Full text of Report on the Status of the Community Reinvestment Act : Views and Recommendations : Report on the Status of the Community Reinvestment Act : Views and Recommendations, Volume II of II

View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

102d Congress

S. PRT.

COMMITTEE PRINT
102–121 Vol . II

R E POR T
ON THE

STATUS OF THE
COMMUNITY REINVESTMENT ACT

VIEWS AND RECOMMENDATIONS

VOLUME II OF II

NOVEMBER 1992

Printed for the use of the Committee on Banking, Housing, and Urban

U.S. GOVERNMENT PRINTING OFFICE
60-893

For sale by the U.S. Government Printing Office

Superintendent of Documents, Congressional Sales Office, Washington, DC 20402

COMMITTEE ON BANKING, HOUSING, AND URBAN AFFAIRS

SUBCOMMITTEE ON HOUSING AND URBAN AFFAIRS
ALAN CRANSTON , California , Chairman
JIM SASSER , Tennessee

( II )

DONALD W.RIEGLE , JR . , MICHIGAN , CHAIRMAN

ALAN CRANSTON , CALIFORNIA

United States Senate
COMMITTEE ON BANKING , HOUSING, AND
STEVEN 8. HARRIS , STAFF DIRECTOR AND CHIEF COUNSEL
LAMAR SMITH, REPUBLICAN STAFF DIRECTOR AND ECONOMIST

November 20, 1992

Hon . Donald Riegle
Chairman

Committee on Banking, Housing and Urban Affairs
U.S. Senate

Washington , D.C. 20510
Dear Mr. Chairman :

I submit this report on the status of the Community Reinvestment Act
( CRA ) . I ask that the report ( which includes a hearing, a Subcommittee

analysis, and a coinpilation of studies) be printed as apublic document for use
by the Senate and distribution to those interested in matters pertaining to
community reinvestment.

As an original sponsor of CRA, I have followed with care its

implementation and enforcement. I consider CRA a critical component of the
body of laws enacted to combat discrimination and redlining -- laws that I
worked to enact and strengthen throughout my Senate career .
Simply passing laws, however , is not enough. They must be enforced

and it is one of the principal responsibilities of Congress to ensure that these
laws are enforced . Thus, the Housing Subcommittee held an oversight hearing
on CRA on September 15, 1992. The Subcommittee solicited testimony from
government officials, banking industry representatives and members of the non

profit community. In addition, I instructed the statt to conduct an
investigation into CRA's implementation and enforcement. This report contains
the findings and recommendations of these oversight activities.

The riots in Los Angeles in April of this year caused the nation to focus
its attention the rage andfrustration of a specific community. South Central
Los Angeles servedto remind us that the needs of our urban communities
nationwide have been neglected for too long -- the needs of our youth who
cannot find jobs, the deep poverty, the lack of affordable housing, the lack of
opportunity , the lack of hope.
( III)

One particular problem that emerged clearly from the riots was the fact

that South Central Los Angeles, like many communities in this country, suffers
from a lack of access to capital. It is a disinvested community that has been ,

for the most part abandoned by traditional lenders. How could this have
happened when CRA has been effect for fifteen years? What role should CRA
have in communities like South Central Los Angeles ? The Subcommittee's
review of CRA sought to answer these and other questions that pertain to the
implementation and enforcement of CRA .

The Subcommittee concluded that the agencies charged with CRA's
enforcement -- the Federal Reserve, the Officeof Thrift Supervision, the Office

of the Comptroller, and the Federal Deposit Insurance Corporation -- and the
financial institutions which they regulate have yet to make a full commitment

to CRA's implementation . The regulators' record of inconsistent and lax
enforcementhas encouraged indifference and disinterest by the financial
institutions. As a consequence, the agencies bear significant responsibility for
the poor performance of many of the financial institutions.
The first volume contains the hearing record; the second volume contains
the results of the Subcommittee's investigation; and the third contains a
compilation of studies to be used as a resource base for Congress and other
interested parties.

I want to thank the many people who contributed to this effort .
Sincerely

herita

Alan Cranston
Chairman

DONALD W RIEGLE , JR ., MICHIGAN , CHAIRMAN

ALAN CRANSTON , CALIFORNIA

United States Senate
COMMITTEE ON BANKING, HOUSING, AND
STEVEN B HARRIS , STAFF DIRECTOR AND CHIEF COUNSEL

LAMAR SMITH, REPUBLICAN STAFF DIRECTOR AND ECONOMIST

Dear Friend:

In the course of the Housing Subcommittee's investigation on the current status
of the Community Reinvestment Act, staff reviewed numerous studies, papers and
other material on lending discrimination. Although it is not possible to reprint all the
material, the Subcommittee has selected a few representativestudies that combined
with the accompanying report and with the hearing record may provide a starting
point for future discussion.
Sincerely

Carton

Alan Cranston
Chairman

)
( V)

CONTENTS
COMPILATION OF STUDIES

VOLUME II OF II

Page
Transmittal letter

159

424

Las Vegas Alliance for Fair Banking
CIO CULINARY WORKERS UNION LOCAL 226
1630 S. COMMERCE STREET . LAS VEGAS, NV 89102 • ( 702) 386-5123

Cashing Out
A Report on Home Mortgage Lending to Minorities and
Low- and Moderate- Income Neighborhoods in Las Vegas

June 1992

( 1)

SUMMARY

Scope of the report
There were approximately 60 banks, savings and loans, credit unions, mortgage
companies, and other lenders active in the Las Vegas area in 1990. In the
aggregate, they made 10,658 loans on one - to four-family residences, with a value
of $ 922.2 million. These loans consisted of home purchase mortgages,

refinances, and home improvement loans.
Of the 60 active lenders, we found that just six ( or 10 percent) made more than
half of all loans. We examined the lending records of all lenders in the
aggregate, as well as the top four deposit -taking institutions on this list: Citibank,

Valley Bank and Valley Mortgage Company ( hereafter " Valley " ) , First Interstate
Bank of Nevada, and Primerit Bank .

Findings
1. Home loans to the Westside .

• Of 10,658 home loans made by all lenders in Las Vegas, only 59 went
•
•

Just three -tenths of one percent of the nearly $ 1 billion in home loans
Of the four leading lenders, First Interstate made only $ 59,000 in loans

2. Deposits in branches close to or serving the Westside, compared with dollars
•

Evidence suggests that bank deposits originating in the Westside and

•

Valley Bank's two branches close to the Westside had $ 96 million in

First Interstate's two branches close to the Westside had $ 94 million in

3. Denials of loan applications from African -American and white applicants.

•

Las Vegas lenders rejected mortgage loan applications from blacks 1.5

A black applicant for a home loan with an income above $ 41,500 was

Among leading banks, First Interstate rejected middle income black

4. Home purchase loans to African -American borrowers throughout the Las

•

African - Americans make up about 9 percent of Clark County's

Valley and Primerit each made about 3 percent of their home purchase

5. The leading banks' market share of home purchase loans to African -American

•

First Interstate's market share of loans to blacks is half its market share

Valley and Citibank have a slightly greater market share of loans to

6. Home purchase loans to all neighborhoods with predominantly minority
•

Of the 8,581 home purchase loans made by all lenders in Las Vegas,

•

Eleven predominantly white census tracts ( with less than 20 percent

7. Home purchase loans in low- and moderate- income neighborhoods.

10 •

•

Only 6.8 percent of the 8,581 home purchase loans in 1990 went to

•

Primerit and Citibank made 10.6 percent and 8.5 percent, respectively,

•

First Interstate made only 2.7 percent of its 563 home purchase loans in
Upper-income neighborhoods received more than 57 percent of the

8. The leading banks' market share of home purchase loans in low- and
•

•

Valley came close to having equal market shares, at about 8 percent, in
Primerit had a 50 percent greater share of the low -and moderate-income
First Interstate's share of the upper income market, at about 10 percent,

Recommendations:
•

•

The appropriate regulatory agencies should investigate every Nevada

The Federal Reserve should not approve any application subject to CRA

Regulatory agencies must take lending patterns in low- and moderate

State and local governments should develop legislation, after suitable

LAS VEGAS ALLIANCE FOR FAIR BANKING

Cashing Out
A Report on Home Mortgage Lending to Minorities and Low- and
Moderate - Income Neighborhoods in Las Vegas

" A few miles and a world away from the blazing neon and flashing
billboards where tourists stroll, this city's black neighborhoods have
been wracked by mob violence almost every night since the riots began
in Los Angeles."

" Local critics have warned gaming moguls that they need to worry less
about image and more about the deteriorating, long -ignored west Las
Vegas ghetto ...."
" By midnight, police had cordoned off the area around the Golden

West and approximately 80 policemen remained in the area . Black
community leaders had warned city officials last summer that trouble
was brewing ...."

A. Introduction

Las Vegas's reputation through the 1980's as the fastest growing city in the nation
was not inaccurate . It was simply not the whole story. Some parts grew faster
than others. Some parts grew very little if at all. By the end of the decade, in

1990, whole communities existed where a few years before there was open land.
The Census Bureau carved 31 new tracts into Clark County in 1990, home to a
largely white and largely affluent population , 50 percent greater in numbers than

it had been ten years earlier. The minority population, almost doubling in the
same period, lived elsewhere: by 1990 , nearly 40 percent in just 15 older, more
centrally situated census tracts.
This report examines some of the ways in which Las Vegas's banks and other
mortgage lenders have set the stage for Las Vegas's passage through the 1990's.

B. Scope of the report

Following the Federal Reserve's release of the 1990 Home Mortgage Disclosure
Act reports in October, 1991 , the Las Vegas Alliance for Fair Banking reviewed
the aggregate data for all financial institutions that made home loans in Las Vegas
in 1990.

There were approximately 60 banks, savings and loans, credit unions, mortgage
companies, and other lenders active in the area in 1990. In the aggregate , they

made 10,658 loans on one- to four-family residences, with a value of $ 922.2
million . These loans consisted of home purchase mortgages, refinances, and
home improvement loans. ( See Table 1 , page 20 of this report.)
Of the 60 active lenders, we found that just six ( or 10 percent) made more than
half of all loans. These top lenders are listed in Table 2 , page 20 .

We examined the lending records of the top four deposit -taking institutions on this
list: Citibank, Valley Bank and Valley Mortgage Company ( hereafter " Valley " ) ,
First Interstate Bank of Nevada, and Primerit Bank. These institutions are subject

to the 1977 Community Reinvestment Act ( CRA ) , as amended , which imposes a
responsibility on banks to meet the credit needs of all the communities they serve,
including low- and moderate - income communities.

We looked at the following patterns, utilizing both aggregate data and data on
individual leading lenders:
1. Home loans to the Westside.

2. Deposits in branches close to or serving the Westside, compared with
3. Denials of loan applications from African - American and white applicants.
4. Home purchase loans to African -American borrowers throughout the Las
5. The leading banks ' market share of home purchase loans to African
6. Home purchase loans in neighborhoods with predominantly minority

8
7. Home purchase loans in low- and moderate - income neighborhoods.

8. The leading banks' market share of homepurchase loans in low- and

C. Findings
1. Home loans to the Westside.

•
•

•

Of 10,658 home loans made by all lenders in Las Vegas, only 59
Just three -tenths of one percent of the nearly $ 1 billion in home

Of the four leading lenders, First Interstate made only $ 59,000 in

The Westside is defined by five census tracts : 2.01 , 3.01 , 3.02, 35 and 37- ( see

Figure 1 , page 24, and Table 3 , page 21 ) . ' In 1980, the Westside's population
was 18,519. By 1990, the population had decreased to 16,204, of whom 13,796
( 85 % ) were African -Americans.

A description of the role of the Westside in the social and economic history of
Las Vegas is beyond the scope of this report. It is worth noting, however, that

the Westside's exclusion from the flow of credit and capital in the city extends to
the 1930's and beyond . The concentration of African -Americans there began in

the 1940's, determined by such measures as covenants restricting the sale of
housing and land in Las Vegas to whites only , the refusal of landlords to rent to

blacks, and conditions on the renewal of business licenses for blacks requiring
them to move to the Westside.

A housing survey of the Westside in 1949 reported that 80 percent of the
community's structures were substandard . Streets were unpaved well into the
1950's. By the early 1960's, as historian Eugene P. Moehring has observed, the
redlining of the Westside had become a contentious political issue, and " it was

well known that ( First National Bank ] had granted few home loans or mortgages
to the zone . "

The practice continues, as do some of the players, although generally under new
names. As Table 3 and Figure 1 show, the Westside received only 59 of the
10,658 purchase, refinance and home improvement loan made by Las Vegas's
sixty mortgage lenders in 1990 — less than one loan per lender.

Of the $ 922 million of mortgage finance allocated to Las Vegas in 1990 by these
multimillion- and multibillion -dollar lenders , just three -tenths of one percent found
its way to the Westside.

Among the four leading home mortgage lenders, the aggregate pattern was
replicated , with some variations. First Interstate, Primerit, and Citibank made
three loans each to the Westside; Valley made 12 loans ( 1.3 percent of its total
loans) .

In monetary terms, First Interstate loaned less than one-tenth of one percent of its
mortgage funds to the Westside: $ 59,000 out of more than $ 62 million in all of

Las Vegas. Primerit and Citibank each loaned $ 114,000 , out of $ 62 million and

$ 73 million , respectively. Valley loaned $ 564,000, or about six -tenths of one
percent of all its mortgage lending .
These patterns are neither accidental nor the result of some temporary economic
fluctuation . As other data discussed below demonstrate , the flow of credit and
capital for home loans follows deliberate courses . Financial institutions make and

pursue decisions to penetrate existing as well as emerging markets . Products are
developed and promoted , staff is encouraged to achieve defined goals, deposits
are directed into carefully considered investments, and so on .

For lenders in Las Vegas as a whole , and for the leading lenders in particular, the
exclusion of the Westside has been as deliberate as the inclusion of other parts of

Las Vegas, whether old or new.

2. Deposits in branches close to or serving the Westside, compared with

•

Evidence suggests that bank deposits originating in the Westside

Valley Bank's two branches close to the Westside had $ 96 million in

•

First Interstate's two branches close to the Westside had $ 94 million

In 1990 there were no bank branches in the Westside. There were, however, a
number of branches of the major lenders near the Westside, and it is instructive to
measure the deposits in these branches in relation to the value of home loans
made in the Westside .

First Interstate Bank, Valley Bank, and Primerit Bank each have one or more
branches that serve or are near the Westside. Table 4 ( page 21) lists these

branches and their deposits in 1990, as well as the value of home purchase loans
made by the banks in the nearby communities.

The rate of home mortgage lending to communities that may be presumed to
make considerable use of these branches for personal as well as commercial

deposits is quite striking.
Valley Bank's two branches close to the Westside, for example, had a total of $ 96
million in deposits in 1990 ; in the same year, Valley made 12 mortgage loans to
the Westside, worth $ 564,000.
First Interstate's two branches close to the Westside had a total of $ 94 million in

deposits in 1990 ; in the same year, First Interstate put just $ 59,000 back into the
Westside by way of three home loans.

Primerit's branch close to the Westside had about $ 15 million in deposits in 1990 ;
in the same year, Primerit made 3 home loans to the Westside, worth $ 114,000 .
For a standard of comparison, we looked at the same relationship in a wealthier,

predominantly white community about five miles to the west of the Westside ( see
Table 4 , page 21 ) .
This community , bounded on the north and south by Vegas Drive and Charleston
Boulevard, and on the east and west by Rainbow Boulevard and Buffalo Drive

( census tract 30.02) , had a population in 1990 of some 11,000, of whom about 85
percent were white . Las Vegas's 60 lenders made 152 home loans, worth nearly
$ 11.2 million, to tract 30.02 in 1990. Recall that the five tracts of the Westside,
combined , received only 59 loans, worth $ 3.1 million , from these same lenders .

First Interstate, Valley, and Primerit each have a branch near the Rainbow
Boulevard /Westcliff Drive junction , on the eastern boundary of tract 30.02.
5

First Interstate's Rainbow -Westcliff branch held nearly $ 58 million in deposits in
1990, about the same as the bank's North Las Vegas branch . In contrast to three
loans worth $ 59,000 in the Westside, however, First Interstate made 23 loans,
worth nearly $ 1.2 million , on properties in tract 30.02 .
Valley Bank had approximately $ 32 million in deposits at its Rainbow -Westcliff

branch , about one-third of the combined deposits at its two branches near the
Westside. Valley Bank and Valley Mortgage made 17 loans, worth $ 925,000, in
tract 30.02 .

Primerit's Rainbow branch had some $ 28 million in deposits in 1990, about

double the North Las Vegas branch . Primerit made 13 mortgage loans, worth
$ 1,693,000, to tract 30.02, four times as many loans, and nearly 15 times as
many dollars, as it loaned to the Westside .

In the absence of more precise data on the flow of deposits from specific
communities to specific branches, the relationships examined here can only be

suggestive of a capital movement out of minority and low- and moderate -income
neighborhoods and into predominanty white and affluent neighborhoods. The

point is that minority communities generate substantial deposits in bank branches
in or close to their neighborhoods; there is nothing to indicate, however, that the
dollars are returned to those neighborhoods at a rate at all comparable to what
occurs in predominanty white neighborhoods.
Indeed , a recent 14 -city study by the Association of Community Organizations for

Reform Now ( ACORN ) found that, on average, for every dollar on deposit in
predominantly white neighborhoods, banks and thrifts made eight cents in
mortgages available locally. In minority neighborhoods, the ratio fell to four
cents .

3. Denials of loan applications from African - American and white

•

Las Vegas lenders rejected mortgage loan applications from blacks

A black applicant for a home loan with an income above $ 11,500

•

Among leading banks, First Interstate rejected middle income black

The home mortgage information released by the Federal Reserve in October,

1991, marked the first time data on the disposition of applications by the race and
income of applicants had been made publicly available.
One of the most consistent patterns disclosed by the data was the difference in the
denial rates of applications from minority and white applicants. Across the
nation , in institution after institution, applications from minorities were denied
more often than for whites of similar income.“

Early in 1992, the Office of the Comptroller of the Currency disclosed that it was
investigating more than 250 banks under its supervision whose rejection rates for
minorities were disproportionately high, or which received fewer than 1 percent
of their applications from minorities."

