, eventhen, not more than 50 percent of her incomewould be credited. He would not under anycircumstances, however, count any income of awoman in her child-bearing years who held ablue-collar job.

A branch manager of another lendinginstitution stated that he allows 50 percentcredit toward the income of a working wifeunder 35 and 100 percent credit of a workingwife over 35, regardless of the type of job sheholds, provided that she has been employed forat least 1 year. Another branch manager of thesame institution, however, stated that hewould never allow 100 percent of the incomeof a female blue-collar worker. Jobs in thatcategory, he said, are unstable.

The arbitrary character of the applicationevaluation with regard to wife's income isunderscored in the following case. A PuertoRican couple applied for a mortgage with asavings bank. They were both 29, childless, andhad been schoolteachers for the previous 5years, earning a joint income of $20,000annually. The couple applied for a $16,000mortgage on a $20,000 home but were told thattheir income was insufficient. They subsequentlyobtained a mortgage at a savings and loanwhere half the wife's income was counted,giving them an adjusted income of $15,500.

This case demonstrates two operational rulesof conventional mortgage lending in Hartford.First is the arbitrary nature of lender decisionson counting a working wife's income. At thefirst institution, rejection on grounds ofinsufficient income clearly suggests that theincome of the wife, who held a well-payingprofessional job, was totally discounted because,at 29, she was within child-bearing age. At thesecond institution half of the wife's income wascounted. Secondly, this case reflects the tradi-tional view of lenders that the income of the

24

working wife, under the best of circumstances,cannot really be relied upon. Even wherecredit was given to the wife's income, she wastreated as half a wage earner rather than as afull wage earner.

POLICIES OF FEDERAL AGENCIES

Federal Housing Administration (FHA).—The FHA's policy on giving credit to the wife'sincome traditionally differs somewhat fromthat of conventional lenders. FHA policy is tocount either all of the wife's income or none ofit. By the mid-1960's, according to FHArecords, all of the wife's income was counted in75 percent of the FHA mortgages made whereboth husband and wife worked. Today, thisfigure approaches 90 percent. The significanceof these statistics must be qualified by the factthat they apply only to mortgages that wereactually made. FHA does not maintain recordson unsuccessful mortgage applicants. Thus,there is no way of knowing how many familieswere unsuccessful in obtaining an FHA mort-gage because the wife's income was discounted.

The standard used by FHA in crediting theincome of working wives is that "income andmotivating interest may normally be expectedto continue throughout the early period ofmortgage risk."38 The underwriting manualstill addresses somewhat conservatively therisk involved in basing net effective income ontwo wage earners.

When the effective income is derived fromdual sources of occupational income, as in acase when both husband and wife areexpected to be employed during the earlyperiod of mortgage risk, risk due to possiblereductions in total occupational incomefrequently will be increased because of thegreater probability that one or the othermortgagor may suffer a loss of income. Thisfactor of risk is of particular importancewhen the dual income is represented by thesalaries of young married couples.89

On the relationship between the income of theworking wife and pregnancy, FHA softens itsposition:

The principal element of mortgage risk inallowing the income of working wives aseffective income is the possibility of itsinterruption by maternity leave. Mostemployers recognize this possibility andprovide for maternity leave, with job reten-tion, as an inducement of employment. Withstrong motives for returning to work anyfailure to do so after maternity leave wouldrepresent such a very small percentage ofvolume that it could be accepted as a cal-culated risk.40

The "strong motive" standard, however, isvague and open to individual interpretation.

FHA policy, while liberal at the nationallevel, is implemented at the local level accordingto the facts of each case, thereby permittingwide latitude in the exercise of judgment byindividual FHA officials. This local autonomycauses some disquiet and misunderstandingamong the real estate community. For example,despite the high national percentage of mort-gages made in which the income of bothhusband and wife is counted, two brokers inHartford believe that FHA will not count theincome of a married woman under age 36.The brokers' belief is not groundlessbecause the mortgage table on which manylocal brokers and lenders rely states, incor-rectly, that FHA will not count the income of awife under 32 years of age.41

Veterans Administration (VA).—Until veryrecently, the Veterans Administration treatedwives' income more restrictively than the FHA.In February 1973, the Washington Post pub-lished an article which alleged that a lenderrequired a veteran and his wife, who wereapplying for a GI loan, to disclose their methodof birth control through a physician's statementand to take measures to prohibit conceptionand/or terminate pregnancy in order to qualifythe wife's income.42 In response to this incident,

38 U.S., Department of Housing and Urban Develop-ment, Mortgage Credit Analysis Handbook forMortgage Insurance on One to Four-Family Properties(1972), sec. 1-22. See also, U.S., Commission on CivilRights, Housing, Report for 1961, vol. 4, ch. 3.

39 Mortgage Credit Analysis Handbook, sec. 2-21.2.

40 Ibid., sec. l-22.b.41 Mortgage Guides, Inc., The Mortgage Guide

(Portland, Oregon: 1962). The table summarizes FHAguidelines in use in 1962.

42 Washington Post, Feb. 24, 1973, p. A-l. See alsoletter from Carol K. Lewicke to the Federal Trade

25

the Veterans Administration issued a circularto its field stations which attempted to establishguidelines on treatment of wives' income.43

The circular denied the foregoing allegations:. . . it is not now and never has been the policyof the Veterans Administration to request ordemand veterans and their spouses to makeany such disclosures....

Further on, it qualified this assertion by saying:If such a medical statement (supportingevidence that a couple are unable to havechildren) is voluntarily submitted by theveteran to the lender, it cannot very well berefused upon receipt in VA. However,...VA would prefer that any such statementsreceived by builders and lenders be retainedby them.

The circular's counsel on the treatment ofwives' income was just as ambiguous:

A proper conclusion that the wife's incomemay be considered toward the repayment ofthe loan obligation requires a determinationas to whether her employment is a definitecharacteristic of the family life; i.e., acondition which normally may be expected tocontinue. Her entire income may be includedif it is derived from steady employment andher age, the nature and length of her employ-ment, and the composition of the familyindicate it is reasonable to conclude that suchincome is likely to be reliable in the future.Unless that condition is met, only such por-tions of the wife's income as is determinedto be reasonable may be considered.44

On July 18, 1973, however, the VeteransAdministration approved a new circularstating "in consideration of present-day socialand economic patterns, the Veterans Adminis-tration will hereafter recognize in full 'both theincome and expenses of the veteran and spousein determining the ability to repay a loan... ,"45

All of the VA's regional offices have beeninstructed that they should no longer discountincome on account of sex or marital status inmaking this determination.

Commission, Bureau of Consumer Protection, Dec. 20,1972, available in Commission on Civil Rights files.

43 U.S., Veterans Administration, Department ofVeterans Benefits Information Bulletin no. 26-73-1,Feb. 2, 1973.

44 Ibid.45 U.S., Veterans Administration, Department of

Veterans Benefits Circular no. 26-73-24, July 18, 1973.

The Federal Home Loan Bank Board.—Responding to a 3-year drive by 13 publicinterest groups, the Federal Home Loan BankBoard issued potentially far-reaching guide-lines in December 1973 on nondiscriminatorylending practices. The guidelines cautionagainst practices which may result in dis-crimination even without actual intent todiscriminate. They state that "each loan appli-cant's credit worthiness should be evaluatedon an individual basis without reference topresumed characteristics of a group."46

The guidelines describe those underwritingdecisions as discriminatory which distinguishcredit-worthiness on the basis of age, sex,and/or marital status. The discounting of aworking wife's income may result in discrimi-nation against race as well as sex, they argue,because of the greater reliance of the minorityfamily on the wife's income for housing andother necessities.

THE SINGLE WOMANThe women with the greatest difficulty in

gaining access to mortgage finance are singlewomen—unmarried, widowed, separated, ordivorced women. Each is treated somewhatdifferently by mortgage lending institutions,but for all it is their status as women who arenot part of a male-headed household that is ofgreatest significance to mortgage lenders.

Bias against the single individual is evidentin the FHA's underwriting manual:

The mortgagor who is married and has afamily generally evidences more stabilitythan a mortgagor who is single because,among other things, he [s'he] has responsi-bilities holding him [her] to his [her]obligations.47

Thus, unmarried persons—men or women—are at a disadvantage in seeking to obtain amortgage. This disadvantage, however, is notshared equally by men and women.