As Figure 2.1 illustrates ( page 25) , among Las Vegas lenders in the aggregate,
loan applications from blacks, regardless of where they may live, were denied 1.5
times as often as applications from whites.
The patter persisted even when applicants came from the same income
categories.' Middle income black applicants, for example, with incomes between
$ 27,665 and $ 41,500, were 1.4 times more likely to be denied a loan than white
applicants in the same income category .

Upper income black applicants ( with incomes above $ 41,500, or 120 percent of
the area median ) were turned down at exactly the same rate ( 23.5 % ) as white
applicants with low- to moderate - incomes ( less than $ 27,665, or 80 percent of the
area median ) .

The leading lenders among banks showed significant and dramatic variations in
their pattems of loan rejections ( see Figures 2.2-2.5 , page 26 ) .

First Interstate denied loan applications from blacks on the whole more than four
times as often as from whites. Even within the same income categories, blacks
were denied loans considerably more often than whites: up to six times more often
in the case of middle income applicants, and nearly three times more often among

upper income applicants. An upper -income black applicant ( with an income of at

least $ 41,500 ) , was twice as likely to be denied a mortgage by First Interstate as
a low -to moderate - income white applicant ( with an income of no more than
$ 27,665 ) .

The pattern at Valley Bank and Valley Mortgage differed considerably from the
aggregate as well as from the other leading lenders in that overall loan denial
rates were very low, and there were no loan denials involving black applicants.
Of the 34 applications Valley received from African -Americans, 32 resulted in

loans and two were withdrawn .' Among white applicants, too, Valley's rejection
rate , at 3.5 percent overall, was well below the other leading lenders.
At Citibank , middle income blacks were turned down nearly twice as often as

middle income whites, although in the upper income category denial rates were
approximately the same. On average, Citibank denied loans to blacks at just
about the same rate as to whites.

At Primerit Bank, low- and moderate-income blacks were denied loans twice as
often as whites in the same income category. For every two upper income white
applicants denied a loan by Primerit, three upper income black applicants were
rejected.

The fact that racial differences in the rate of loan application rejections in Las
Vegas replicate findings in other cities is not surprising. The city's history

partakes of a national history of legal and informal racial segregation in every
area of life, including the pursuit of domestic tranquility. That blacks are more

likely, and often much more likely, than whites to be denied access to credit
mirrors other differences experienced in daily life by the two races, such as the
mortality of their infants, violence at the hands of police, unemployment,
incarceration , and life expectancy .
Despite the African -American community's persistent demands to be included in

the definition of what constitutes a good and credit -worthy market, financial
institutions in Las Vegas, for the most part, conduct business as they do
elsewhere: denying that race plays a role in their product development, marketing,
credit extensions, or hiring, while compiling a record that demonstrates the
opposite

4. Home purchase loans to African -American borrowers throughout the

•

African - Americans make up about 9 percent of Clark County's

Valley and Primerit each made about 3 percent of their home

Clark County's black population has increased by about 50 percent since the 1980
census, and10 is now close to 70,000, or 9 percent of the county's total
population."

It is a population , as well, with substantial assets in real property. The 1990
census estimates the aggregate value of owner-occupied residential property
owned by blacks in Clark County at $ 592.4 million .

Nonetheless, the aggregate record shows that only 308, or less than 3 percent, of
the 10,658 home loans made in 1990 went to African -American borrowers ( see
Table 5, page 22 and Figure 3.1 , page 27) .

Of the dominant bank lenders, only Valley and Primerit exceeded the aggregate
share of loans to blacks, and in neither case by very much: Valley made 3.5
percent of its home loans to blacks, and Primerit made 3.2 percent. Citibank , on

the other hand , made only 2.6 percent of its loans to African -Americans, while
First Interstate , at a mere 1.6 % of all its home loans, was well below the

aggregate pattern ( see Table 5, page 22, and Figures 3.2-3.5 , page 28) .
In monetary terms, Citibank, with $ 2.6 million in loans to blacks ( $ 1.8 million of
which took the form of conventional home purchase mortgages to upper- income
borrowers) was the only major lender to exceed the aggregate average of 2.7 % of
dollars loaned . Primerit matched the aggregate average, while Valley, at 2.4 %
( $ 2.2 million ) was somewhat below it. First Interstate, which loaned only

$ 754,000 to African - American borrowers, or 1.6 percent of its total lending , was
well below the aggregate's already unimpressive benchmark .

5. The leading banks' market share of home purchase loans to African
•

•

First Interstate's market share of loans to blacks is half its market

Valley and Citibank have a slightly greater market share of loans to

Market share is a useful indicator of the outcome, over time, of the priorities

banks establish and pursue. By breaking the HMDA data down into submarkets
by race or income of applicants, and then comparing the shares obtained by
competing lenders, we begin to glimpse the " culture “ prevailing in the individual
lending institutions. Market share, after all, is to a great extent the result of
conscious and deliberate actions on the part of lenders. An institution that more

effectively penetrates the market of white borrowers than black borrowers, for
example, or upper income as against low- and moderate -income borrowers, does
so most likely because it intends to .

While no major bank in Las Vegas originates a significant portion of its loans
among black borrowers, and blacks in any case obtain relatively few of the home

loans originated , there are market share variations among leading banks worthy of
note ( see Figure 4 , page 29 ) .

Valley leads the four top bank lenders with a 10 percent share of the 267 home
purchase loans made to blacks. This is 20 percent above Valley's share of loans
to whites, and suggests, at the least, the inclusion of blacks as a " customary "
component of Valley's corporately -defined marketplace.

Primerit, with a somewhat smaller overall market share than Valley, has about the
same share among both black and white borrowers. Citibank , smaller still than

Primerit, has approximately one percent more of the market among blacks than
whites, about the same proportion as Valley .
First Interstate , on the other hand, has half the market share among blacks as it
does among whites. While First Interstate's market share among white borrowers

is almost as great as Valley's, Valley's share among black borrowers is three
times First Interstate's.

6. Home purchase loans to all neighborhoods with predominantly
•

•

Of the 8,581 home purchase loans made by all lenders in Las

Eleven predominantly white census tracts ( with less than 20 percent

In 1980, there were nine census tracts out of 89 in Clark County with 50 percent
or more minority residents. Their total population was 35,971 , or 7.8 percent of
the county population.

By 1990, the number. of predominantly minority census tracts had increased to
fifteen ( out of 120 tracts in Clark County ) , with a population of 69,003, or 9.3
percent of the total in the county . These tracts also contain approximately 8.2
percent of the county's single family residences ( see Table 6, page 22) .
It is worth noting that no census tract that was more than 50 percent minority in
1980 has dropped out of the list. Among those are the tracts that comprise the
Westside: 2.01 , 3.01 , 3.02, 35 and 37 .
There has been , in other words, a tendency not only for predominantly minority
tracts to remain so over time, but for the proportion of the area's population

living in predominantly minority neighborhoods to increase over time, as well.
Indeed, data from the 1990 census indicate that, while the county's minority
population has now reached 182,000, or almost 25 percent of the total ( see Figure

5) , nearly 40 percent of this population lives in the 15 predominantly minority
tracts listed in Table 6.

The containment and concentration of minorities in certain sections of the city is

one of the consequences of the patterns of bank lending discussed in this Report.
In 1990, as Table 6 shows, Las Vegas lenders made a total of 183 home purchase
loans to the 15 predominantly minority census tracts . These loans had an

aggregate value of $ 9,954,000. The loans to these tracts constituted 2.1 percent
of the total loans made in 1990 , and just 1.2 percent of the total dollars loaned by
all mortgage lenders."

By way of contrast, Table 7 ( page 23) lists eleven 1980 census tracts ( and their
1990 equivalents) that each received more home purchase loans than the 15
predominantly minority tracts combined .

These census tracts, now subdivided into 31 tracts, represent areas of Las Vegas
that expanded rapidly in the 1980's. They include communities to the east, south
and west of the city. The minority population in most of these tracts is between
10 percent and 15 percent on average .
Individually, in 1990, these tracts had about four times the population of the 15

predominantly minority tracts , but received nearly 30 times more home purchase
loans ( 5364 to 183) , and 56 times the number of home purchase dollars

( $ 557,194,000 to $ 9,954,000 ) .

7. Home purchase loans in low- and moderate- income neighborhoods.
•
•

Only 6.8 percent of the 8,581 home purchase loans in 1990 went to

•
•

Primerit and Citibank made 10.6 percent and 8.5 percent,

First Interstate made only 2.7 percent of its 563 home purchase
Upper -income neighborhoods received more than 57 percent of the

Financial institutions are required under the CRA to serve " the convenience and
needs of the communities in which they are chartered to do business .“ Meeting
the credit needs of low- and moderate - income neighborhoods in these communities

is one of the criteria on which financial institutions are evaluated when applying
for deposit insurance, a branch or deposit facility, a merger or acquisition, and
other regulated activities. Low- and moderate - income neighborhoods, for HMDA
reporting purposes, are those census tracts with a median family income below 80
percent of the area median .

In 1980, when the Las Vegas area median family income was $ 21,056 , there
were 27 tracts that met the low- and moderate -income definition . At that time,

these tracts had a population of 117,934 , or about 25 percent of the total
population of the county . For 1990 HMDA reporting purposes, these tracts
continue to define the universe of low- and moderate - income neighborhoods in
Las Vegas. "

Of the 8,581 home purchase loans made in Las Vegas in 1990, only 6.8 percent
went to low- and moderate - income neighborhoods ( see Figure 6, page 31 ) .
Among the leading bank lenders, Primerit made 10.6 percent of its loans in these

neighborhoods, Citibank made 8.5 percent, and Valley effectively matched the
area aggregate at 6.7 percent ( see Figure 6 ) . Only First Interstate fell below the
aggregate , making just 2.7 percent of its 563 home purchase loans to low- and
moderate -income neighborhoods.

By way of contrast, upper income neighborhoods received 53.1 percent of all
home purchase loans, and 57 percent of the $ 811 million conveyed by those
loans. Valley once again matched the area aggregate , making 53.1 percent of its
loans to upper income areas. Primerit, too , was close to the area aggregate , at
54.0 percent, while Citibank was slightly above, at 58.7 percent.

First Interstate made a remarkable 83.3 percent of all its home purchase loans in
upper income neighborhoods, 31 times greater than its loans to low- and
moderate - income neighborhoods.

8. The leading banks' market share of home purchase loans in low- and

•
•

Valley came close to having equal market shares, at about 8

•

Primerit had a 50 percent greater share of the low - and moderate

First Interstate's share of the upper income market, at about 10

As discussed in Section 5 , above, market share in a competitive arena reflects the
cumulative results of a bank's policies, objectives, products and financial
commitments . It is not something that can be changed overnight or by directive .
Thus it is a fairly discriminating indicator, in HMDA terms, of an institution's
more durable priorities.

Of the four leading bank lenders in Las Vegas, Valley's market share of home
purchase loans, by income of tract, was most consistent, at between 8 percent and
9 percent in each of the three income categories ( see Figure 7, page 32 ) .

Citibank, with a somewhat smaller overall share of home purchase loans than
Valley, had a slightly greater penetration of the market for loans to low- and

moderate -income tracts than to middle and upper income tracts.
Primerit, as is often the case with S & Ls, whose commitment to home mortgage
lending predates the aggressive entry of commercial banks into the market, has
about a 50 percent greater market share in low- and moderate -income tracts than
in middle and upper income tracts .
First Interstate , among the leading bank lenders, has the smallest share of the

low- and moderate -income market ( less than 3 percent) , and the largest share
( about 10 percent) of the upper-income market. The latter share is due in large
measure to some 375 home purchase loans in just one census tract ( 1980 tract 32;

1990 tract 32.02, the Sun City development) , a clear reflection of a major
commitment by the bank to capture a specific market. The bank's small share of
loans to the low- and moderate -income market reflects the opposite: little or no

commitment to penetrate a market that is clearly underserved .
D. Conclusions

The Home Mortgage Disclosure Act data, which is the basis for this report, has
been described as " the major vehicle forN 13monitoring the results of an institution's

( Community Reinvestment Act] efforts. “ '3 It is the only data banks are required
to make public that objectively reflects the underlying principles on which the
institutions base their operations, including their credit extensions. Nothing
comparable is disclosed regarding consumer loans ( other than home improvement
loans) , commercial or business loans, investments, or other uses of depositors'
7

federally insured funds.

Our examination of the aggregate HMDA record of lenders in Las Vegas, and of
four leading individual lenders, shows an entrenched and consistent set of

patterns. Among them are the following:
•

•
•

Minority and low- and moderate -income communities generate substantial

Minority borrowers, regardless of their income, are denied loans more
Lenders compete aggressively for market share in predominantly white ,

The 1990 HMDA data confirm that financial institutions in Las Vegas, as in other

cities of the United States, continue to play the lending game by rules they write
themselves.

The consequences of this one -sided practice are extensive and not quickly
remedied . Urban riots and attacks on property are but a signal of what happens
when entire communities are kept under a financial chokehold .
The Las Vegas Alliance for Fair Banking will work with other groups and
individuals in Las Vegas, as well as with the financial institutions that operate

here , to address the patterns of inequity and discrimination reflected in this report.
E. Recommendations

The Las Vegas Alliance for Fair Banking makes the following recommendations:
•

The appropriate regulatory agencies should investigate every Nevada

The Federal Reserve should not approve any application subject to CRA
involving financial institutions with disproportionate denial rates between
white and minority applicants until the institutions demonstrate to the
satisfaction of regulators and the communities in which they operate that
they have taken steps to correct such disparities.

•

Regulatory agencies must take lending patterns in low- and moderate

State and local governments should develop legislation, after suitable

Notes

1. The trade names and marks used in Figure 1 are owned by Citibank, First
Interstate Bank, Primerit Bank and Valley Bank .

2. The details cited here are from Eugene P. Moehring, Resort City in the Sunbelt:
Las Vegas 1930-1970 ( Reno & Las Vegas) , 1989, pp . 176-179 and 185. First
National Bank later became First Interstate Bank. Moehring's chapter, " Civil Rights

in a Resort City ," examines the history of the Westside with considerable attention to
housing and living conditions.

3. In the context of its acquisitions of Valley Bank and Security Pacific Bank, and
following negotiations with the Southern Nevada Reinvestment and Affordable

Housing Committee, Bank of America plans to open a branch in the Westside in
1992. Las Vegas Sun, November 20, 1991 .

4. Data on branch deposits of First Interstate Bank and Valley Bank are from the
FDIC's Operating Banks and Branches Data Book, v.6, and reflect deposits as of
June 30, 1990. Data on branch deposits of Primerit Bank are from the Office of

Thrift Supervision's Survey of Branch Deposits, as of June 30, 1990.
5. ACORN , " Take the Money and Run: The Siphoning of Deposits from Minority

Neighborhoods in 14 Cities, " Washington, D.C. , June 4, 1992. See also Paulette
Thomas, " Minority - Area Lenders Faulted in Acorn Study," The Wall Street Journal,
June 5, 1992, A2. A comparable examination of the ratio of deposits to lending in
Las Vegas for First Interstate, Valley, and Citibank produces much greater

disparities. Using total deposits reported for all Clark County branches, and the
dollar value of all home loans made in tracts with less than 20 % minority population
and tracts with more than 50 % minority, the following ratios result:
DEPOSIT TO LOAN RATIOS - LAS VEGASICLARK COUNTY
CITIBANK

FIRST
INTERSTATE

VALLEY

906,383

1,613,322

1,614,380

69,426

60,483

75,928

8 CENTS

4 CENTS

5 CENTS

131

176

771

.01 CENT

.04 CENT

TOTAL DEPOSITS

CLARK COUNTY ( S'000)
LOANS TO WHITE

NEIGHBORHOODS ( 8,000)
AMOUNT LOANED PER
DOLLAR OF DEPOSITS
LOANS TO MINORITY

NEIGHBORHOODS ( $ .000)
AMOUNT LOANED PER
DOLLAR OF DEPOSITS

6. Summaries of the application denial patterns may be found in several articles by
Paulette Thomas in The Wall Street Journal: " Mortgage Rejection Rate for Minorities
Is Quadruple That of Whites, Study Finds, " October 21 , 1991 , A2 ; " U.S. Examiners
Will Scrutinize Banks With Poor Minority -Lending Histories, " October 22 , 1991 ,
A2; " Federal Data Detail Pervasive Racial Gap In Mortgage Lending," March 31 ,
1992, A1 .

7. " U.S. Probing Banks' Records For Race Bias," The Wall Street Journal, May 19 ,
1992 , A2. First Interstate Bank was the only Nevada institution under OCC
supervision identified by The Journal as a potential target for investigation. See also
Bill Atkinson , " 250 Banks Probed for Race Bias in Lending," American Banker, May
14, 1992 , 1 ; Paulette Thomas, " U.S. Intensifies Its Investigation Of Lending Bias, "
The Wall Street Journal, May 15 , 1992 , A3; and Phil Roosevelt, " Banks Face Long
Heat Wave over Loan Bias," American Banker, June 5 , 1992, 1 .
8. For 1990 HMDA reporting purposes, the median family income for the Las Vegas

MSA was $ 34,582 ( data provided by the Federal Financial Institutions Examination
Council) . This was an adjustment upward by a factor of 1.6424 from the 1980
median family income of $ 21,056. The applicant income categories for 1990 HMDA
purposes, therefore, were as follows:

Low- and moderate - income ( < 80 % median ) : $ 27,665 or less
Middle income ( 80 % -120 % median ) : $ 27,665 to $ 41,500
Upper income ( > 120 % median ) : $ 41,500 or more

9. Valley Bank published a correction notice to its 1990 HMDA data stating that an
additional 30 loans were declined by the bank ( not the mortgage company) that were
not originally reported. The breakdown of the additional denials by gender, race or
income was not available .

10. The 1990 census category of " Hispanic origin " includes persons of different
ethnicity, including whites and blacks. For purposes of this section , and later
discussions of demography, references to " white ," " black " and other ethnicities apply
to persons not of Hispanic origin.
11. In 1980, tract 36 was classified for HMDA purposes as more than 80 %

minority.