A widely held view in the mortgage lendingcommunity is that single women must present astronger paper position than single men. Theircredit and income must be more secure than

46 38 Fed. Reg. 34653 § 531.8 (b).47 Mortgage Credit Analysis Handbook, sec. 2-7a.

26

those of men of the same status, and theircredit histories must be more closely scrutinzed.As a result, the filtering process is applied morerigidly at each stage of the mortgage applica-tion procedure, from broker to branch managerto formal application. The myth generatingthis stringency holds that the female isinherently unstable and incapable of conductingher own affairs. She allegedly needs the pro-tection of a male, usually a husband or father.In the lending industry the myth translates intoa reluctance to grant a woman a mortgageloan outright and often, a requirement of anassumption48 or a male cosigner.

A peculiar consequence of the theory thatwomen need protection is the lenders' disin-clination to grant a loan to a woman who wantsto purchase a multif amily dwelling, reasoningthat she would be unable to perform the neces-sary maintenance. Normally, "a man would dothe repairs himself whereas a woman has tohire someone," one lending official argued.According to this official, VA policy is evenstricter than his own. He cited the case of awidow who worked part time and receivedsocial security. She wanted to buy a multi-family property. He approved the loan, but VArejected the application on the ground of insuf-ficient income. Based on his previous experiencewith VA, the lender felt that this excuse wasused to camouflage the true reason forrejection, which was the repair issue.

THE UNMARRIED WOMAN

In the case of an unmarried younger woman,the principal reason cited for denyingmortgages is the likelihood of marriage andpregnancy, and consequent shift in economicstatus, which serve as risk factors. One brokertold Commission staff that in his experience,of all categories of single women, the youngunmarried woman has the most troublesecuring a mortgage.

Another broker said that the only way anunmarried woman can get a mortgage isthrough an assumption. He tried unsuccessfully

48 In an assumption, ultimate responsibility restswith the original mortgagor. The second mortgagortakes on, or assumes, payments of the original loan.

to secure financing on a home for a professionalwoman who had worked for 23 years. Suchfailures not only disappoint the individual;they also influence the real estate communityto discourage similar clients. Assumptions,however, are considered a safety device; shouldthe woman default, the institution can holdthe principal mortgagor liable.

Policies vary both among and within institu-tions with regard to unmarried women.Officials of lending institutions at the centrallevel generally stress length of employment,while branch managers and brokers emphasizeage of applicant and type of job. One branchmanager went so far as to assert that anunmarried woman could obtain a loan only ifshe had a professional career. The marginalcase, according to a loan officer, would be thatof the older, unmarried woman in a nonpro-f essional occupation, such as a waitress orstore clerk, and with only a modest downpay-ment. Even if she has a reasonably long recordof employment, it is unlikely that she wouldbe approved for a home mortgage, he said,because her employment is considered unstable.

THE WIDOW

The widow applying for a home loan gener-ally has a better chance of obtaining1 it thanother single women. She more often can relyon life insurance proceeds, social securitypayments, or settlement on an estate to providea healthy downpayment and assure-a regularincome. One lending official said that althoughwidows represent a variety of ages and economiccircumstances, if they have a sufficient down-payment and a guarantee of regular income,their prospects for homeownership often arerelatively bright.

THE SEPARATED WOMAN

In Hartford, the separated woman, particu-larly if informally separated, is in an awkwardposition in trying to purchase a home. UnderConnecticut law, a husband is liable for his"wife's reasonable support while abandoned[by him]."49 The husband's liability extends

49 22 Conn. Gen. Stat. Ann. Ch. 809, § 46-10(1949).

27

even to cases where the wife has an independentincome.

One of the primary concerns of lenders inmaking a mortgage loan is assurance ofunambiguous liability. For this reason, theinformally separated woman falls into a highrisk category. If she were approved for amortgage loan and then defaulted on payments,lenders could not be certain that the husbandwould be required to make the payments, evenassuming he were financially able to do so.The reason for uncertainty is that thehusband's liability exists only if the separationresults from his abandoning the wife, not fromher abandoning him. Thus, under an informalseparation arrangement, the lender could notbe entirely sure that the husband was liable.

Under formal separation, the legal rightsand liabilities are much clearer. In Connecticut,legal separation carries all the conditions of adivorce except that the parties are not yet freeto marry again. Neither party is liable for thedebts incurred by the other, and a wife's rightto alimony, support of children, and financialallowances are specified by the court. For thisreason, one mortgage company requires proofof legal separation as a condition to consideringseparated women for a mortgage loan.