By 1990, tract 36 had been divided into 36.02, the southern portion, and tract 36.01 ,
the northern portion. Tract 36.02 remains a predominantly minority tract ( 96 % ) ;
tract 36.01 , on the other hand, is not ( 23 %

minority ) . Aggregate and individual

institution HMDA data for 1990 reports loans to tract 36 as if it were still a high

minority tract, when it is reasonable to assume that most if not all loans reported
were actually made in the section that is now 36.01 . To avoid skewing the data in

this report, we either exclude tract 36 or count no loans there. We will gladly adjust
the calculations if any lender provides evidence that loans reported for 1980 tract 36
were made in the section that is now 36.02 .
18

12. The 1990 census data on income are only now being made available . Thus,
changes in median tract income over the 1980-1990 period will not be reflected in

HMDA categoriès until the release of 1992 HMDA reports, in the latter part of
1993. Applicant income, on the other hand, as described in Note 8 above, has been
adjusted to 1990 levels.

13. Fannie Mae, " Investing in Your Community, " Washington, DC, 1990, p.5 .

TABLE 1 .

HOME LOANS IN LAS VEGAS , 1990. ALL LENDERS .

TYPES OF LOANS

# OF LOANS

Home Purchase
FHA, FMHA, VA
Conventional
Refinance

Home Improvement
TOTAL

$ ( MILLIONS )

3333

279.0

5248

532.0

653

73.5

1424

37.7

10658

922.2

TABLE 2.

TOP MORTGAGE LENDERS IN LAS VEGAS, 1990

LENDER

# OF LOANS

MKT SHARE ( % )

$ ( MILLIONS)

1907

17.9 %

192.7

1020

9.6 %

73.0

908

8.5%

91.9

First Interstate Bank NV

754

7.1 %

62.2

Primerit Bank
Margaretten & Co

689

6.5%

61.7

674

6.3%

65.6

5952

55.8%

547.1

Weyerhaeuser Mtg Co
Citibank NV
Valley Bank /Valley Mtg Co

TOTAL

60-893 O - 92 - 2
-

TABLE 3.
LOANS TO THE WESTSIDE. 1990. ALL LENDERS AND FOUR LEADING BANKS.
LOANS

738
301

88
86

381

520
10

0
0
1

3
1

3075

114

10658
0.6 %

922114

1020

72988

0.3 %

754

62220

689

1135

WESTSIDE TOTAL
AGGREGATE /BANK TOTAL
WESTSIDE AS OF TOTAL

。

3.01
2.01

。

37 "
3.02

61762

0.4%

0.4 %

TABLE 4.

BRANCH DEPOSITS AND HOME LOANS - FIRST INTERSTATE BANK AND VALLEY BANK
THE WESTSIDE AND WESTERN TRACT 30.02

VALLEY BANK

564

114

50,130

Rancho Lane
North Las Vegas

46,365

PRIMERIT BANK
North Las Vegas

15,249

NEAR TRACT 30.02

1.172

57,671

VALLEY BANK
Rainbow -Westcliff

31,557

PRIMERIT BANK
Rainbow

28,461

925

1,693

SOURCE: FDIC. Operating BanksandBranches Data Book, 6/30/90; 1990 HMDA Reports;
OTS, Survey of Branch Deposits, 6/30/90

srooo )

114
188
62

190

364

008

91898

1.3%

Note: Population percentage based on 1990 census.

jsia

VALLEY
LOANS

PRIMERIT

TABLE 5.
APPUCATIONS FROM BLACKS AND LOANS TO BLACKS - LAS VEGAS 1990

ALL LENDERS AND FOUR LEADING BANKS
ALL LENDERS

CITIBANK

FIAST
INTERSTATE

PRIMERIT

VALLEY

16960

1433

1004

1226

963

BLACK APPLICANTS

569

% BLACK APPUCANTS

3.4%

3.3 %

2.69

3.9%

3.5 %

10658

1020

754

689

908

ALL APPUCANTS

ALL LOANS
LOANS TO BLACKS
% LOANS TO BLACKS

1.6 %

3.2

TABLE 6.
HOME PURCHASE LOANS TO TRACTS WITH MORE THAN 50 % MINORITY POPULATION
RANKED BY % MINORITY

LAS VEGAS - 1990

3223

996

36.02

3992

1155

3.02

4193

1619

2458

3452

3865

88
86
39
62

503

5
8

21
16

36
9

71
66

5612

1821

1111

5637

1955

1187

5.04

6304

2529

347

4867

2452

250

4134

1253

775

5.03

5478

23
38

2240

1305

6887

3263

2878
6023

2.01

851

183

9954

3.5%

TABLE 7.
TRACTS RECEIVING MORE HOME PURCHASE LOANS IN 1990

THAN ALL PREDOMINANTLY MINORITY TRACTS COMBINED
1980 TRACT

1990 TRACTS

TOTAL HP

LOANS #
55

Tract 55.01 -.02-.03 -.04

199

29.02

Tract 29.08 -.09.11..12

215

Tract 49.01-.02 -.03
Tract 29.05..07
Tract 36.01 -.02 *
Tract 29.06-10
Tract 58.01 -.02
Tract 34.03..04..05..06 -.07

242
249

58
34.02
28.02

Tract 28.03 -.04 .
Tract 53.01.02
Tract 51
Tract 32.01.,02

290
344

464
505
512
618
725
1001

TOTAL HP
LOANS S

18392
15584
19193
31060
37782
49915
54007
47387

49049
62002
74169
. 98654

POP'N

AVG %

NON -HISP

12567

5.8%

22311

16.8 %

26394

22.0 %

11840
18560
20594

13.2 %

11307

13030

22.7 %
13.7 %
16.3%

15.4%
12.2 %
10.7 %

11.9 %
10.3%

LAS SIDE
VEGA
WEST
REDLI
THE
S
NING
Loans
Home
in
1990
,refinance
purchase
of
Number
and
:improvement
loans
home

,refinance
purchase
of
Number
and
improvement
:home
loans

:
Westside
to
loans
of
Number

:
Westside
to
loans
of
Number
rValue
, efinance
purchase
of
:
loans
improvement
home
and

FUN

,refinance
purchase
of
Value
and
improvement
loans
:home

思聞 Ha
EU

:of
Westside
to
loans
Value

2017

of
:Value
Westside
to
loans

2
29

PRIMERIT

CMTBANKO

and
, efinance
rof
purchase
Number
:home
loans
improvement

廷

and
rpurchase
,ofefinance
Number
:home
loans
improvement

的到

A221
:to
Westside
of
loans
Number

:loans
to
Westside
of
Number

rpurchase
,ofefinance
home
Value
:loans
improvement
home
and

Vegas
Las
in
loans
home
10,658

and
,ofefinance
rpurchase
Value
:home
loans
improvement
:to
Westside
loans
of
Value

:to
Westside
loans
of
Value

Vegas
Las
in
loans
of
worth
$922,100,000

Figui

Figure 2.1

%Applications
Denied

Denial Rates for Home Loan Applications
By Race and Income of Applicant

32.3 %

Denlals/Applications:
White :

27.3 %

26.5 %

Low -Mod : 460/1911
Middle: 604/3115

Upper: 1140/7826
23.5 %

23.5 %

SSSR

Black:

19.4%
17.1 %
29

Low -Mod: 42/130
Middle: 42/164
Upper: 66/281

14.6 %

Low -Moderate

Middle

Upper

MSA Average

Income of Applicants
White Applicants

Black Applicants

Valley
VBank
& alley
Mortgage
Company

Interstate
First
Bank
Nevada
of
Denlal
Rates
Home
for
Loan
Applications

Denial
Rates
Home
for
Loan
Applications
Race
By
Income
and
Applicant
of

Race
By
and
Income
of
Applicant
080

Total
:Applicatione
wako.O
%
10

Black

Total
Applications
WARY

% Applications Denied

Applications

fon

son

200

11
7.7

Middle
income
Upper
Income

Income
-Mod
Low

.8Blact
4

101

Bank e
Averag

Middle
Upper
Income

Incomo
M
- od
Low

whne
Appleanta
Applicants
Black

Bank
Average

Applicants
Black
Whne

Denial
Rates
Home
for
Loan
Application
s
Race
By
and
Income
Applicant
of

27.3

Citibank

Primerit
Bank

Denlal
Rates
Home
for
Loan
Applications
Race
By
Income
and
Applicant
of
06

Total
Applications

620

Total
Application
s

%
%0

Middle
Income
Upper
Incomo
Income
M
- od
Low

Bank
Average

% Applications Denied

128
1QS

101

om
347

383

31.4
301
206

Low
Incom
.-Mod

Middle
Income
Upper

Whhe
Applicanto

Applicanto
Black

Bank ge
Avera

Figures 2.2-2.5

Applications Denied

Who
1
- 120
Blact40
17.4

WR
.
Block
4

son

Figure 3.1

All Lenders in Clark County - 1990
Home Loans to African -Americans

African -Americans : 308 loans ( 29% )

10658 Loans to all Residents

Total Includes :

Home purchase loans
Refinances

Home improvement loans

First
Interstate
Bank
Nevada
of
Home
Loans
AAfrican
-tomericans

Valley
VBank
& alley
Mortgage
Company
Home
Loans
AAfrican
-tomericans

Clark
1-County
990

-1Clark
County
990

AAfrican
2
1loans
%)(1:-.6mericans

Total
loans
9=08

Total
loans
=54
7

value
million
$9Total
=
1.9

Total
value
$6
=
22
million

Citibank
Primerit
Bank

Home
Loans
to
African
-Americans

Home
Loans
AAfrican
-tomericans

Clark
1County
- 990
1Clark
- 990
County

AAfrican
loans
76
%)(2:- mericans

AAfrican
loans
2:- mericans
22
%)(3

Total
7value
million
$=3.0

Total
value
=$6million
1.8

rigures J. :-) . )

Total
6loans
=89

Total
1loans
=020

Figure 4

Home Purchase Loans - Las Vegas 1990
12 %
Total Loans Originated by
An Lenders In MSA :

Market
Share

10 %

Patani

0 %

Citibank

First Interstate
Whito

Primerit
Black

Valley

Figure 5

Clark County Population by Racial Groups

Asian ( 3.3 % )
24,483

Hispanic ( 11.2 % )

White ( 75.4 % )

Source: U.S. Census, 1990

Figure 6

All Lenders and Leading Banks
Distribution of Home Purchase Loans

Loans
%of

83.3 %

531 %

54.4 %

53.1 %

40.1 %

523%

Figure 7

Home Purchase Loans - Las Vegas 1990
Market Share by Income Characteristics of Census Tracts
=

12 %

Market
Share

10 %

8 %

6 %

4 %

0 %

Citibank

'First Interstate

Low -Moderate Income

Primerit

Middle Income

Valley
MV Upper Income

TUESDAY, MAY 19, 1992

U.S. Probing
Banks' Records
For Race Bias

Banks With Lending Disparities
Listed below are banks and mortgage companies where the rejection rate for
black and Hispanic mortgage applicants was more than twice that for whites in
1990.

Group of Lenders Confirms

A WALL STREET JOURNAL News Roundup

Alaska

Oklahoma

The Wall Street Journal

May 19 , 1992 , A2

The Comptroller's letters of inquiry
note specifically that the government isn't
presuming that racial discrimination is
behind the lending patterns it is scrutiniz:
ing. And bankers are quick to note that the
Federal Reserve's mortgage data doesn't

take into account credit history or debt

income areas. Many also are offering
credit counseling programs through local
community groups and churches. Others
have altered their loan approval stan

dards, allowing, for instance, rent pay.
ments to be examined for credit history for
a borrower who never had a credit card.

load .

Many banks contend that the rec
ord compiled as part of last year's Federal

Reserve study distorts the true lending
picture. Most importantly , it doesn't in

clude a borrower's credit history, the No. 1
reason for loan rejections, banks say.
" The only criterion we apply in mak
ing credit-granting decisions is the credit
worthiness of individual borrowers, " said
Jim Lestelle, a spokesman for Hibernia .
" That would be the reason for the differ

ence in those figures."
Credit Standards

Expanded HMDA Data

onResidential Lending: One Year Later
Glenn B. Canner, of the Division of Research and
Statistics, and Dolores S. Smith, of the Division of

Consumer and Community Affairs, prepared this
article .

Questions about the access of minorities and lower
income households to home morgage loans contin

ued to draw considerable attention in the past year.
Indeed, the release of new data in October 1991
documenting the credit experiences of various
groups during 1990 intensified the discussion and
stimulated initiatives in the private and public sec
tors to address perceived inequities. The data on
home lending, which cover metropolitan areas
throughout the United States, are available as a
consequence of the 1989 amendments to the Home
Mortgage Disclosure Act ( HMDA ) , which greatly
expanded the scope of the act.

A first study of the expanded HMDA data,
reported in the November 1991 Federal Reserve
Bulletin , depicted certain statistical relationships
that the data revealed about lending activity nation
wide. ? Among the findings, the one that aturacted
the most attention was that black and Hispanic loan

applicants were denied credit in greater proportions
than white applicants, even within the same income
groupings. The data showed similar variations in
rates of loan disposition among neighborhoods
classified by their racial composition and income
characteristics. The HMDA data have clear limita

Lions. Foremost among them is the general lack of
information about factors important in assessing
the creditworthiness of applicants and the adequacy
of collateral offered as security on loans. Without

such information , determining whether individual
applicants have been treated fairly is not possible.
Nonetheless, the lending patterns depicted by the
data have led many persons to conclude that wide

spread racial discrimination characterizes the
home-lending process.

ments to Regulation C ( 12 C.F.R. 203 ) . These amendments will
establish a new set of criteria as of January 1 , 1993, for determining
coverage for independent mongage companies. The new rules are
expected to bring the soul of independent mortgage companies
covered by HMDA to more than 1,000 institutions.

1. The 1989 changes to the ad also extended coverage to some
independent mongage companies those unaffiliated with a depos
itory institution. The Federal Deposit Insurance Corporation
Improvement Aa of 1991 ( FDICIA ) extends coverage to even

more independent mongage companies. To implementthe provi
sions of FDICIA , the Federal Reserve Board is adopting amend

802

Federal Reserve Bulletin

November 1992

After presenting national aggregates from the
1991 reports, this article describes some of the
responses within the public and private sectors to
the data released a year ago. These responses
include research projects that seek objective expla
nations of the statistical patterns, investigative and
enforcement eſlors by federal regulators to ensure
compliance with fair-lending and community rein
vestment laws, educational measures to increase
awareness of lenders' responsibilities and to inform
consumers better about the mortgage loan process,
and practical ideas for identifying and eliminating
lending practices that may discriminate against
minorily applicants, including a careful examina
tion of any unintended adverse effects of underwrit
ing standards.3 The aricle discusses as well the
special role that entities in the secondary mortgage
market play in the home-lending process and steps
such institutions have taken to promote affordable
housing.

1. Residential lending activity reponied by financial

SUMMARY RESULTS FOR 1991 HMDA DATA

ity of a particular institution in a specific metropol
itan statistical area ( MSA) . Although the number of
reporting institutions in 1991 remained about the
same as in 1990 , the volume of reported applica
tions and loans increased substantially.

For lending activity in 1991 , the FFIEC prepared
disclosure statements for 9,358 reporting
institutions — 5,551 commercial banks, 1,536 sav.

Number
Year

1981
1982
1983
1984
1985
1986

1.28
1.13
1.71
1.86
1.98
2.83

1987
1988
1989
19902
1991

3.42
3.39
3.13
6.59
7.89

Number of
Number of
reporting
institutions

metropolitan

8.094
8.258

10.945
11.357
10.970
11.799
12.567
12329

8.050
8.491

9,072
8.898
9,431
9319
9.203
9.332

9.358

13.033
13.919
14,154
24.041
25.934

1. Before 1990, includes only loans originated by covered institutions;
beginning in 1990 ( hurst year under revised reporting system ) , includes loans
originated and purchased, applications approved but not accepted by the
applicant applications denied or withdrawn, and applications closed because
information was incomplete

ings and loan associations, 1,436 credit unions, and
835 mortgage companies, of which 528 were
unaffiliated with a depository institution ( table 1 ) . “
These disclosure statements consisted of 25,934

individual reports, each covering the lending activ
Volume of Applications and Loans
3. The federal banking agencies include the Federal Reserve
Board, the Federal Deposit Insurance Corporation ( FDIC ) , the

Office of the Complroller of the Currency ( OCC ) , the Office of
Thrift Supervision ( OTS ) , and the National Credit Union Adminis
tracion ( NCUA ) . The other enforcement agencies are the Depart
ment ofJustice, the Department of Housing and Urban Develop
ment ( HUD ) , and the Federal Trade Commission .

In 1991 , lenders covered by HMDA acted on
roughly 6.56 million home loan applications
3.26 million for purchasing, 2.09 million for refi
nancing, and 1.18 million for improving dwellings
for one to four families, and the balance for loans

on multifamily dwellings for five or more families
( table 2) . As in 1990 , nearly three - quarters of the
reported applications for home purchase loans were
for conventional mortgage loans; the remainder
were for government-backed forms of credit
loans insured or guaranteed by the Federal Housing
Administration ( FHA ) , the Veterans Administration
( VA ) , or the Farmers Home Administration
( FmHA ) .

Expanded HMDA Dala on Residential Lending: One Year Loter
Among the various types of loans used to pur
chase homes, those backed by VA guarantees
changed the most from 1990 to 1991 , increasing

803

costs , they are particularly appealing to prospective
borrowers who have limited financial resources.

27 percent. The lotal number of conventional loans
was virtually unchanged, while the number of

FHA -insured loans fell about 1.5 percent:

Use of Various Loan Products
for Home Purchase
Like the 1990 data, the 1991 HMDA dala reveal
large differences in the types of home purchase
loans that applicants, grouped by their income and
racial characteristics, sought during the year
( table 3) . In general, government-backed home pur
chase loans are more likely to be requested by
households with relatively low incomes than they
are by borrowers with higher incomes. In 1991,
38.8 percent of applicants with low incomes
( income less than 80 percent of the median family
income for their MSA) applied for government
backed loans, compared with 15.4 percent of appli
cants with high incomes ( income more than
120 percent of the median family income for their
MSA) . The heavy reliance of lower-income appli

Disposition of Loan Applications
HMDA data for 1991, like the data for 1990 ,
indicate that lenders approve most applications they
receive for home purchase loans. In 1991, lenders
approved roughly 71.2 percent of applications for
convencional home purchase loans and 71.7 per

canis on government-backed loans reflects two

cent of applications for government-backed loans

principal factors. First, such households are much
more likely to buy homes that are within the maxi
mum limits of FHA loan insurance ( between
$ 67,500 and $ 124,875, the latter amount for locali

( lable 2 ) . Among the applications for conventional
loans, 18.9 percent were denied by lenders; for the
balance, either the consumers withdrew their appli
cations or the lender closed the application file
after the prospective borrower was asked for, but
failed to submit, information required for the credit
decision

ties with relatively high prices for homes) . Second,
households with lower incomes, which on average

have substantially fewer liquid and other financial
assets than do higher-income households, are much
more likely than households with high incomes to
face significant liquid -asset constraints. Because
government-backed loans allow very low down
payments and the financing of a portion of closing
5. See Arthur Kennickell and Janice Shack -Marquez " Changes

in Family Finances from 1983 60 1989: Evidence from the Survey
of Consumer Finances," Federal Reserve Bulletin, VoL 78 ( January
1992) . pp . 1-18.

6. The HMDA data indude Hispanics of all races in the His
panic category, in contrast to dau compiled by the US. Census

Bureau,
which differentiate between while Hispanics and nonwhite
Hispanics
.

804

Federal Reserve Bulletin

November 1992

2. Disposition of applications for home loans, by purpose and type of loan, 1991
Number in thousands, and percenuge distribucion
Homne purchase
Disposition

Lom originaled .

Valarans

' Federal Housing
Administracion

Adminigration

Fara Home

Convencional

Administration

Nurnber

Percent

Number

Percent

Nurnbe

Percent

Numbes

Percent

450.6

67.7

134.6

720

1.0

627

1.615.2

67.2

Applicauan approved but na accepied by applicant
Application denied
Appucation withdrawn
Fie closed ( information incompiac)

1. Components may na sum lo louls because of rounding.
• Fewerthan 500

can likely qualify. Results of a recent consumer
survey sponsored by the Federal Reserve Board

100

SOURCE FFIEC, Home Morgoge Disclosure Act

most frequently cited reason for credit denial is
credit history.

indicate that nearly 70 percent of the families that
purchased a home within the past three years, and
that financed the purchase with either a mortgage
or a land contract, received information from real
estate agents or loan officers about whether they

were likely to qualify for a loan ? Also receiving

7. Board of Govemors of the Federal Reserve System , “ Survey
of Consumer Credit Shopping Aaivities " ( forthcoming ) .

8. Data from the Morgage Bankers Association show that, after
reaching a len - year low in the first quarter of 1990, delinquency
rates on conventional mongages rose sharply through the second
quarter of 1991. Delinquency rates have moderated some since
then

Expanded HMDA Daia on Residential Lending : One Year Later

805

2.- Continued

Loans on one - ho four-family dwellings
Launs on

multifumily dwelling

Disposition
Lou originated ....
Application approved but not accepted by applicant
Application denied
Applicationwithdrawn
File closed ( information incomplae)

Homerehouncing
Number

Percent

1.488.3

71.1

Home
improvement

Nurnbes
7323

Parcent

Nurnber

Percent

622

18.0

56.1

1.2
1.1

1.176.4

321

writing guidelines of the various secondary market

for 61.2 percent of all applicants for conventional

institutions and thus frequently follow these guide
lines in assessing loan applicants. Lenders may
also approve nor.conforming Loans for their own
portfolios, however. Sometimes they originale
these loans under special lending programs that
apply highly flexible underwriting standards; at

cent in 1991. ( For purposes of comparison , appli
cants in both 1990 and 1991 were categorized
using the median family income figures for each
MSA as estimated by HUD.) This change results
from both a decline in the number of high -income

home loans in 1990, they accounted for 53.6 per

other times, the lender's familiarity with the pro

applicants and an increase in the number of low

spective borrower allows an extension of credit
when a strict application of the underwriting guide

income applicants. The larger number of low
income applicants may be due to enhanced market
ing and outreach efforts by lending institutions and
the implementation of innovative lending programs
by the secondary market ( see the discussion on the

lines might suggest otherwise. Some evidence that
lenders are selling more loans 10 the secondary
market comes from the HMDA data. In 1990 ,
lenders covered by HMDA reported selling 46 per
cent of the conventional home purchase loans they
originated ; in 1991, they reported selling 51 per

cent. The proportion of loans for refinancing that
were sold to the secondary market increased even
more significanuy, from 39 percent in 1990 to
51 percent in 1991.

secondary market below ) . The growth at the lower
income end of the market is consistent with dala

from other sources that indicate a greater propor
tion of first-time homebuyers in 1991 than in 1990.
Such homebuyers likely have lower incomes than
current owners buying new homes. Finally, the
growth at the lower - income end of the conven
cional market may reflect a shift in preferences
among some home purchasers away from FHA
insured loans toward conventional loans. In July
1991, the FHA loan program was modified in
several ways that made these loans relatively less

desirable to prospective mortgage borrowers ( for
instance, only 57 percent of closing costs instead of
100 percent could be financed ) .
Disposition Rates for Different Groups
ofApplicants
Although most applications for home loans are
approved, the rates of approval and denial vary

806 Federal Reserve Bulletin

November 1992

3. Number of home loan applications, by purpose of loan, characteristics of applicant, and characteristics of census
Home purchase
Horne refinancing

Applicant or census tract characteristic
Governmen backed '

Conventional

3,865
12201
84.450
54.749
621.482

10.977
92,018
93,051
106,809
1,882,748

MSA median )
Less than 80 .
80-99
100-120

264.240
133.285
104 548

417.442
209,629
214,499

More than 120

177 330

970,689

Less than 10 .

333.557

1,086,285

Race of applicant
American Indian/Alaskan native
Asian Paciſc Islander
Black
Hispanic
White
Other

7,952

92,104
62,182
103 575
1,608,452

Home improvement

6,746
21.294
95,671

Joint ( white minority)

Income ofapplicant ( percentage of
281.647
190,739
213 218
989,908

307,526

974,916
323 554
240.010
84.670
52.992

523,105

127,973
113,496

348.328

Racial composition of Census tract ( minorities
as percentage of population)
10-19

180,309

20_49
SO_79

93.868

80-100

22,182

26.200

139.319
116.983
49,694
61,043

Income of census tract'
Low or moderate
Middle

Upper

83.336
437,00
135.758

183.520

178.379

1. Loans backed by the Federal Housing Administration , the Veterans
Administration, and the Farmers Home Administration

considerably among applicants grouped by their
income and racial characteristics ( table 4) . Nation
wide in 1991 , 79.1 percent of the applicants for

conventional home purchase loans whose incomes
placed them in the highest income grouping were
approved for loans, compared with 59.8 percent for
the lowest income grouping. Similar relationships

between approval rates and applicant income are
evident for other types of loans, including those for
refinancing and for home improvement.

and experience more frequent periods of
unemployment.

Expanded HMDA Dala on Residential Lending: One Year Laler
The differences in denial rates for applicants
categorized by their race or national origin partly
reflect differences in the proportion of each group

807

home loan products and granting mortgage
loans.

with relatively low incomes.In 1991, for instance,
22.9 percent of white applicants who applied for
conventional home purchase loans had incomes
that were less than 80 percent of the median family
income for thcir MSA . The comparable perceni
ages for blacks, Hispanics, and Asians were
39.8 percent, 26.2 percent, and 12.1 percent
respectively. Although income levels may account
for some of the variation in loan disposition rates
among racial groups, other factors account for most
of the differences. This conclusion is evident
because, after controlling for income, white appli
cants for conventional home loans in all income

groupings have lower rates of denial than black and
Hispanic applicants ( table 6) .

Department of Justice Investigation
EVALUATING THE DATA

in Atlanta

The HMDA data provide little insight into the

For the past several years, the Department of Jus
tice has been investigating home-lending practices
in Auanta and , in particular, the practices of Deca
tur Federal Savings and Loan Association , one of
the largest home lenders in the city. The Depart
ment of Justice launched its investigation after a
series of articles, published in 1988 in the Atlanta
Journal Constitution, documented wide differences
in the number of home mortgage loans extended in

financial circumstances of loan applicants or the
characteristics of the properties that applicants seek
to purchase , refinance, or improve. The data reveal
that credit history problems and excessive debt
levels relative to income are the reasons that lend-

ers most frequently give for credit denial; but specific information for applicants on their level of
debt, debt repayment record , employment experience, and other factors pertinent to an assessment
of credit risk - is not available . Moreover, the
HMDA data include no information about the specific underwriting standards used to assess each

prospective borrower's application. Thus, the data,
by themselves, provide liwe basis to assess the
faimess of the loan process.

Auanta neighborhoods grouped by their racial com
position. " The investigation has been wide -ranging
but has focused on a detailed review of the files of
more than 4,000 applicants for mortgage loans,
using statistical techniques to control for differ
ences in the financial and economic circumstances

808

Federal Reserve Bulletin

November 1992

4. Disposition of home loan applications, by purpose of loan and characteristics of applicant 1991
Home purchase
Applicant

Governinent-backed ?

Conventional

characteristic

With
drawn

Approved

Denied

64.2
74.9
61.4
68.4
74.3
68.7
74.0

221
125
26.4
18.9
16.3
16.3
15.9

121
11.4
10.3
11.0

66.2
77.6
79.1
79.8

25.2
13.6

75
7.8
7.8
8.1

..With
drawn

Fiic clases

Toul

8.8

1.4

11.0

1.4

100
100
100
100
100

File closed

Toial

Approved

Denied

1.6
1.2
1.8
1.7
1.1
24
1.0

100
100
100
100
100
100
100

624
726
53.3
615
73.7
67.2
71.9

273
15.0
37.6
26.6
173
19.9
175

1.1
1.0
1.0
1.1

100
100
100
100

59.8
75.0
77.8
79.1

328
16.8

6.7
7.4

13.7

7.7
8.9

Race
American Indian /
:୯

1.0
15

8.2
11.S
9.8

1.4
.8

100

.7
.8
.8

100
100
100

100

Income ( percentage of
MSA median ) '

121
11.0

1. Components may na sum to louls because of rounding.

。

Less than 80
80-99
100-120
More than 120

11.1

3. MSA median is median family income of the metropoliin sulistical
area in which the property related to the loan is located.

Decatur Federal had violated the Fair Housing Act
and the Equal Credit Opportunity Act ( ECOA ) by
treating black applicants less favorably than white

excluded large portions of the black community

from its defined lending market and that the institu

applicants. 12

Lion “ rarely or never ” advertised its home loan
products in media oriented to the black community.

12. Both the ECOA and the Fair Housing Aa prohibit discrimi
nation on the basis of race or athnic origin , gender, and religion . In
addition , the Fair Housing Act prohibits discrimination on the basis

Federal Reserve Bank Study of Institutions

of handicap or familial sualus; and the ECOA prohibits discrimina

Lion on the basisof age and mariial salus, because income is
derived from public assistance, or because a right under the Con
sumer Credit Prolection Act is exercised. The civil righus nos of

1866
and 1870, 600, have been interpreted to bar racial discrimina
Lion in lending

in Boston

Racial disparities in patterns of mortgage lending
have long been a concern in Boston . A 1989 study

Expanded HMDA Dala on Residential Lending: One Year Laler

809

4. - Continued

Hane refinancing

Home improvement

Applicant
characteristic

With
drawn

File closed

66.0
68.7
58.1
59.6
76.8
62.6
70.7

21.2
18.7
29.5
26.9
13.7
24.2
18.8

11.1
10.6
10.9
11.8

1.6
1.9
1.4
1.7

693
75.1
76.3

20.6
15.9
14.6

75.5

14.3

9.2
8.2
8.2
9.1

Toul

Approved

Danied

100
100
100
100
100
100

23.9
32.0
44.2
395
21.1

100

69.0
59.0
50.9
S5.7
74.0
58.6
693

100
100
100
100

59.2
67.0
69.8
723

With

by the Federal Reserve Bank of Boston docu
mented differences in lending patterns across
neighborhoods grouped by their racial composi
tion. 13 That study was based primarily on informa

tion from records of property transfers and on data
about neighborhood characteristics from the 1980
U.S. census of population and housing. Information
about individual borrowers or loan applicants was
unavailable to the researchers. The study found
that, after controlling for a wide variety of factors
related to the economic characteristics of neighbor
hoods, the number of mortgage originations rela
tive to the number of owner -occupied housing units
was 24 percent lower in predominantly black
neighborhoods in Boston than in predominanty
white areas. The researchers could not, however,

conclude with certainty the causes of the observed
differences in lending.

13. Katherine L Bradbury, Karl E Case, and Constance R
Dunham , " Geographic Paiems of Mongage Lending in Boston ,
1982-1987," Federal Reserve Bank of Boston , New England Eca
nomic Review ( September /Oaober 1989) . pp. 3–30 .

4.6
7.8
4.3
4.1
4.4

34.2

6.3

25.2

4.8

35.7
27.6
24.7
215

4.7
4.8
4.9

‫وا‬
‫ب‬
‫منم‬

1.0
.9
.9
1.1

File closed

Toul

drawn

5.3

sà onraña

Denied

100
100
100
100
100
100
100

1.2

nyooo

Approved
Race
American Indian /

100
100

eral Reserve Bank of Boston augmented the
HMDA data with information for about 1,000 black
and Hispanic applicants who had applied for con
ventional home purchase loans in the Boston area
in 1990 and for a control sample of roughly 3,100
white applicants. The additional data were re
quested from the 131 financial institutions that had
received lwenty -five or more mortgage applica

lions, from among 352 lenders that had filed 1990
HMDA data for the Boston metropolitan area.
Lenders assembled data for applicants identified by
the Federal Reserve Bank and reported thirty -eight
additional pieces of information about each one
pertaining to financial characteristics, employment

experience, and credit history. The data were items
available to the lender on residential loan applica

tion forms, credit bureau reports, and loan
underwriting worksheets.

810

Federal Reserve Bulletin

November 1992

5. Disposition of home loan applications, by purpose of loan and characteristics of census tract

Home purchase
Census traat

Governmeni-backed ?

Conventional

characicosuc

With
Denied

10-19
20_49
SO- 79
80-100

80.1
61.8
73.1
68.7
66.0

11.1
30.6
15.6
18.5
20.5

7.8
6.9
10.1
113
11.9

Income
Low or moderate
Middle

71.1
71.6

17.1

Upper

795

10.5
7.6
8.6

drawn

Racial composition
( minorities as percentage

of population)
Less than 10

20.0
10.9

Filc closed

Tuwasio

Approved

.8
1.3
1.5
1.7

1.3

1. Components may not sum to touls because of rounding.

quite similar 10 that of white applicants. Other
substantial differences between the two groups
were that minority applicants were much more
likely to be seeking to buy two- to four -family
properties than single -family properties and were
more often applying for loans with high loan-10
value ratios, most of which required private mort
gage insurance for approval. Black and Hispanic
applicants were also more likely than whites to be
seeking a loan under special lending programs

offered in the Boston market in 1990.

Approved

Denied

With

Toul

100
100
100
100
100

79.4
626
69.5
65.6
59.9

121
28.1
18.9
25
2.4

7.8
8.5
10.4
10.6
11.2

100
100
100

66.1
71.5
79.8

22.4
19.8
105

10.3
8.0
8.8

Toul

100

100
100

1.2
1.4
1.6

1.2
.8
.9

100
100

100

meropolitan suatistical area ( MSA ) as a whole; in middle - income census
Lracts , median family income is 80 percent to 120 percent of the median
MSA fardy income; in upper- income census tracts, median family income
is more than 120 percen : of themedian MSA family income.

accounted for much of the disparity that was appar

ent in the HMDA data, but they do not appear to
explain it entirely. Specifically, the study found that
if minority applicants had the same economic and
property characteristics as white applicants ( there
by differing only by race ) , they would have experi

enced a denial rate of 17 percent, compared with
11 percent for whiles. Stated another way, the
analysis indicates that the denial rate for minority
applicants would have been 20 percent if the race
of the applicant had not been a factor - compared
with the actual denial rate of 28 percent revealed

by the HMDA data.

New York State's Review of 1989 Data
In March 1992, the New York State Banking
Department released the findings of a study of the

Expanded HMDA Data on Residential Lending: One Year Laler

811

5. - Continued

Home refinancing

Home improvement

Census tract

characteristic
Approved

Denied

With
drawn

8.5
10.3
10.8
11.7
129

100
1.3
1.4
1.5
1.7

100
100
100
100

735
64.6
S9.5
S21
465

215
28.8
33.9
413
47.1

4.6
5.7
5.8
5.7
5.6

11.5
9.0
9.6

1.4
1.0
1.2

100
100
100

53.9
69.0
71.8

40.2
25.6
21

5.2
4.9
5.4

77.8
70.5
67.3
63.0
56.6

128
17.9
20.4
23.8
28.8

Low or moderate
Middle

63.5

Upper

75.2

23.6
15.6
14.0

Racial composition
( minorities as percentage
of population)
Less than 10
10-19
20_49
SO - 79
80-100

File closed

9.9.9.

With
drawn

j
aoova

Toul

Denied

vinnin

Füe closed

Approved

Toul

100
100
100
100
100

Income '
74.4

mongage-lending practices of ten savings banks in

metropolitan New York. 14 The department exam
ined the banks' mortgage loan files in detail, focus
ing on the treatment of minority and female appli
cants and of applicants seeking to buy homes in
areas of low income and in those with high percent
ages of minorities. It first reviewed the underwrit
ing criteria that the lending institutions used, to
determine whether they were in keeping with or
more restrictive than industry and secondary mar

ket standards. It then considered the actual applica
tion of the criteria, to determine whether they were
consistently applied or whether exceptions had
been made - or not made — in a way that indicated
discriminatory treatment.

14. Emest Kohn, Cyril E Foster, Bemard Kaye, and Nancy J.
Terris, " Are Mongage Lending Policies Discriminating ? -A Swdy
of 10 Savings Banks," New York Sute Banking Department
( March 1992) .

록

S
.7

100
100
100

criteria were generally in line with indusưy and
secondary market standards. In some of the other
institutions, the standards were more suingent

than the industry norm and hence could dis
advantage minorities, women, and lower - income
applicants. Specific policies in question included a
maximum fifteen - year maturity.on mortgages
( which could require higher monthly payments than
a person with low income could afford ) , the
requirement of a 20 percent down payment, and the
charging of nonrefundable fees to applicants who
could have been turned away after a preliminary
review .