Other mortgage lenders simply will not dealwith separated women at all, whether theseparation is formal or informal. The vicepresident of one savings and loan associationflatly stated that, because of ambiguous lia-bility, separated women are not eligible formortgage loans from his institution.

Despite the fact that FHA policy generallyweighs heavily against separated women, in arecent case in Hartford, the local FHA officedid approve a separated woman for a loanunder the section 235 program.50 Initially, herapplication was not processed because she wasonly separated, not divorced, from her husband.FHA counsel, however, reversed the decisionbecause in his view there was no question ofambiguous liability, arguing that State law

permits a wife to own property with clear titlein her name even when informally separatedfrom her husband.

Aside from the problems associated withtheir legal status, another reason why sepa-rated women have difficulty obtaining mort-gages is that their status allegedly reflectsdomestic strife. For example, the FederalHousing Administration, traditionally skepticalof discordant marital relationships, states inits underwriting manual:

It has been demonstrated that inharmoniousdomestic relationships are an importantcause for foreclosure. The determination asto this risk will be dependent upon recogni-tion of items in the credit report and personalhistory of the mortgagor which give evidenceof family discord, pending divorce suits,reconciliation after initiation of divorcesuits, and other items which point to unstablefamily conditions.51

This policy underscores the stigma imposed on"domestic strife" and is especially unfair towomen who continue a long and stable informalseparation. It is contradicted by a recent studythat found no statistically significant relation-ship between marital status and loan delin-quency or foreclosure.52

THE DIVORCED WOMAN

The divorced woman also has considerabledifficulty in obtaining a mortgage, both becauseof the alleged probability of an unstableeconomic situation and because of her socialposition. Her financial circumstances are oftencomplex: she may rely on alimony, childsupport, or an independent income alone, orany combination of these sources.

The divorced woman who has a substantialwork history and a separate source of incomewill be treated as any other single woman.Differential treatment occurs when alimony orchild support payments are listed as contribut-ing and fundamental sources of income.

Lenders generally will not rely on supportpayments which are not court ordered even if

so Title I, Section 235, of the Housing and UrbanDevelopment Act of 1968 provides subsidies in theform of interest reduction payments to mortgagelenders on behalf of lower-income families to enablethem to purchase their own homes

51 Mortgage Credit Analysis Handbook, sec. 2-7.52 John P. Herzog and James S. Earley, Home

Mortgage Delinquency and Foreclosure (New York:National Bureau of Economic Research, 1970).

28

there is a long history of reliable payments.63

One lender stated, however, that the patternof payments would determine their inclusionin total income. The divorcee whose support iscourt ordered, and her sole source of income,has a fair chance of negotiating a loan if shecan make a reasonable downpayment or obtainprivate mortgage insurance, and if her totalsupport can sustain the monthly payments.

Credit ratings are a significant factor inmortgage applications, and a divorced womanfinds establishing credit particularly difficultif she does not have her own credit ratingalready. A divorcee shares the taint of anyadverse information in her former husband'srating, even if she paid their bills promptly.Thus, any negligence on his part would impairher effort to secure independent credit.

Even when income and credit ratings aresound, divorced women still may be turneddown arbitrarily for mortgages as in the fol-lowing case. A 51-year-old divorcee with nodependents, working as a supervisor at aninsurance company in Hartford, sought to pur-chase a three^bedroom, two-bath house inNovember 1971. The asking price was $21,500;the sales price was $20,000. At the time, herannual income was $8,600, and she had workedat the insurance company for 15 years. Hercredit standing was excellent and she main-tained two department store charge accounts.She was willing to put $5,000 down, thusapplying for a $15,000 mortgage.

She tried to apply for a mortgage at themain offices of four lending institutions. Atone savings and loan she was told not to fillout an application because she was "notqualified" and the loan committee would notapprove it. No explanation was offered as towhy she was not qualified. At two others shewas told that she did not "fit their formula,"i.e., 30 percent of income for housing expenses.Having been turned down by three institutions,

she went to a fourth, a savings and loan, andobtained a 25-year, T1/^ percent mortgage.