812 Federal Reserve Bulletin

November 1992

6. Disposition of home loan applicacions, by purpose of loan and income and race of applican . 1991 1
Home purchase
Applicant
income and nice

Ceavencional

Government-becked '
With
drawa

Approved

Deriod

61.0
723
58.8
64.6
67.3
68.9
67.1

28.6
15.2
30.1

685
78.8
66.8
70.8
795
69.4
76.4

18.8
10.3
221
17.1
125
14.4
14.8

115
9.8
9.4
105

71.2
78.7
66.9
71.2
81.0
75.9
79.1

15.0

11.1
10.6

69.8
783
693
75.7
81.8
73.2
79.2

15.3
10.6
19.5
13.3

Fie closed

Toul

Approved

Denied

With
drawn

Filc closed

Toul

1.2
1.3
1.7
1.6

100
100
100
100
100
100
100

53.9
685
44.9
S4.0
61.7
63.0
59.2

38.6
20.2
48.2
37.1
315
29.1
33.2

64
10.4

1.1
1.0

100
100
100
100
100

1.2
1.2
1.7
1.6

100
100
100
100
100
100
100

69.4
75.8
60.7
64.9
71.0
69.7
729

19.9
13.9
30.0
253
153
193
185

9.7
9.1
8.1
8.7
7.0
10.1
7.8

1.0
12
1.2
1.1
.7

26
1.0
1.6
1.4

100
100
100
100
100
100
100

73.9

17.1
13.7
26.1
23
122
183
15.0

8.0
9.5
8.8
9.6
7.2
10.6
9.3

1.0
13
13

1.7
1.0
1.4
1.4
1.0
24
.6

100
100
100
100
100
100
100

73.4
743
66.0
685
81.0
70.4
768

15.7
13.6
232
19.8

9.7
10.8

1.2
13
1.1
13

Less than 80
Amcdian Indian

25.2
17.7

23.5

93
11.2
9.4
9.9
6.6
11.8
8.4

1.6
1.0

62
7.8
62
8.1
7.2

100

80-99

American Indian

755
63.9
67.0
79.9
70.1
75.0

8888888

23.9

1.1

100
100
100
100
100
100
100

More than 120
American Indian

13.2
10.2
9.7
9.6
75
9.1
8.9

1. Components may not sum to touls because of rounding.

REGULATORY EFFORTS :
ENFORCEMENT OF FAIR LENDING AND CRA
Federal regulators are expanding data analyses to
strengthen enforcement of fair lending and of
compliance with the Community Reinvestment Act
( CRA ) . 15 Acting in concert, the agencies are devel
oping techniques using automated access to the
data in looking for evidence of differential treat
menl Through these efforts they are also seeking to
identify the factors that underlie disparate lending

100
100
100
100
100
100
100

3. Loans backed by the Federal Housing Administration, the Veterans
Administration , and the Fumers Home Admninistracion.

pattems. In October, they issued a joint statement
that addressed the issue of disparate treatment,
autempting to shift the focus from a debate about
whether unequal treatment is occurring to initia

tives that willensure fair lending practices.

Interagency Statement on Disparate
Treatment in Mortgage Lending
In a joint statement dated October 9, 1992, the
regulatory agencies outlined initiatives for ensuring

15. The CRA requires federal agencies to encourge depository

instituzions to help meet the credit needs of their communities,
including low- and moderate - income neighborhoods, in a manner
consistent with safe and sound lending praaices.

that minorities have equal access to home lending
and reemphasized concerns about fair treatment
of applicants for mortgage loans. The statement

813

Expanded HMDA Data on Residential Lending : One Year Later

6. - Continued

Home improvement

Home refinancing
Applicant
income and race

With

Approved

Denied

621
64.4
S5.7
56.9
73.4
S6S
65.1

26.4
23.5
326

67.4
71.6
59.5
62.1
78 S.
62.0
70.2

21.8
18.4
29.4
26.3
13.5
24.5
20.9

10.2

66.9
725
58.9
62.1
79.6
67.5
72.4

21.3
17.0
29.3
26.0
123
222
18.4

10.8

68.6
69.1
60.6
622

19.3
18.5
27.7
24.8
12.4
22.9
17.4

10.7
10.4
105
11.5

File closed

drawn

Toul

Approved Denied

Wuh

File closed

Toul

drawn

71.7
59.7
51.0
S5.2
75.1
57.1
67.1

22.0
31.9
43.9
40.4
20.4
37.1
28.9

4.8
7.4
4.4
3.7
4.1
SS
35

100
100
100
100
100
100

2.8
31.9
40.1
37.4
18.6
28.2
23.7

3.9
7.9

100

71.7
58.9
54.2
57.8
76.8
63.9
71.1

1.5
20
1.2
1.6

100
100
100
100

73.5
620
59.1
61.4

20.1
29.4
34.9
33.6

6.0
7.2
53

100

78.4

163

1.6
.9

100
100

624
729

30.0

4.0

4.0
5.0
4.7

80-90
American Indian

783

64.9
72.5

1.4
1.0
1.2

.9
1.4
1.3
1.0

pointed to increased evidence that the differences
in loan approval rates between white and minority
home mortgage applicants that characterize some
lending may be unwarranted by economic factors.

215

1.0

4.4
4.7

6.6
5.0

100
100
100

100
100
100
100

100
1.4

5.0
3.9
4.2
6.8
4.8

100
100
100
100
100
100
100

100

100
100

1.1

‫ܬܘܚܬܝܪܝܝ‬

100
100
100
100
100
100
100

18.0
31.6
24.4

100
100
100
100
100
100

‫بیا‬

30.4
41.9
48.8
46.6
265
43.6
33.2

3.8
7.1
4.1

100

6SS
50.2
46.5
48.8
69.2
S0.9
61.7

1.4
1.3
1.4
1.2

awawson

10.1
10.8
10.4
11.4

dodania

30.4

‫ماهمماا‬
‫م‬
‫ه‬

Less than 80
American Indian /

100

100
100

100
100
100
100
100

100

ment; development of training programs to ensure

fair treatment of prospective borrowers; credit
counseling for groups of prospective loan appli
cants; participation on mortgage review boards;
and use of “ shoppers " hired by an institution
10 test its personnel's adherence to its own
procedures. 16

16. Mongage review boards are organizations in which partici
paring lenders review the underwriting decisions on loan applica.
cons.The objcaive of a mongage review board is to help ensure
nondiscriminatory treatment of loan applicants and 10 encourage

additional lending. For example, in the Delaware Valley Mongage
Plan implemented by lenders in the Philadelphia ares, reviews are
automatically obuined before the lender denies an application. ( For

an evaluation of this plan, see Paul S. Calem , " The Delaware
Valley Mongage Plan: An Analysis Using HMDA Daua, " Federal
Reserve Bank of Philadelphia, Working Paper No. 92-3. February
1992 ) In Boston and Decroit, loan applicants can appeal decisions
to the review boards after being wmed down

814

Federal Reserve Bulleun

November 1992

Fair -Lending and CRA Compliance

Expanded HMDA Daia on Residential Lending : One Year Laler

815

Lender Education

The FFIEC's brochure " Home Mongage Lending and
Equal Treatment " published in March 1992 , alerts lend
ers to sublle forms of discrimination that may occur

Employment stability. Standards that require a fixed
number of years on the same job may exclude individuals

despite an institution's stated policy of fair lending. It

with one employer.

who have consistent employment but not necessarily

suggests that lenders take a closer look at long -acccpied

practices in loan origination, underwriting, appraisal, and
marketing that can have discriminatory effects, such as
the following:

• Property standards and minimum loan amounts.
Setting limits on the maximum age or minimum size of
the property may exclude homes in minority and low

income areas, especially in urban centers characterized
by older and smaller dwellings. So may minimum loan
amounts that are unrcalistically high for these areas.
Sometimes lenders believe loans must be of a certain size

to qualify for purchase on the secondary market. Neither
FNMA nor FHLMC has a standard for minimum loan

amount that prevents their purchasing small mongage
loans.

ing the Process and Your Right to Fair Lending" to

each covered institution to adopt a branch -closing

inform consumers about the mortgage application
process and about their rights under fair-lending

policy and to provide ninety days' notice of any
proposed branch closing to its customers and to the
appropriate federal regulator. The notice to the
regulator must include a detailed statement of the
reasons for the decision to close the branch and

and consumer protection laws.

statistical information to support the decision. A
notice of the decision must be posted at the branch
thirty days before its closing.

816 Federal Reserve Bulletin

November 1992

reports used in evaluating their applications for
home purchase and home improvement loans.

credit issues, and are often developed in conjunc
Lion with state bankers ' associations.

Community Affairs Programs

EXPANDING OPPORTUNITIES: PRACTICAL

The banking agencies conduct educational and
informational programs to promote fair lending and
10 foster affordable housing and small-business
lending, frequently on an interagency basis. The

RESPONSES FROM LENDERS AND OTHERS

Federal Reserve System, the FDIC, and the OCC
conduct - primarily through community affairs
programsmongoing outreach and educational ac

Livities to help banks and the public better under
stand and deal with community credit issues,

The financial services industry continues to address
the questions of how and to what extent unlawful
discrimination takes place in mortgage lending.
Industry trade associations and individual lenders
are also focusing on ways to improve access to
home mortgages for minorities and lower -income
applicants.

including discrimination. These activities fall pri
marily into four categories:

Trade Association Task Forces

• Sponsoring conferences and workshops for
and with bankers and community and business
representatives. These programs focus on issues

and opportunities concering community reinvest

Several of the national trade associations have
formed HMDA task forces to evaluate the issues of
accessibility raised by the HMDA data and have
sought to identify areas that members could

ment, community -development lending, and related

address. For example, the American BankersAsso

Expanded HMDA Daia on Residential Lending: One Year Laler
ciation ( ABA) has created a center for community
development to provide member banks with a
clearinghouse of information and products that
could help increase the availability of mortgage
loans to creditworthy minority and low- and
moderate - income applicants: Key components of
the association's plan for the center include the

817

with interpretations and applications, are based on
historical data that reflect primarily nonminority
participants in mortgage loans, and thus they may
unintentionally reflect racial bias; and ( 3) members
of minority groups may receive unequal treatment
in the prequalification and loan application stages
of the lending process.

following:

• Lender training. The center will develop
resources to help bankers conduct community
development lending and will work with govern
ment agencies , nonprofit organizations, and other

development groups; it will provide training in
preventing subtle, unintended forms of discrimina
Lion. The ABA currenuly makes available a video,
“ Fair Lending Compliance: Understanding Equal
Treatment,” developed to help lenders avoid dis
criminatory treatment of applicants.

In September 1992, the Mortgage Bankers Asso
ciation of America ( MBA) issued thc “ HMDA

Initiatives to promote Affordable Housing

Task Force Report " to heighten awareness of issues
of mortgage availability , encourage lender self

Financial institutions across the country have
developed many creative activities in partnership
with each other and in conjunction with local com
munity organizations. These efforts are assisting
banks in making credit available to their entire
communities. The following programs are repre
sentative of numerous efforts by financial institu
cons, individually and jointy, to address the need
for affordable housing:

analysis, and provide guidance on ways in which
members could increase their loan approval rates,
particularly for black and Hispanic applicants.

Three of the task force's findings were that
( 1 ) many mortgage companies are conservative in
applying the underwriting guidelines of the second
ary market and of the FHA and VA because of

concems about forced repurchase and indemni

fication of loans approved under nontraditional

• Slatewide lending consortiums developed in

guidelines; ( 2) underwriting guidelines, together

California, Washington , Nevada, and Hawaiicreate

818

Federal Reserve Bulletin

November 1992

a pool of fixed -rate, long -term morigage funds for
low- and moderate - income housing.

60-893 0 - 92 - 3

crimination in individual cases. The results of test

Expanded HMDA Data on Residential Lending: One Year Lolcr
Housing Initiatives Program . In 1993, HUD will
competitively award $ 1 million to ſund a testing
project on mortgage -lending practices. Nonprofit
organizations and other privateentities were invited
in August 1992 to submit proposals to identify

819

the institution located in nonminority, middle

income neighborhoods. Each bank branch was vis
ited separately by a white and a minority shopper.
The research firm characterized the survey results
as revealing subue forms of discrimination on the

specific unlawful discriminatory acis or practices - part of bank representatives. Differences were
that prevent or impede minorities from obtaining

financing for the purchase of homes. The project,
which will support testing in three metropolitan
areas, is expected to produce a method for testing
mortgage lending practices and for providing evi
dence of the existence or nonexistence of discrimi
natory lending based on race and national origin.
HUD also continues to issue contracts to state and

noted, for example, in the amount of information
given to minority shoppers, in the time spent with
minority shoppers, and in the effort to inform
minorities about alternative mongage products.

RECENT ACTIVITIES INVOLVING THE
SECONDARY MORTGAGE MARKET

local agencies and fair -housing groups to conduct
testing for the detection of discrimination in the
sale and rental of homes.

Institutions in the secondary mortgage market play

820

Federal Reserve Bullcuin

November 1992

7. Morgage loans soldby type of purchase,characteristics of borrower, and characteristics of census oracl

636,954

830,402

Total loans sold
Race ofborrower

66,833

1,965

603,776

American Indian

1325
5,094
26.632
21 205
231,779

1.827
1.6
8.4
6.7
79.8

29.989
10,024

23,488
422,854

60
20
4.7

84.3

.2
5.4
26
2.1
86.9

266
1.179
4,049
2508
50.957

ទីនេះ
៩-

4.3
2.3
3.9
86.6

995
65*

2447
28.958
15.917
26.294
586.907

84.5
1.9
100

1.798

Income of borrower
( percentage of MSA
median ) ?
Less than 80 .
80-99

76.994
72658
84.901
342063

100-120
More than 120

576,616

13.4
12.6
14.7
59.3
100

80344

30.7

56356

S5.147
46.383
79.698

21.1

53.575
63,632
256,472
430.035

261 372

17.7
30.5
100

351

11.268

23.9

7,780
6,943

165
14.7
44.9
100

59.6
100

1,723

20.4
14.1
135
521
100

61.3
19.1
13.5
4.2

1.009
374
225
57

59.8
22
133
3.4

33.413
9,124
6.141
1.561
1.326
S1.365

64.8
17.7
11.9
3.0

8.6
57.9
33.6
100

158
1.033

9.4
61.2
29.4
100

5,428
30.091
16.046
51.565

105
58.4
31.1
100

13.1
125
14.8

243
232
897

21.158
47.149

Recial composition of
census tract ( minorities as

percentage olpopulation )
443.069

Less than 10
10-19
20-49
SO - 79
80-100

117.099

77,932
23.068

65.9
17.4
11.6
3.4

256.691
86.672

7.6
SSS
36.9
100

SS.297
281 524
101 545
438 366

11.533
672,701

64,186
17,762
13.055
438.366

58.6
19.8
14.6
4.1

292857

126
64.2
232
100

40,902
26563

91 322

64,660
19.985

Income of census tract '
Low or modau
Midde

Upper

...

51,203
373.328
248.170

672701

1. Components may not surm lo louls because of rounding.

160.593

478,058

MSA as a whole; in middle - income census acts, median family income is
80 percent 10 120 percent of the median MSA family income; in upper.
income casus ras, median family income is more than 120 percent of the
median MSA family income

FNMA and FHLMC follow essentially the same
guidelines, which they themselves set for the con
ventional loans they purchase. For GNMA, under
writing standards are established by HUD for FHA
insured loans and the VA for VA - guaranteed loans.
Given that HUD and the VA impose less stringent
loan standards than do originators of conventional
loans, and that they have different rules about the
maximum size of loans they will back , one can
expect that, overall, FHA and VA borrowers will
differ markedly from users of conventional loans.

Consequendy, borrowers whose loans are securi
lized by GNMA are also likely to differ from those
whose loans are sold 10 or securitized by FNMA or
FHLMC.

tions ( particularly the GSEs) has long been pub
licly available, but before 1990 it was mostly of an
aggregate nature. The 1989 amendments to HMDA
fundamentally expanded the information available
about secondary market activities by requiring
lenders covered by HMDA to report, for loans
originated or purchased during a year, the loans
that they sold , classified by the type of secondary
market purchaser. The release of the 1990 HMDA
data provided the first opportunity to examine the
secondary market's patterns of loan purchase and
securitization by the characteristics of mortgage
borrowers and neighborhoods in which their homes

are located. It also allowed comparisons between
the activities of the primary market institutions and
those of the secondary market institutions along
these dimensions.

821

Expanded HMDA Data on Residential Lending: One Year Laier

7.- Continued

Borrower or census tract

Savings bank of
savings andloan association

Life insurance
Nurnbes

Toual loans sold

Affiliate of

242,197

14,129

83,017

Other purchaser
- Nurnber

Parcent

Peroen

498,775

Race ofborrower
American Indian
107
888

3
24
4.1
28
88.1

100

5,872
4,478

19.1

4.365
15.868
30,783

14.8
SIS
100

23,798
9.017
3,012

9.1

40
688
387
235

15
25.2
14.2

1.062

636

5.786
6,708
4.569

165,652

3
3.1

ក

1.502

3.6
24

88.0

2,143
16.202
23,984
22.093
347,756

.s
3.8

5.6
5.2
81.8
‫کی‬
25
100

237

2733

2472
1.532

23.1
143

1.400
10,718

13.1
49.6
100

25,283
18.390
19.059
91.296
154,028

16.4
11.9
124
593
100

77,066
50.838
44,914
168.799
341,617

226
14.9
13.1
49.4
100

72.1

7.713

65.8

133,455

68.7

272,913

15.2

2186
1,292
319

18.6
11.0
27

32339
20.310
4.946

16.6

78.143
54.136
13.827

59.0
20.7

8.2

881

58.1
33.7
100

6,313

75
53.9

15,841
106,098

145

5314

Racial composition of
census tract ( minorities as
percentage of population )
Less than 10
10-19
20-49
S0-79 .
80-100

105

14.3
3.7

Income of census tract
Low or moderne
Midde

2708
19.164

Upper .