63 This criterion has no basis in fact because support,whether or not court ordered, is infrequently paid. Forinstance, the Citizens Advisory Council on the Statusof Women has reported that in Wisconsin in 1972 only13 percent of ex-spouses were in full compliance 10years after the settlement; 70 percent had discontinuedsupport altogether.

Chapter 5Employment in Hartford Lending Institutions

The policies and practices of Hartford lend-ing institutions discussed earlier operate witha discriminatory effect on minorities andwomen seeking mortgage loans. Many of thesepolicies and practices are not overtly or con-sciously discriminatory but are viewed by thelending community as necessary elements ofprudent banking. In fact, they can be traced tolong-standing bank tradition and are based oncertain assumptions concerning the subordinaterole of racial minorities and women in society.

Banking traditionally has been a professiondominated by white males. Those in positions tomake bank policy have been white males, andthe policies they have established have beengeared to facilitate credit for white males.Racial minorities and women rarely have beenin positions to change or even influence thesepolicies.

Title VII of the Civil Rights Act of 196454

prohibits discrimination in employment, andall lenders in Hartford are covered by thisstatutory prohibition. The Commission exam-ined the employment records of four Hartfordmortgage lending institutions to determineminority and female representation generallyand their representation in decisionmakingpositions.55 (See tables 9 and 10.) The basicfindings of this examination are:

• Women are abundantly represented asemployees of these institutions but almostentirely in low-level positions which carry no

54 42 U.S.C. 2000 e-2.55 Data for one institution cover only the main

branch for 1971 because there were no comparable datafor 1969. The inclusion of this institution does notdistort the interpretation because it shows the samepatterns as the other institutions. For convenience, the1971 data are treated as though they were for 1969.

authority to influence the policies of theinstitutions.

• Applying the discriminatory occupationalcriteria which these lenders use in weighinga woman's mortgage application, these womenwould not qualify for a loan from the veryinstitutions which employ them.

o Minorities—men and women—are grosslyunderrepresented in any positions with theselending institutions but especially in decision-making positions.56

• Of equal importance, the situation has notappreciably improved over the 3-year periodbetween 1969 and 1972.

During this period the number of womenemployed at these institutions increased by 198,all in office, clerical, and service jobs. Thenumber of female officials, managers, andprofessionals declined by one, although thetotal number of such positions increased by 29.The number of minority women working asofficials, managers, and professionals increasedfrom one to five. The number of minority menin such positions declined, from five to four.

The number of office, clerical, and serviceemployees increased by more than 260 positionsfrom 1969 to 1972, but the great majority ofnew employees (75 percent) were white women.The institutions employed an additional 64white men in this category but added only 3minority females and 8 minority males.

In 1969, women constituted 58.4 percent of

56 The May 1972 HUD Private Lending InstitutionsQuestionnaire showed that 10 percent of the employeesin 50 cities were minority employee?, with 4 percent ofthe decisionmaking positions being filled by minoritiesand 12 percent of the cashiers and tellers beingminority members.

81

the employees and by 1972 this figure hadincreased to 59.7 percent.57 But, whereas in1969, 6 percent of the female employees wereofficials, managers, or professionals, by 1972this figure had declined to less than 5 percent.68

The comparative statistics between 1969 and

57 The percentage of female employees who wereminority women remained constant at 2 percent bothyears.

BS Minority women increased their percentage in this

1972 of minority and female employment byHartford lenders do not suggest that substantialprogress is being made. Until minorities andwomen are more adequately represented inpositions to determine or influence lendingpolicy, it is unlikely that the policies and prac-tices that serve to discriminate against thesetwo groups of homeseekers will be changed.

group from just under 2 percent in 1969 to 9 percentin 1972.

TABLE 9—OFFICIALS, MANAGERS, AND PROFESSIONALS, BY RACE AND SEX,FOUR HARTFORD MORTGAGE LENDERS

Total numberWomen

Minority womenBlackSpanish speakingAsian American _

MenMinority men

BlackSpanish speakingAsian American _

1969353

561100

2975500

1972382

555311

3274400

1969-72Change

29—1

4211

30—1—1

00

Source: Commission staff review of employment records of four Hartford mortgage institutions, April 1973.

TABLE 10—OFFICE, CLERICAL, AND SERVICE WORKERS, BY RACE AND SEX,FOUR HARTFORD MORTGAGE LENDERS

Total numberWomen

Minority womenBlackSpanish speakingAsian American _

MenMinority men

BlackSpanish speakingAsian American _

19691,250

8781917

20

372514452

1969-72Change263199

3111

648062

Source: Commission staff review of employment records of four Hartford mortgage institutions, April

32

Chapter 6Summary and Conclusions

This report has examined mortgage lendingpolicies and practices in Hartford, Connecticut,that bear on homeownership opportunities forminorities and women. The Commission'spurpose has not been to uncover individualinstances of discrimination but to answer thisquestion: If left to operate in accordance withtraditional banking processes and standards,will the system of mortgage finance in the cityassure fair treatment for minorities andwomen ? The basic finding of this report is thatit will not. For minorities and women, themortgage finance system is a stacked deck—stacked sometimes inadvertently, oftenunthinkingly, but stacked nonetheless.

As the Commission and other agencies andorganizations have documented, mortgagelending traditionally has been a closed com-munity, operated largely by white male deci-sionmakers, and its standards have been gearedto facilitate service to white male customers.Minorities and women have had great difficultyin joining this community, either as decision-makers or customers.

The principal problem for minorities in thepast was overt discrimination. Regardless oftheir personal or financial worth, minorities,simply because of race or national origin, wereconsidered less desirable risks than whites.Moreover, it was deemed virtually unthinkableto provide them mortgages for houses in non-minority areas. In the past, these discrimina-tory policies were openly admitted by repre-sentatives of the mortgage lending communityand stoutly defended as essential elements ofprudent banking, even by agencies of theFederal Government charged with responsi-bility for supervising and regulating mortgagelenders.

Since the enactment of Title VIII of the CivilRights Act of 1968, mortgage lending dis-crimination on the basis of race or ethnicorigin has been unlawful. Public pronounce-ments advocating such discrimination are nowrarely heard. It is unlikely, however, that thetraditional banking attitudes and perceptionsabout minorities, which developed and hard-ened over decades, have changed substantiallyin the 6 short years since Title VIII wasenacted. The fact that mortgage lenders nolonger openly avow discriminatory practicesdoes not mean that they no longer engage insuch practices. Rather, it suggests that theyhave gone underground.

In examining the mortgage lending systemas it relates to minorities, the Commission'sprincipal concern was to determine the extentto which safeguards were provided to assureagainst decisions based on racial or ethnicdiscrimination rather than on objective creditcriteria. It also sought to determine whetherthe system afforded sufficient opportunity forminorities subjected to discrimination to havetheir grievances redressed. The Commissionfound that the mortgage finance system inHartford not only lacks sufficient safeguards toassure against discriminatory decisions butfacilitates discriminatory rejection of minori-ties. The Commission also found that thesystem affords little opportunity for suchdecisions to be reversed.

The process of obtaining a mortgage isarduous for everyone. But running the three-stage gauntlet required to obtain a loan is evenmore difficult for minorities. Before reachingthe final stage of formal submission of a mort-gage loan application to the loan committee,

33

minorities must first clear two formidableobstacles: the real estate broker and the loanofficer. Each of these first two stages carriesbroad opportunities for discriminatory orother irrational rejection.

Brokers may reject minority applicants onthe basis of their own bias. Further, becausebrokers seek to stay in the good graces oflending institutions, they prefer not to referquestionable applicants who may not qualify or,if approved, may default on the mortgage pay-ment. Thus, brokers may reject minorityapplicants because they think, rightly orwrongly, that the lending institution frowns onminority applications, or because they misun-derstand the criteria used by the lendinginstitution. Moreover, this stage of the trans-action is entirely .informal. Rejected minorityapplicants usually are merely told that they donot qualify for a mortgage loan and the matterends there.

Even if minority applicants clear the hurdleof the broker, they are still one difficult stepaway from a formal application. The secondhurdle is the loan officer, who also handles thetransaction on an informal basis. The loanofficer's decision to recommend or discouragesubmission of a final mortgage loan applicationis based upon a number of factors, includingsome that are highly subjective, such as moti-vation, character, and the quality of theapplicant's domestic life. Furthermore, loanofficers, whose opportunities for career ad-vancement depend in part on the default rateon applications they refer for approval to theloan committee, prefer to err on the side ofextreme caution. Indeed, they are reluctant toaccept applications which they feel may berejected by the loan committee, since thisreflects adversely on their judgment.