11,124

32.996

The first comprehensive evaluation of the
HMDA dala, as it pertains to secondary market
institutions, was completed in May 1992.20 The
pattems of home loan purchase by the major enti
ties in the secondary market appear, in general, to
mirror loan origination activity in the primary mar

ket. In particular, the distribution of loan purchases
arrayed by borrower and neighborhood characteris
tics among the secondary market agencies reflects

38.6
100

8.2

54.6
373

100

37.588
206.913
133,492

377.993

9.9
54.7
35.3

100

tics of home purchase loans backed by GNMA
guarantees direcuy reflect that agency's legislated
specialization in government-backed loans. Simi
larly, the characteristics of loans acquired by
FNMA and FHLMC derive, for the most part, from
the borrower and geographic composition of con
ventional home- purchase loan originations.

closely the distribution of loan originations by
applicant and neighborhood characteristics. More
specifically, the borrower and location characteris
20. Results were presented at the Annual Housing Conference
sponsoredby FNMA in May 1992 and subsequently published. See
Glenn B. Canner and Swart A. Gabriel, “ Market Segmentation and
Lender Specialization in the Primary and Secondary Mongage
Markeus. " Housing Policy Debate, vol. 3 ( September 1992 ) .
pp. 241-329; and Frank E. Nochaft and Vanessa Perry, " Home

Mortgage Disclosure Act Data , " Secondary Morgage Markers,
vol. 8 ( Winter 1991-92 ) . pp. 2–6.

21. The ILMDA dou do not reflect all the loans purchased or
backed by secondary markel entities in a given year - only those
that were originated or purchased by a covered lender and that were
sold in the same year. The characteristics of borrowers whose loans
are not included may differ from those reported by institucions
covered by HMDA.

822

Federal Reserve Bullcun O November 1992

HMDA dala also reveal thal, compared with other

FNMA and FHLMC have responded to these

secondary market purchasers, GNMA guaranteed
relatively morc loans to borrowers purchasing
homcs in low- or moderale -income and middle

concems by continually revising their guidelines
and by emphasizing 10 lenders and others that the
rules are not absolute. Over the past five years, the

income neighborhoods.

agencies have significantly changed the guidelines

to help ensure their flexibility in teftecting the
special circumstances of lower- income borrowers
and properties in older, urban areas. In the past
year, for example, guideline changes have dealt

with the treatment of credit history ( such as a
record of slow or late payments on debes ) and the
acceptability of certain types of short-term income
( such as seasonal employment and child support
payments) as compensating factors to justify the
use of higher qualifying ratios for debt- to - income.
The use of information in appraisal reports has
been clarified. For instance, the guidelines now

clearly indicate that a loan on propery receiving a
" less than average " rating is not automatically
ineligible for sale to the secondary market.

22. Conference repon on H.R. 5334. Housing and Community
Development Ad of 1992, sections 1332 and 1333 , Congressional .
Record ( daily edicion ) , October 5, 1992, pan 5. p . H12019.

Expanded HMDA Dala on Residential Lending : One Year Laler

istics of their communities. A FNMA product

violation of HMDA is subject to administrative

called FannicMaps allows loan originators to use a

sanctions, including the imposition of civil money
penalties where applicable.

computer-accessible pictorial representation of the
income and racial characteristics of a designated

arca , such as the city in which they have offices or
groups of selected neighborhoods, on which 10

823

824

Federal Reserve Bulletin

November 1992

1990 census tract boundaries for identifying the
location of properties underlying their loans. Dis
closure statements displaying the 1992 HMDA data
will use the 1990 census information.

Data Availability
The FFIEC makes the HMDA data available in
various forms and formats . These include disclo
sure statements for individual institutions and

aggregate reports ( in paper copy, microfiche, and
computer tape) for each MSA; a set of tables

showing nationwide aggregates; and a series of
tables highlighuing key information for cach MSA.
An edited version of the LAR records for the

nation as a whole is available on data tape and will

be made available in early 1993 on PC diskelle for
each MSA separately.

Home Mortgage Disclosure Act:

Expanded Data on Residential Lending
Glenn B. Canner of the Board's Division of
Research and Statistics and Dolores S. Smith of
the Division of Consumer and Community Af
fairs prepared this article with assistance from

on loans they originated or purchased. Now , in
disclosure statements released to the public in
October 1991, lenders for the first time have
reported on all home loan applications they re

Nancy E. Bowen , Florence M. Benkovic, Sylvia

ceived and their disposition, plus the race or

A. Freeland, Cynthia H. Johnson , Thomas A.
Orndorff, and Mark R. Schultz ofthe Division of
Information Resources Management.

national origin , gender, and annual income of the
applicants. In addition, more lenders are now

860 Federal Reserve Bulletin

November 1991

laws ( the Fair Housing and Equal Credit Oppor

tunity Acts) and the Community Reinvestment
Act ( CRA ) . Because bank examiners have access
to loan application files, they will be able to
overcome most of the limitations of the HMDA
data. By using the HMDA data in conjunction
with loan application files, related information,
and other materials related to evaluating CRA
performance, the agencies will be able to carry
out their enforcement responsibilities more effec
tively .

Following the most recent amendments to
HMDA , contained in the Financial Institutions
Reform , Recovery and Enforcement Act
( FIRREA) of 1989, the data may serve a fourth
purpose: to assist in identifying possible discrim
inatory lending patterns and in enforcing anti
discrimination laws.

Recent Changes in Coverage
For more than a decade, HMDA applied only to
depository institutions commercialbanks, sav
ings banks, savings and loan associations, and
credit unions and their subsidiaries. Among
that group, only those with assets exceeding $ 10
million and a home or branch office in a metro
politan statistical area ( MSA) have been cov
ered.

HMDA'S PURPOSE : IDENTIFICATION OF
HOME LENDING PATTERNS IN
URBAN AREAS
The Congress passed the Home Mortgage Dis

closure Act in 1975 in response to concerns that,
by failing to provide adequate home financing to

qualified applicants on reasonable terms and con
ditions, some depository institutions “ have
sometimes contributed to the decline of certain
geographic areas." The law was intended to
provide information about residential lending ac
tivity that could be used on several fronts:
• Generally, the data could help determine
whether financial institutions are serving the

housing needs of the communities in which they
are located, by identifying pockets in which they
are and are not providing credit.

1. AD MSA typically consists of a central city having a
population of 50,000 or more , the county in which the city is
located, and any surrounding counties that are tied econom .
ically and socially to the central city .

Home Mortgage Disclosure Act: Expanded Data on Residential Lending 861
1. Residencial lending activity reported by financial
Numbo
Number of

Number of
metropolitan

8.09
8.258
8.050

10.945
11.357
10.970

Year

1981
1982
1983
1984
1985
1986
1987
1988
1989
1990

Loans on properties outside the MSA were
grouped to show the total number and the dollar
value of such loans by type of loan.

1.28
1.13
1.71
1.86
1.98
2.83
3.12
3.39

3.13
6.37

8,491

11.799

9.002
8.898

12.567
12.329
13.033
13.919
14.154
23.891

9.431

9.319
9.203
9.281

1. Except for 1990. includes outy loans originated by covered insccutioas:
for 1990 ( first year under revised reporting system ) . includes looss origio

1990 Data : Disclosure Procedures and
Scope ofInformation

ated and purchased applications approved but not morepiedby the applicant.
applications denied or withdrawe, bed applicacions closed because informan

With the 1989 FIRREA amendments , institutions

dion was incomplete.

must continue to disclose information about res .
idential loans extended and purchased and also
must report on applications that did not result in

an extension of credit. They are also making
the lending activity reported by roughly 15 per
cent .

Pre- 1990 Data : Focus on
Geography of Lending

public for the first time information about loan
applicants — their race or national origin , gender,
and annual income. Further, for loans originated
or purchased during the year, institutions must

report the loans they sold, classified by type of
secondary market purchaser. Finally, they may ,

Through 1989, lenders reported only their origi
nations and purchases of home purchase and
home improvement loans, under conventional

if they wish, report their reasons for denying
loans.

and government-backed lending programs ( those

Loan Application Register. The Federal Re

insured or guaranteed by the Federal Housing
Administration ( FHA ) , the Veterans Administra
tion ( VA) , or the Farmers Home Administration
( FmHA) ) . Lenders prepared two reports for each
MSA in which they had offices one for loans
originated and the other for loans purchased
during the calendar year.

serve Board is charged with implementing the

HMDA amendments. The Board's approach to
collecting the data ( developed in consultation
with the other supervisory agencies “ ) is a rela
tively simple one that minimizes the burden on

the reporting institutions and, at the same time,
provides a reporting format that offers a large
base of information for use by the public and the
2. The FFIEC is composed of represcatatives of the
Federal Reserve Board , the Federal Deposit lasurance Core
poration, the Ofice of the Comptroller of the Currency, the
Office of Thrift Supervision , and the National Credit Union
Administration .

862 Federal Reserve Bulletin November 1991
supervisory agencies. Covered institutions
record data for each loan application acted on
and each loan purchased on a separate line of a

report for an MSAmay contain as many as
thirty -three tables. The first thirty -one are an

aggregate version of the individual institution

reporting form , the Loan / Application Register

disclosure tables. The other two show the dispo

( LAR ) . At the end of the year, the institutions
submit the LARs to their respective supervisory
agencies, which send them to the Federal Re.
serve Board for processing. The Board , acting on
behalf of the FFIEC , produces disclosure state

sition of loan applications by median age of
homes in census tracts in the MSA and by the
central city or non -central city location of the
property .

ments and sends them to the reporting institu
tions for release to the public. Under this system ,
institutions collect the required information but
do not have to undertake the additional costly
step of preparing their own disclosure state
ments , which would involve sorting and aggre
gating their data in multiple cross -tabulations.

Disclosure Statements and Aggregate MSA
Reports. The disclosure statements made avail
able to the public consist of a series of tables. An
individual institution's statement may consist of
as many as thirty -one tables for each MSA in
which it has offices. The tables show the follow
ing:

• Disposition of loan applications, by type of

loan and geographic location of the property ( in
most instances the census tract number)

Scope and Volume of Disclosures. However
measured, the 1990 effort to collect and process
the data has been immense. The disclosure re
ports contain data on nearly 6.4 million loan and
application records. At the Federal Reserve, the
volume of HMDA data processed on behalf of
the FFIEC this year was greater than that for any
other single subject handled by the System . To
put the effort in context, the amount of data

processed was roughly eleven times the quantity
of HMDA data handled prior to the 1989 amend
ments. Moreover, given the relatively weak
housing market in many sections of the country
through most of 1990 , the volume of loan activity
reported can be expected to be significantly
greater in subsequent years.

6. The federal supervisory agencies incurred an estimated

The disclosure statement is available to the pub
lic at the lender's home office and at one branch
office in each other MSA in which the lender has
a branch. Copies of the disclosure statements for
all lenders in an MSA also are available to the
public at the central data depository in that MSA .
In addition , the FFIEC compiles and provides

one -time cost of $ 2.8 million to develop the system for

processing the expanded HMDA data ( primarily for com.
puter software developmcat) . The agencies have spent és
proximately $ 2.6 million to process the 1990 daia . The annual
processingcost is cxpected to decline in future years as more

institutions submit the data in machine-readable form . De
spite a comprehensiveeffort to identify errors in the data and
have them corrected, at the time the disclosure statemen's
were distributed to the public the agencies were aware that

about 4 percent of the LAR records contained errors . In
addition, a number of institutions have contacted the FFJEC
during the thirty -day review period with questions abou! :
completeness of their reports.

Home Morgage Disclosure Acı: Expanded Data on Residential Lending 863
2. Financial institutions covered by HMDA, by cumber

increase public access to the data. These librar.
ies - of which there are some 1,400 across the

nation - are repositories for a wide range of doc
uments and data produced by federal govern
80.6
8.6

ment agencies.

3.0
1.8
S
6-9
10-19
20-49
SO or more

2.1

100

MEMO

Totalsumber of financial institucions
Total number of MSA reports ..

9.281
23.891

1. Components do not sum to local because of rounding.

For lending activity in 1990, the FFIEC dis
tributed disclosure statements to 9,281 reporting
institutions , consisting of 23,891 individual MSA
reports ( table 2, memo item) . Disclosure state
ments for the vast majority of institutions ( 81
percent) covered a single MSA; for roughly 275
lenders, the reports encompassed ten or more
MSAs.

PRE - 1990 STUDIES: FINDINGS AND
DATA LIMITATIONS
HMDA data have long been the primary source
of public information about the geographic dis
tribution of home loans originated and purchased

by financial institutions. Dozens of studies have
examined the distribution of home loans across
neighborhoods stratified by residents' income
and race .

Efforts to Facilitate Public Access
In paper form , the HMDA data can be awkward
to use and costly to duplicate. Consequently, the

FFIEC is exploring ways to distribute the data in
forms that reduce the volume of paper and facil
itate public use , including microfiche, PC disk
ette, and CD -ROM discs. The FFIEC also is
investigating the possible use of the govern
ment's Federal Depository Library System to

7. To help ensure the confidentiality of loan applicants, the
edited version of the LAR excludes three reported items: the
loan identification number, the date of application, and the
date action was taken on the application .

864 Federal Reserve Bulletin

November 1991

activity across neighborhoods within local communities when the neighborhoods are grouped by

median family income or racial composition.
Although these differences in lending activity
vary greatly among different institutions, depending on their specific circumstances, overall
the HMDA data show that a smaller proportion

of home purchase loans made by reporting lenders are for properties in low- or moderate -income
neighborhoods ( those where median family in-

come is less than 80 percent of the median family
income of their MSA) . Although the proportion
varies somewhat from year to year , since 1985 it

more home improvement loans per single family
housing unit in minority neighborhoods than in
similar-income predominantly white areas.
Although the statistical disparities cited in

these studies clearly exist, opinions on the rea
sons for the differences vary widely. Some peo
ple believe racial discrimination by commercial

banks and thrift institutions is a contributing, if
not the primary, source of these patterns. Others
suggest that the patterns reflect fundamental dif
ferences in the economic circumstances of pop

ulation groups ( whether already living in or seek
ing to reside in the different areas) and in market

generally has been between 10 and 12 percent of

specialization by different types of lending insti

all the home purchase loans granted inMSAs. In
comparison , roughly one -third of the home purchase loans are for properties in upper-income
neighborhoods ( those where median family income exceeds 120 percent of the median family
income of their MSA) .. The remainder are for

tutions.

properties in middle-income neighborhoods.

Consider, for example, the analyses that focus
on the level of home lending per housing unit in
seemingly similar minority and nonminority
neighborhoods. An assumption underlying these

analyses is that by selecting neighborhoods that
have certain similarities in aggregate characteris

Home Mortgage Disclosure Act: Expanded Data on Residential Lending 865

In Atlanta, another factor that appeared to
reduce demand for home purchase loans from

depository institutions covered by HMDA was a
much heavier reliance on government-backed

forms of credit in the minority middle-income
neighborhoods than in the predominantly white
areas. Mortgage bankers, most ofwhich were not

then covered by HMDA, are much more likely to
be the source of such credit. Nationwide, they
extend roughly 80 percent of FHA and VA loans.
Thus, the use of government-backed loans by
home buyers in the minority community in effect

reduced demand for credit offered by lenders
covered by HMDA .

The findings about FHA financing patterns are
consistent with the results of two recent studies
that were based on nationwide consumer sur
veys. The first found that black and Hispanic.
purchasers of moderately priced homes are
roughly 70 percent more likely to use FHA
insured loans than are similarly situated white
home buyers. Although all the reasons for these
differing usage patterns are not clear, they may

reflect differences in loan product recommenda
tions made by real estate agents, self steering by
loan applicants, or differences in marketing ef
forts by lenders.16

866 Federal Reserve Bulletin November 1991

3. Disposition of applications for home loans, by purpose and type of loan . 1990'
Number. in thousands, and percenuse distribucco

Conveaconal

Loon onprated ...

484.2

68.6

103.6

SS.O
15.1
20.0
9.2

70.3

Application approved but not accepled by applicant
Application denied
Applicanor withdran
Fie closed ( informatica incompless )

22.1
18.4

1.0

1. Components may not sum vo toals because of rounding.

Nurnber

Percent

1.565.5

68.6

100

Soura . Preliminary date. Hows Mortgage Disclosure As , Board of
Governors of the Federal Reserve Systern.

race of residents.16 This study used title lien
records to gather information about lenders and

the geographic distribution of their loans. As in
the other studies, the researchers did not have
information about the prospective home buyers
and how their applications were treated by lend
ers. The study sought to determine whether
differences in economic and other nonracial char
acteristics ( primarily neighborhood characteris
tics) as reported in census data might account for
the disparities. The researchers found that, after
controlling for a wide variety of neighborhood
factors, predominantly minority neighborhoods
in Boston had been granted 24 percent fewer
mortgage loans per housing unit than predomi
nantly white areas. They concluded from this
evidence that race may have been a factor in the
lending patterns. They also indicated, however,
that from their data it was not possible to deter
mine with certainty the causes of the observed
differences in lending.

keting and community outreach efforts, and in
some cases to establish or join with others in
offering or participating in special lending pro

grams to expand affordable housing opportuni
ties .

SOME PRELIMINARY FINDINGS

FROM THE 1990 DATA
Because the 1990 HMDA data have just been
released, little is yet known about what the
expanded data may reveal once they are thor
oughly analyzed. This section takes a first look at
some loan and application patterns discernible
from the data. Myriad levels of analyses are
possible, particularly with respect to different
geographic areas and different groupings of finan
cial institutions. The focus here is on nationwide
totals and on some potential uses of the new and
expanded data . In reviewing the nationwide data,

it should be noted that the lending records of
individual institutions may vary greatly, both
from one another and from pattems for the
nation as a whole , depending on their location ,
the types of applicants they serve , the types of
loan products they offer, and their credit stan
dards.

16. See Katharine L. Bradbury , Karl E. Case , and Con
stance R. Dunham, “ Geographic Patterns of Mortgage Lend
ing in Boston, 1982-1987," New England Economic Review
( September/October 1989) , pp. 3-30 .