Even criteria that appear objective on thesurface often turn out on closer examination tohave subjective factors that discriminateagainst minorities. Criteria such as the maxi-mum permissible ratio of monthly payment tomonthly income, the definition of income,credit standards, and appraisals are appliedinconsistently by Hartford lenders and have theeffect of screening out qualified minorityapplicants.

Thus, the system of mortgage finance, repre-sented by the screening process and the sub-jectivity of many of the criteria on whichqualifications are measured, affords ampleopportunity for discriminatory rejection ofminorities. Even though discrimination inmortgage lending may have gone largelyunderground since enactment of Title VIII,the Commission's assessment of the policiesand practices of Hartford lenders offers littleassurance that such discrimination has ceasedto exist.

While discrimination against minorities 'byHartford lenders is subtle, discriminationagainst women is blatant. Minority womensuffer the double effect of both sex and racediscrimination.

Women as a class, unlike minorities, areunprotected against discrimination in mortgagelending under Federal law. Connecticut and anumber of other States now prohibit sexdiscrimination in credit transactions, butwomen, by their very status as women, never-theless are openly considered questionablerisks by mortgage lenders. Their treatmentvaries, depending upon whether they aremarried, unmarried, widowed, separated, ordivorced, but none receives treatment equal tothat received by men.

The income of married women, a substantialpercentage of whom are gainfully employed, isconsidered "secondary" and rarely creditedmore than 50 percent, if at all, in determininga family's financial status. This policy worksconsiderable hardship on all families for whomthe wife's income is essential. However, itcompounds the problems of minority families,a substantial percentage of whom rely on thewife's earnings to meet their needs. The policyof discounting- some or all of a working wife'sincome has a doubly discriminating effect onthese families, as they are thus penalizedbecause of both race and sex.

The policies of mortgage lending institutionson the discounting of a working wife's incomevary widely from institution to institution andfrequently within the same institution. Somegeneral standards do exist, however. The mostfavored working wife is beyond child-bearingyears, holds a job considered "professional,"

34

and demonstrates a consistent pattern of con-tinued employment. She has a good chance ofhaving all or most of her income credited bythe mortgage lending institution.

The least favored working wife is the youngwoman of child-bearing years, who may haveyoung children, and who holds .a job considered"nonprofessional." This class of working wivesis viewed skeptically by the mortgage lendingcommunity as "unstable." Certain assumptionsare made about her regardless of her real situa-tion. One is that, if she has no children, she islikely to and will leave her job, at least for atime, thereby reducing the family income.Another is that, if she already has youngchildren, she is even more likely to have addi-tional children. Again, she allegedly will leaveher job, at least for a time. The third assump-tion is that her "nonprofessional" job is un-stable and that somehow makes her unstable.These assumptions add up to a compositepicture of a person in whom lending institu-tions place little reliance. Therefore, her incomeis automatically discounted, either substan-tially or entirely. Moreover, whether the familyis approved or disapproved, they frequentlynever learn if, or to what extent, the wife'sincome was credited.

The single woman has great difficulty ingaining access to mortgage finance. The factthat she is without the protection of .a malemakes her suspect to the lending community asa credit risk. The image of women as "weakervessels" also makes lending institutions reluc-tant to approve mortgage loans, particularly onmultifamily dwellings, on grounds that womenare unable to perform the necessary mainte-nance.

The young, unmarried woman has the mosttrouble securing a mortgage, principally be-cause of the assumption made by lenders thatshe will marry, have children, and stopworking, thereby reducing her economic status.

The widow generally is in a more advanta-geous position, since she often can rely on lifeinsurance proceeds, social security payments,or the proceeds of her husband's estate toprovide a healthy downpayment and assure asteady income.

The separated woman is in an awkward posi-tion in seeking to obtain access to mortgagefinance. One reason for this is her uncertainlegal status for debt liability, particularly ifthe separation is informal. Another reason isthat her status is assumed to reflect domesticstrife, an added suspect factor, according tolenders.

The divorced woman also has considerabledifficulty obtaining a mortgage, both becauseof the alleged probability of an unstableeconomic situation and because of the socialstigma attached to divorce. If her support iscourt ordered, her chances for obtaining amortgage are enhanced. If not court ordered,they are diminished. Furthermore, unless shehas established independent credit, any adverseinformation in her former husband's creditrating reflects on her and further diminishesher chances to obtain a mortgage.

The Commission also inquired into employ-ment of minorities and women by Hartfordmortgage lenders, particularly in decision-making positions. It found that, while womenconstitute a majority of the employees at theinstitutions surveyed, few occupy positions inwhich they can determine or influence lenderpolicy; the vast majority, in fact, are in cleri-cal and service occupations. Blacks and PuertoRicans—both male and female— are scarce inevery level of mortgage finance institutions.

The system of mortgage finance in Hartford,under which minorities and women are in-equitably treated, reflects a reluctance by thelending community to alter traditional policiesand standards, even though many are unrealis-tic and others facilitate illegal acts. Discrimi-nation on the basis of race or ethnic origin isprohibited under Federal and State law, butthe mortgage lending community adheres to asystem under which such discrimination per-sists, unabated and undetected. Sex discrimi-nation in mortgage finance, although not yetprohibited by Federal law, is prohibited bylaws recently passed in several States, includingConnecticut, and is totally at odds with thereality of modem-day America in whichwomen are an increasing and substantial partof the labor force. Yet mortgage lendinginstitutions cling to images of women as un-

35

stable, unreliable, and in need of maleprotection.

Reform is clearly needed. One alternative isfor mortgage lending institutions, themselves,to reexamine their policies and institute thechanges necessary to assure that minoritiesand women are treated equitably. The Commis-sion, however, recognizes the severe limitationsof relying on voluntary reform. The agenciesthat supervise and regulate most of the mort-gage lending institutions in Hartford andthroughout the country should, therefore, re-quire that the needed reforms be instituted.

On several occasions in recent years, theCommission, other Federal agencies, andprivate groups have sought to point the direc-tion for such reform. These recommendationsare still valid. Lending institutions should berequired to maintain records of the race orethnic background and sex of mortgageapplicants and persons who make oral inquiriesabout home loans. Such data is urgentlyneeded as a basis for uncovering and correctingpatterns of discrimination. Otherwise, theburden of redress rests entirely on the com-plaint process, a most inadequate means ofremedying discrimination. The regulatoryagencies, especially the Federal DepositInsurance Corporation and the Federal HomeLoan Bank Board which have considered sucha requirement, should move with speed torequire member institutions to initiate aneffective data collection system.59

Another way to generate reform is to requirelenders to provide homeseekers with writtenstatements of the criteria used in determiningtheir eligibility for a mortgage. Lenders alsoshould be required to notify in writing rejectedapplicants and persons advised against makingapplication, informing them of the reasonsfor their rejection. These steps would dis-

59 The Federal Reserve System, the Federal DepositInsurance Corporation, the Comptroller of the Cur-rency, and the Federal Home Loan Bank Board recentlyhave implemented an experimental program to collectracial and ethnic data over a 6-month period in 18cities during 1974. The objective is to develop a record-keeping requirement for racial data on the dispositionof mortgage loan applications. The information requiredunder the experiment includes the applicants' age, sex,marital status, and financial situation.

courage loan officers from arbitrarily dis-suading or rejecting applicants, as such actionswould be matters of public record.

Examiners from the financial regulatoryagencies should be charged with the responsi-bility, of detecting discrimination by lenders. Tothis end audits and examinations of lenderpractices must be expanded. In particular, theFederal Home Loan Bank Board shouldestablish a monitoring system to assure thatits newly established guidelines are beingfollowed.

Title VIII of the Civil Rights Act of 1968(the Federal Fair Housing Law) must beamended to prohibit discrimination based onsex. Single women and working wives couldthen more easily fight sex discrimination inthe courts.

Unless these measures are instituted, equalhousing opportunity will remain for mostminorities and women a slogan withoutsubstance.

36

U. S. COMMISSION ON CIVIL RIGHTSWASHINGTON, D. C. 20425

POSTAGE AND FEES PAIDU. S. COMMISSION ON CIVIL RIGHTS

OFFICIAL BUSINESSPENALTY FOR PRIVATE USE $300