Home Mortgage Disclosure Act: Expanded Data on Residential Lending 867

3. - Continued

Las on one to four- family dwellings
Loans on
multifamily dwellings

Disposition

Rehnancing
Nezba

Loua originated
Application approved but not acceprod by applicant.
Application denied
Application withdrawa

Homs

improvement

691.1

Perceuk

Number

Perceert

Number

Percent

67.5

716.1

65.1

27.2

61.5

4.6

5.3

File closed ( information incomplete )
100

Volume of Applications
In 1990, lenders covered by HMDA took action
on roughly 3.26 million home loan applications
3.09 million for purchase, 1.02 million for refi
nancing, and 1.10 million for improvement of
residences housing one to four families, and the
balance for loans on multifamily dwellings for

five or more families ( table 3) .17 Among home
purchase loan applications, 74 percent were for
conventional mortgage loans, and the remainder
were for government-backed forms of creditFHA, VA, and FmHA loans.
Use of Various

Home Purchase Loan Products

1.100.7

100

The new data also indicate that black ( and to a
much lesser extent Hispanic) applicants are more
likely than either white or Asian applicants to
seek government-backed home purchase loans.18
Blacks in particular are relatively more likely to
seek FHA and VA loans: Blacks constituted 4.3
percent of all applicants for conventional home
purchase loans in 1990, but they accounted for
10.5 percent of all applicants for FHA loans and
11.7 percent of all applicants for VA credit
( detailed data not shown in tables) . Viewed in
another way , 46 percent of all black home loan
applicants applied for either an FHA or a VA
loan, while only 28.6 of Hispanic applicants, 24.4
percent of white applicants, and 10.2 percent of
Asian applicants sought such loans.

Application patterns for various kinds of home
purchase loans differ according to applicant in
come. Government-backed loans are much more
likely to be used by households with relatively
low incomes than by households with high in
comes. The 1990 HMDA data indicate that 39

percent of applicants with low incomes ( less than
80 percent of the median family income for their
MSA) applied for government-backed home pur
chase loans, compared with only 15.6 percent of

applicants with high incomes ( more than 120

percent of the median family income for their
MSA ) .

18. Data compiled by the U.S. Census Bureau differenti.
17. Covered institutions also reported data for 1.1 million
loans they purchased during 1990 .

ate between white Hispanics and non -white Hispanics. In

the HMDA data, both are included inthe Hispanic

868 Federal Reserve Bulletin

November 1991

Overall Approval Rates

whites.20 The 1990 HMDA data reveal a similar
pattern for all lenders covered by HMDA.

Lenders approved the majority of home purchase
loan applications they received — roughly 72.3
percent of applications for conventional loans
and 71.7 percent of applications for government
backed loans ( table 3) .19 Among the applications
for conventional loans, 16.6 percent were denied
by the lender and 10.2 percent were withdrawn
by the consumer ; in a relatively small number of
cases ( less than 1 percent) the application file was
closed after the applicant was asked for but failed
to submit information required for the credit
decision. For government-backed home pur

Conventional Home Purchase Loans. Nation
ally, about 14.4 percent of white applicants for
conventional home purchase loans were denied
credit in 1990. In sharp contrast, the rate for
black applicants was 33.9 and for Hispanics 21.4
percent ( tables 4 and 5) .21 At 12.9 percent, the
denial rate for applicants of Asian extraction was
lower than for any other racial or ethnic group.

chase loans, the denial rate was 16.5 percent and
the withdrawal rate 10.6 percent.

Approval Rates for Minorities
Although the majority of home purchase loan
applications are approved, many are not. Ap
proval rates vary according to the applicant's
income and demographic characteristics and the
characteristics of the area in which the applicant
resides or seeks to purchase a home .

20. Office of Thrift Supervision, Data Submission Reports,
selected years. These reports contain information on the
disposition of mortgage applications filed with savings and
loan associations. The data, which have been collected for
more than ten years, include information on the race or
national origin of the applicants.

19. Among loans approved, in a relatively small proportion
of cases the consumer did not take out the loan , perhaps
because the property sale did not go through or because the
consumer filed applications with more than one lender and
accepted the most attractive offer .

Home Mortgage Disclosure Act: Expanded Data on Residential Lending 869

greater when comparisons are made at the ex

4. Number of home loan applications, bypurpose of loan, characteristics of applicant, and characteristics of census
Home parchase
Applicant or census rct characterisdic

Goverranea -becked '
Race of applicant
American Indian / Alaskan masive
Asia Pacific Islander..

Black
Hispanic
Wuste
Other
Joiss ( white/minority ) ,

3.281
10.721
76.983
*4.45
561.735

Refinancing

Home improvement

1.960
39.897
42.669
61.622
760.490

5.727
16.968
74.106
40.232
679.292

174.962
120.701
680.605

199.944
155.212
596.803

101.720
70.973
79.494

240.042
103.061

Conventional
11.320
9.264
90.414
110.602

1.733.522

Gender of applicans
Male '

Female '
Jour ( male / fernale) .
Income of applicor ( percentage of
MSA median )
Less than 80
80-99
100-120 ..
More than 120

146.277
109.375
478.079

420.667

192.214

238.461
154.421

113.509

$ 33.143

Racial composinon of census tract ( winorines
as peremege of population )
Las than 10
10-19 ..
20-49
SO - 79

80-100

318.464
106.831
07.125

1.010.345

25.171
21.534

421.329
162.894
131.275
$ 3.470
41.447

484.935

120.556
100.650
43.353
57.016

brcome of casus tracta
Low or moderate

Middk .

Upper
Income of consument and racial composicion
( muronnas as percentage ofpopulation '

81,443
354.883
122.579

204,107
931,665

115.763

129.581

$ 20.2006

149.578
245.074

44.459
192.470

49.906
26.059
51.835
37.477
38,830

21.387
12.602
26.918
24.641

29,742
13.427
25.121
20.575
40.716

585.7os
169.225
131.20
34,070

246.019
90.095
79.598
24.539
9.377

316.852

153.923
60.197
24,799
4.290
1.905

138.341
34.213
15.441
J.177
1.298

Low or moderne

20.JSO
13,617
21.247
11.939
14.310

30.215

Middle

213.219
68.859
53.42
12.114
6.849

11.30

72.916
60.088
19.601
15.002

Upper
Less than 10
10-19

20-19
50-79
0-100

14.895
24.355
12.036
1.098

1. Loans backed by the Federal Housing Adeninistration, the Veterans
Administration, and the Farmers Home Adminstrado .

374,734
98.568
39.376
6,122
1.946

family income is less than 20 percent of the medias family incorse of the
10 percent to 120 percent of the median MSA fornity we come: in upper-ncome
census tract , nedisa family income is more than120 percent of the methan

MSA as I whole: in middle-icone crisus trucs, medius famaly vecom is
MSA family income

75
870 Federal Reserve Bulletin o November 1991

5. Disposition of bome loan applications, by purpose of loan and characteristics of applicant, 1990 '
Hows puretese
Governancar -becked '

Convensonal

Applicant
chanderistic

With

Fik
clored

Total

Approved

Deaiad

100
100

66.0
72.7

100

SS.7
65.1
75.5
63.2

22.
12.9
33.9
21.6

Approved

Danied

63.5
74.8
60.9
68.7
17.4
66.3
75.6

2.S
12.8
26.3
18.4
12.1
18.4
14.1

12.8
11.6
11.3
11.6

71.6
74.7
75.0

14.9
14.5
14.3

12.1
9.9
9.3

1.3
.9
.9

100
100
100

72.0
77.9
79.1
79.7

18.1
13.0
11.3
10.4

8.9

1.0

100
100
100
100

Wida .
drawn

File
closed

Toul

Race
Americana Loding

1.2
9
1.5
1.3

100
100
100
100

10.6
13.5

1.0

100
100

1.1
1.0
.7
.8

100
100
100
100

73.3

19.0
14.9

68.1
69.8
75.3

20.0
19.9
14.2

10.9
9.5
9.8

1.0
.7

100
100
100

6.5
75.5

26.0
15.7
12.9

100

79.0

9.9

7.7
8.2
8.5
10.4

73.0

100

Income percentage of
MSA median ) '
Less than 80
80-99
100-120
More than 120

8.6
9.2

I. Components may nor sum lo locals because of rounding.

100
100

S. MSA media is meeting facsiły income of the metropolias sexoscical area
in which the property related to the loca is locaud.

purchase loans, 14 percent had incomes below 80
percent of their MSA's median family income.
Low - income black and Hispanic applicants, in
contrast, accounted for 25 percent and 16 percent
of all applicants in their respective groups. Low
income Asians accounted for only 8 percent of
the conventional home purchase loan applica
tions filed by Asians overall.

higher for both Hispanic and Asian applicants,
12.4 percent and 13.5 percent respectively.23

Government-Backed Home Purchase Loans.
The patter for denial of government-backed
23. Home purchase loan applications are withdrawn for a
variety of reasons. For example, prospective bome buyers
who ile a loan application may not be able to complete a
purchase because of ao inability to sell their own home. The
1990 HMDA data will enable supervisory agencies, which
will have access to loan application files, to investigate
differences in withdrawal rates across different gender and

racial or national-origin groups for evidence of unfair treat
ment .

Home Mortgage Disclosure Act: Expanded Data on Residential Lending 871

5. - Continued

Renancing

Home improvement

Applicans
charactersec

File
closed

Rece
American Indian
67.7
66.0
61.1
61.9
74.4
60.7
71.1

17.9
17.3

25.1
21.6
14.3

23.2
16.4

19.1
17.7
14.7

67.7
71.9
73.6

21.1
17.4
1S.S

13.3
12.0

10.9

1.2
1.0
.9

Approved

100
100
100
100
100
100
100

73.8
65.0
58.1
60.2
78.1
57.3
75.4

36.9
32.5
17.0
34.1

100

Denied

21.6
24.6

19.3

Widt

9.1

a tài 0 .

13.0

10.3
10.0
10.1
11.7

home purchase loans is similar to that for con
ventional home purchase loans. The rates of
denial were 26.3 percent for blacks, 18.4 percent

Toul

100
100

6.4

100
100

7.8
4.9

100
100
100
100
100

67.3

27.0

5.0

100

66.0

100

74.9

28.2
19.1

5.2
5.4

100
100

62.7
70.0

32.4
24.8

4.5

100
100

13.4

21.4

4.8

76.7

17.0

5.6

tacoma ( percentage of
MSA mediani'
80-99
100-120
More than 120

File
closed

8888

73.6

66.5

13.6
15.2
12.6
15.4
10.5
IS .
11.8

Tood

8888888

Wideo
dnma

‫ܝܚܪܘܩܪܝܼܪܗܿ܆‬

Danied

,7878887
777

Approved

approved. Males were somewhat more likely
than females to have a home improvement loan
approved.

for Hispanics, and 12.8 percent for Asians, com
application withdrawal were 11.3 percent for

Relation ofApproval Rates to
Neighborhood Income and Composition

blacks, 11.6 percent for both Hispanics and
Asians, and 9.7 percent for whites.

The HMDA data make it possible to compare

pared with 12.1 percent for whites. The rates of

lending across neighborhoods grouped by racial
makeup and the income level of their residents.

Considerable caution should be exercised , how
ever, when making such comparisons. The use
fulness of these data is currently limited by the
lack of an up -to - date match with the characteris
tics of census tracts . The recently released
HMDA disclosure statements are based on 1980
Home Improvement Loans. The patterns for

census tract boundaries and population charac

denial and withdrawal of home improvement

teristics ( neighborhood income level, racial com

loan applications are broadly similar to those for
home purchase loan applications. Generally , for
all groups the denial rates are higher than for
home purchase loans, and the withdrawal rates
lower; 36.9 percent of black, 32.5 percent of
Hispanic, and 24.6 percent of Asian applicants
were denied loans, compared with 17 percent of

position, and housing stock characteristics) . This

white applicants.

census information is now more than ten years
old, and in some cases the resulting figures may
be misleading. For example, a low -income, pre

dominantly minority neighborhood in 1980 may
have undergone substantial change and may now
have a much higher average income and a differ
ent racial composition . The Federal Reserve
Board has published proposed amendments to
HMDA reporting requirements, calling for a
switch to the 1990 census tract definitions begin

872

November 1991

Federal Reserve Bulletin

6. Disposition of home loan applications, by purpose of loan and income and race of applicant, 1990 '
Percenage distribucion

income' wad race
Wich

File
cloved

Tool

drawa

File
closed

.9
1.0

Denicd

Toul

Less than 80
American Indian
9.2
10.3

29.4

10.7
9.8

67.7
74.1

70.2
78 .
64.5
72.2
810
72.0
78.2

17.8
12.7
24.8
17.0
10.6
D.S
13.0

68.0
78.1
65.7
73.9
81.9
69.6
77.6

17.0
12.4
23.1
14.7

76.5

71.3
76.0
68.0
72.4

13.6

11.2
20.8
14.2

.9
.9

62.7
68.4

27.7
17.2

8.8
13.4

100

$ 1.4

40.1

7.6

100

58.1
69.0
64.S
64.8

31.1
23.1
26.1
26.3

9.8
7.2

100
100
100
100

100
100
100

73.3
67.7

100
100

100

78.1
70.6
72.2

16.6
13.7
29.3
21.5
13.7
21.1
18.0

9.4
10.5

100

12.7
11.S

8.0
10.2
1.0

11.1

73.1
60.8

9.5
9.9
7.7
13.0
8.2

1.2

13.6

1.5

100
100

72.6
75.0

14.0
12.6

1.1
1.1

100

63.8
69.6

26.3

12.3
12.0
10.3
12.4

1.6
.6

100
100

.7
1.0

100
100

72.1
75.8

.7
.8
1.0
1.0

100
100
100
100
100
100

74.4
75.2
65.7
71.1
81.2

67.3
79.2

1. Components may not sum o boals because of ronding

2.Applicanincome shows us percentage ofthemedisetasnilyincome of
the metropolitansaasocal area in which the property related to the house is
located.

ning January 1992. The FFIEC plans to reflect
socioeconomic information about these areas in
the disclosure tables portraying 1992 lending
activity , which will be released in 1993.
Approval of Home Purchase Loan Applica

tions. Although the majority of applications for
home purchase loans are approved, experience

differs across neighborhoods grouped by racial
composition and the income levels of their resi
dents. The patterns of loan acceptance and denial

do not differ greatly whether the type of home
purchase loan sought is conventional or govem
ment-backed .

100
100

100

80.4

71.0
77.6

19.1
11.2
18.0
19.0

12.8
11.2
21.4
15.8

8.0

.7
1.0
.6
avaiovo
oo

26.5
13.9
2.4
14.7
21.3
17.3

63.5
75.0
$ 8.5
66.5

9.2
8.6

11.9

100
100
100
100
100
100

100

100
100
100
100
100
100
100

100

12.9

11.6
12.2
9.7
12.5

1.3
1.0
.6

11.3

100
100

100
100
100
100
100
100

3. Lonas becked by the Federal Housing Administracion, the Veterans
Administracion, and the Farmers Home Administracion

Home Mortgage Disclosure Act: Expanded Data on Residential Lending 873
6. - Continued

Home improvement

Retoureng

American Ladies

Ota

$ 1.6

22.1
22.6
31.3
27.5
18.5
33.0

Joint ( white/minority)

65.6

22.9

10.8

67.2
66.6

18.3
18.8
27.3
22.1

13.0
13.5
12.2
13.5

$ 6.2

56.3
$ 7.0

72.1

1.0

Approved

Deared

Wid
dro

100
100
100
100
100
100
100

69.6
57.0
52.0
55.7
72.5
50.5
70.8

26.9
36.1
43.3

3.4
6.3
4.2
4.5

100
100
100
100

74.1
60.5
$ 7.9
60.9
77.6
59.6

1.3
1.0

10.9
14.6
1.6

1.1

38.9
23.4

FD
closed

3.9

3.0

6.1

35.7

3.5

76.8
58.6
72.3

12.3
11.6
11.9

13.5

100
100

100
100
100

100

16.9

74.4

77.9
65.8
62.3
61.7
80.1
63.6
76.3

2.3

30.5

7.9
4.5
5.6

32.3
18.0
33.0
21.2

18.1
25.9
32.6
31.0

100
100
100
100
100

7.0

100
100

2.6

1.4

7.6

ខិឱខិ

17.9

100
100

22.7
37.0

6.5

19.6

‫مزذومةبةيا‬

66.3

16.1
24.6
19.5
13.6
28.4

noin
ao

69.8
62.3

100

7878877

15.0
27.2
18.7

100-120
Amencas Indias

71.3

91.9.

Otes

Jour ( white/minority)

62.9
75.7
59.8
68.6

1.6
1.2

‫بمينن‬
‫هييابة‬

20.99
Amenean ladies
58.9

Toul

‫امام‬

13.3
20.3

Las thon 20

Tool

7887887

6.5

File
closed

‫مية‬

Denied

8888888

Widt
drom

Approved

8888888

Applicant
incontre and me

31.5
19.3

100
100
100
100
100
100

100

More than 120
American Loctinn

17.3

12.6
14.5
10.7
15.4
11.4

Oda

61.2

19.1
12.1
22.7

Joirs ( white/minonty )

72.8

15.0

75.7

100
100

13.7

17.3

23.1

1.2
1.1

than 10 percent minority residents and rises to

about 24 percent for areas with 80 percent or
more minority residents. The patter of loan
denial for government-backed loans is virtually
the same as that for conventional loans.

100
100
100
100

100

79.0
63.0
66.1
65.6
02.0
62.9

26.3

9.6

78.3

16.2

100

5.8
21.4

9.1

28.0

5.2

26.0
12.7

7.3

ๆ• ๆ

68.2
67.3
63.1
64.8

1.1

100
100
100
100
100

100

Approval ofHome Improvement Loan Appli
cations. Like home purchase loans, the majority
of home improvement loan applications are ap

proved regardless of neighborhood income or

874

Federal Reserve Bulletin

November 1991

7. Disposition of home loan applications, by purpose of loan and characteristics of census tract in which property is
Home purchase
Governmen -backed '

Convennool

Census tract

charactersuc
Wide
drava

File
closed

11.2
13.4
16.1
21.1
23.2

9.9
11.1
11.6
11.6

1.1
1.2
1.3
1.7

100

69.9
77.1
78.2

17.8
13.0
11 ..

11.1
8.9
9.5

1.3
1.0
1.1

100
100
100

67.2
78.7

20.2
13.9
9.7

75.3
72.6
70.8
65.8
61.8

14.0
14.9
17.3
20.6
24.2

9.8
113
10.7
12.2
12.3

1.0
1.2
1.2
1.4
1.7

100
100
100
100
100

71.9
69.3
67.3
65.4
61.2

17.8
18.9
19.4
21.2
24 .

79.8
76.0
71.9
WS.S
66.3

11.3
13.5
15.8
22.2
21.5

8.0
9.5
11.0
11.0
10.7

9
1.0
1.2
1.3
1.S

100
100
100
100
100

78.6
72.7
70.4

79.7
76.5
71.9
73.0
74.1

10.3
12.0
14.9
13.9
17.1

8.9
10.1
11.9
12.1

1.0
1.3
1.4
1.0

100
100
100
100
100

80.9
73.6
72.7
72.3

Approved

Denied

79.5
75.6
71.7
66.0
63.5

Recial companion
Imironnes as percentage of
population )
Less than 10
10-19

20_19
50-79
80-100

Tocal

Approved

Denied

100
100
100
100

79.1

72.7

11.5
13.8
16.5
19.3
24.0

70.1

67.5
62.1

Wich
drowo

Fik
closed

8.7

12.7

12.6
12.3

Toul

100
100

100
100
100

12.5

Midde .
Upper
Income' and racial
composition ( motorities as

percorage ofpopulanon )
Low or moderate

75.8

11.6

11.0

1.0
.7
.7
woooo

Low or moderate

AJNO

Income

12.5
12.9

1.0
1.5

12.7
14.3
16.3
18.1
23.7

8.0
12.0
12.5
12.1
11.3

.7
.8

8.8
11.3
13.0
15.3
16.8

9.6
14.3

100
100
100

100
100
100
100
100

Middle

68.9
63.8

1.0
1.2

100
100
100
100
100

Upper

1. Components may nor sum to looks because of rounding .
2. Loans bected by the Federal Housing Administration , the Veterans
Administration, and the Formers Home Adorinistracion

71.0

13 .

11.9
11.0

100
100
.8

100

.S
1.2

100

median family income is 80 perceox to 120 percent of the meetins MSA family
income: in upper-icons census tracts, median family incoex is more then

120percentof the median MSA family income .

nently, the data indicate that black and Hispanic

applicants are denied home loans more fre
quently than are white or Asian applicants who
have similar incomes. The data also indicate that
applicants seeking to purchase homes in low- or
moderate -income neighborhoods ( regardless of
the race of the residents) are denied credit more
frequently than are applicants seeking to buy
homes in upper-income neighborhoods.

Home Mortgage Disclosure Act: Expanded Data on Residential Lending 875

7.- Continued

Refinancing

Home improvement

Census tract
charactensuc

Approved

Denied

75.2
69.8
66.9

14.0
16.8
18.3
20.0
2.0

With

File

drawa

closed

9.9
13.4
13.8
14.6

.9
1.0
1.0
1.0
1.2

Approved

Dewed

100
100
100
100
100

76.6
68.9
64.7
58.1
50.3

23.9
28.1
34.9
43.3

Tocal

Wich
drawa

File
closed

Total

1.0
.9
.7

100
100
100
100
100

Racial composinon
( Mironnes as percentage

of populonon )
Less than 10
10-19
20-19
50-79
80-100

61.5

15.3

18.5

6.2
6.2
6.1
5.7

Income
Low or moderate
Middle

Upper

13.6
10.9

1.2

15.8

100
100

58.3
72.7

21.9

14.5

12.7

1.0

100

75.7

17.9

71.2
67.5
64.7
62.7
60.3

17.9
19.1
20.5
21.4
22.4

9.9
12.3
13.6
14.7
15.9

1.1
1.1
1.2
1.3

100
100
100

100
100

67.9
64.0
61.2
S5.9
48.8

27.2
30.1
32.4
37.5
44.7

4.S
5.4
5.6
5.8
9.7

76.2
69.6
67.5
65.1
63.5

14.1
16.8
18.0
19.2
21.8

8.9
12.6
13.5
14.6
13.8

100

76.7

18.7
24.0
27.4
33.7
41.1

4.2
5.8
6.2
6.4
3.7

74.1
63.0
67.3
69.5
70.4

13.4
16.3
16.9
16.1
15.0

11.4
14.7
14.8
13.6
14.1

16.1
21.4
23.9
28.0
24.3

5.1
7.5
7.6
7.0

64.7
72.5
71.8

20.6

35.6

5.4
4.9
5.8

100
100
100

Income' and racial
composition ( muronnes as

percentage of populanon )

100
100

home loan applicants and the adequacy of the
collateral provided by the properties they seek to
purchase or improve. Thus, caution in interpret

ing the numbers is called for. For example,
although the expanded HMDA data show loans
denied by race or national origin , that informa
tion alone does not provide a basis for an inde
pendent assessment of whether an applicant who
was denied credit was in fact creditworthy. Sim
ilarly, the HMDA data do not establish whether
the property involved in the proposed credit
extension was appropriately valued. Thus, it is
not possible to determine, from the HMDA data
alone, whether loan applicants are being treated
fairly and on a racially nondiscriminatory basis .

100
100

78.3
69.9

100

67.2

100
100

67.1

70.3

inn

1.0
1.0
1.0

oooo
ๆ

Upper

65.4
58.9
52.6

100
100
100
100

100

sonini

.8
1.0
1.0
1.0

Middle

GOOOOO

Low or moderne

100

.3
9
1.0
1.0
.7

100
100
100
100

.s
1.2 .

100
100

100
.9
.6

100

across the country. That process seeks to ensure
that individuals granted credit will repay their
debt as scheduled and that, should they fail to do
so , the collateral offered as security will pay off
the loan plus costs associated with foreclosure .
Consequently , lenders evaluate the factors that
they believe allow them to predict an applicant's
ability to repay ; among these factors are several
consumer financial characteristics the propor
tion of the consumer's income that will need to
be dedicated to the repayment of the proposed
loan plus other outstanding debts, the level of
equity ( through the downpayment) that the con

sumer is able and willing to put into the property ,
the consumer's employment experience and
prospects, and the consumer's history of repay .
ing debts. Lenders also consider the appraised
value of the property serving as the collateral for
the loan .

876 Federal Reserve Bulletin

November 1991

cial characteristics of loan applicants- only their
annual income . Even here , two applicants who
have similar incomes may be strikingly different
in their asset levels, existing debt burdens, and
credit histories. Applicants of different race and
gender may differ systematically in their financial
characteristics. Other sources of information,

different from those of whites.26 For instance , in
1996 the mean amount of financial assets held by
black families was $ 5,900, compared with
$ 64,000 for white families. Differences in net
worth were even more pronounced, with black
families having an average net worth of $ 29,000
and white families $ 165,000 .

such as consumer surveys conducted by the

Federal Reserve, provide extensive data on the
financial situations of households grouped, for
example, by annual income, race, or gender.
Here , too, caution is called for, however. Con
sumer surveys generally represent a wider pop
ulation of respondents than do the HMDA data,
which represent only individuals who have ap

plied for a home loan . To the extent that group
profiles developed from these surveys reflect the
characteristics of home loan applicants , such
information may prove helpful in understanding
variations in loan disposition rates among appli
cants grouped by race or gender.

USES OF NEW AND EXPANDED
HMDA DATA

Users of the HMDA data include community

based and other types of consumer - interest orga
nizations, financial institutions, state and local
government agencies, and federal supervisory
agencies. Community -based organizations have
long used HMDA data in assessing the home
lending activities of institutions in their commu
nities. Financial institutions covered by HMDA
use the informatior: to evaluate the success of
their loan marketing efforts and community out
reach programs and to compare their perfor
mance with the home lending activities of their
competitors. State and local governments find
the data useful in identifying areas that may need
assistance.
Supervisory Agencies
Supervisory agencies will be a major user of the

expanded HMDA data. The new information will
help them better assess the performance of finan
cial institutions in satisfying their obligations
under the Community Reinvestment Act and
their compliance with the fair lending laws.

Community Reinvestment Act. The CRA re
quires federal agencies to encourage depository

institutions to help meet the credit needs of their
communities, including low- and moderate
income neighborhoods, consistent with safe and
sound lending practices. Historically , examiners

have used the HMDA data to help them assess
lenders' compliance. The regulations that imple
ment the CRA establish twelve criteria for eval
24. Statistical Abstract of the United States, Money In.
come of Households, 1990 .

26. Board of Governors of the Federal Reserve System ,
1986 Survey of Consumer Finances.

Home Mortgage Disclosure Act: Expanded Data on Residential Lending 877

uating the record of depository institutions. The
HMDA data help measure institution perfor

mance against several of the criteria, including
the following:

• The geographic distribution of the institu
tion's credit applications, extensions, and

if peer lenders are receiving few applications for

home loans, weak demand may be the explana
tion. Few applications might also indicate, how
ever, that outreach efforts and marketing among
all lenders are either ineffective or not aimed at
the community in question.

denials

Fair Lending Laws. Supervisory agencies also

will use the expanded HMDA data in evaluating
compliance with the fair lending laws- the Fair
Housing Act and the Equal Credit Opportunity

Act. For example , during on-site evaluations,
Federal Reserve examiners currently review a
sample of approved and denied loan applications
to determine whether a bank is applying its stated
lending standards consistently and fairly . Exam
iners look for instances in which loan applicants
met established standards but were denied credit

and, conversely, for instances in which appli
cants failed to meet the guidelines but were
nonetheless granted credit. When they find ex
ceptions, examiners seek to determine whether
similarly situated applicants, particularly mem

bers of protected groups, were accorded like
treatment.

27. The CRA requires depository institutions to identify the
boundaries of their primary service areas referred to as
their community delineation. The boundaries must seem
reasonable , and low- and moderate - income neighborhoods
must not be excluded arbitrarily.

878 Federal Reserve Bulletin

November 1991

examiners will be able to look for statistical
indicators of possible discrimination , such as
differences in denial rates among groups . They
will then review individual home loan application
records for specific evidence of any disparate
treatment. Although different denial rates for
majority and minority group applicants, for example, ultimately may be found to have a legitimate basis, the identification of such differences
is one step in the assessment process.

by covered lenders - only for loans they thcin.
selves originated. Although HMDA information

about the census tract location of properties is
available for roughly 75 percent of the loans
sold to , or securitized by, secondary market enti
ties, information on borrowers' race or national
origin , gender, and income is available for only
about two-thirds of the loans ( table 8) . In most
instances when information is unavailable, lend.

Home Mortgage Disclosure Act: Expanded Data on Residential Lending 879
and neighborhoods where properties are lo.
cated will differ as well .

Preliminary Findings from the
HMDA Data
Lenders covered by HMDA sold roughly 2.3
million loans to secondary market institutions in
1990 ( table 8) . Most of the activity ( some 70
percent) was with FNMA , FHLMC, and
GNMA .

IN SUMMARY
The more complete information about home
lending now being gathered under the Home
Mortgage Disclosure Act will give many
groups financial institutions, community orga
nizations, supervisory agencies, and others - a

Home Mortgage Disclosure Act: Expanded Data on Residential Lending 881

8. - Continued

Borrower or census tra

Savings back or
saving and low

Number
Toualboas sold

Life insurance

Affiliate of

Other purchaser

company

Percent Number

Number

Perceel

61,205

‫ז‬
159 , ‫מ‬

Perctos

00,950

Race of borronet

American Indian Alesten
2:40
1.876
2.110
2.081
42.678

3.7
4.2

SS
462

5.7

320

4.0

85.0

33
03.1

2.1
100

2.3
100

481
2.479
5.251

2.1

2.3
88.6

1.752
10.660
23.234

20.335
291.235

3.0
6.5
5.7
81.6

.s
2.4
100

Gender of borrower
Mae
Female

Jout ( male/ female )

9.045
5.995
35.889
50.929

17.8

1.246

152
11.0

70.5
100

13.8
100

21.363
14.691
$ 3.026

119,880

17.8

100

67.662
54,452
202.479
364,593

58.704
46.652
40.797
141.912
207,865

180.644
*3.709
48.447
14.029
11.706
318.535

56.7

48.685
185.781

12.1
58.3
29.5
100

12.3
69.9

18.6
14.9

66.5
100

Income of borrower
( percentage of MSA
mediar

Less dan 80
80-99
100-120

More fan 120

5.672
4.591
4,892
27,927
43.002

13.2
10.7

870

12.0

14.210

705

9.7
10.9
67.4
100

11.309
11.474
60.172

97.165

14.6
11.6
11.8
61.9
100

60.0
20.1
14.0
3.1

81.942
17.032
11.180
3.288

70.8
14.7
9.7
2.8

64.8
100

20.4
16.1
14.2
19.3
100

Recial composition of
Cennus tract ( muronnes

as porerniage ofpopulation )
Less than 10
10-19
20-49

50-79

28.613
9.658
6.797
2.210

Upper

6.555
2.200
1.532

3.3
100

80-100

Incams of Cenu tract
Low « moderate
Middle ...

$ 8.5
19.8
13.9

6,043
28.083
14.766
43.092

12.4
57.4
30.2
100

1.9

100

846

5.319
4.554
10.919

7.1
50.5
41.7
100

10.950
65.009
39.725
115.684

9.5
56.2
14.3
100

20.0
15.2
4.4

Mno

Mean size of lona

123.29

within their communities in a fair and nondiscrim
inatory manner. Because of certain limitations
( the most important being incomplete information
about applicants' financial characteristics ) , the
expanded data alone cannot provide the answers
to these questions. Nonetheless, the data can be
expected to prompt useful dialogue between

financial institutions and members of their com
munities .

123.57

115.08

101.43

oughly lenders ' compliance with community

reinvestment and fair lending obligations. With

access to individual applications and to infor
mation about institution lending standards,
agency examiners are able to overcome most of
the data's limitations. Computerization of the
data will increase their efficiency. Finally, a
switch to 1990 delineations of census tract
boundaries, proposed for the 1992 data, will
make the HMDA information more reflective of
current lending practices.

Mortgage Lending
in Boston:

Interpreting HMDA Data
by
Alicia H. Munnell

October 1992

Working Paper No. 92-7

Federal Reserve Bank of Boston

Mortgage Lending in Boston :

Interpreting HMDA Data

by
Alicia H. Munnell , Lynn E. Browne , James McEneaney , and Geoffrey M.B. Tootell

*Director of Research , Deputy Director of Research for Regional Affairs ,
Research Department Administrator , and Economist , respectively , Federal
Reserve Bank of Boston . The views expressed are those of the authors , and do
not necessarily reflect official positions of the Federal Reserve Bank of

Boston or the Federal Reserve System .

The Home Mortgage Disclosure Act ( HMDA ) data for 1990 , which were
released in October 1991 , showed substantially higher denial rates for black

and Hispanic applicants than for white applicants . These minorities were two
to three times as likely to be denied mortgage loans as whites .

In fact ,

high - income minorities in Boston were more likely to be turned down than low
income whites. The 1991 HMDA data , which are being released currently , show a
similar pattern .

This pattern has triggered a resurgence of the debate on whether
discrimination exists in home mortgage lending .

Some people believe that the

disparities in denial rates are evidence of discrimination on the part of
banks and other lending institutions .

Others , including lenders , argue that

such conclusions are unwarranted , because the HMDA data do not include
information on credit histories , loan-to-value ratios , and other factors
considered in making mortgage decisions .

These missing pieces of information ,

they argue , explain the high denial rates for minorities .
Because the applicant and loan characteristics collected under HMDA are

indeed limited , the Federal Reserve Bank of Boston , with the support of the
other supervisory agencies , asked financial institutions operating in the
Boston Metropolitan Statistical Area ( MSA ) to provide additional information
on the financial and employment variables that lenders have indicated are

relevant to the mortgage lending decision . This information was requested for
all applications for conventional mortgage loans made by blacks and Hispanics
in 1990 and for a random sample of 3300 applications made by whites .

Substantial lender cooperation resulted in a very good response rate and high
quality data . The additional data , combined with Census information on
neighborhood characteristics , were used to develop a model of the determinants
of mortgage lending decisions in the Boston area . This model was then

employed to test whether race was a significant factor in the lending decision
once financial , employment , and neighborhood characteristics were taken into
account .
The results of this study indicate that minority applicants , on average ,
do have greater debt burdens , higher loan - to- value ratios , and weaker credit
histories and they are less likely to buy single- family homes than white

applicants , and that these disadvantages do account for a large portion of the
difference in denial rates .

Including the additional information on applicant

and property characteristics reduces the disparity between minority and white
denials from the originally reported ratio of 2.7 to 1 to roughly 1.6 to 1 .

But these factors do not wholly eliminate the disparity , since the adjusted
ratio implies that even after controlling for financial , employment, and
neighborhood characteristics , black and Hispanic mortgage applicants in the
Boston metropolitan area are roughly 60 percent more likely to be turned down
than whites .

This discrepancy means that minority applicants with the same

economic and property characteristics as white applicants would experience a
denial rate of 17 percent rather than the actual white denial rate of 11
percent .

Thus , in the end , a statistically significant gap remains , which is

associated with race .
The information gathered in this survey provides some insight into how

this outcome emerges . Many observers believe that no rational lender would
turn down a perfectly good application simply because the applicant is a
member of a minority group . The results of this survey confirm this
perception ; minorities with unblemished credentials are almost ( 97 percent )
certain of being approved .

But the majority of borrowers - both white and

minority - are not perfect , and lenders have considerable discretion over the

60-893 O - 92 - 4

extent to which they consider these imperfections as well as compensating
factors .

To take just one example , two key standards for selling mortgage loans
in the secondary market are the " obligation ratios , " which relate the

applicant's housing expense to total income and total debt burden to total
income .

Secondary market guidelines suggest benchmarks of 28 percent and 36

percent , respectively , although they go on to add that " a lender may use a

higher ratio ... when there are fully documented compensating factors ... "
( Fannie Mae 1992 , p . 654 ) .

More than one- half of the applications in this

sample exceeded one of these benchmarks , and lenders approved and sold into

the secondary market some loans with ratios in excess of 36 percent and 44
percent , respectively .
The secondary market's flexibility in this area undoubtedly increases

the general availability of mortgage funds for both minorities and whites .
Moreover , this willingness to lend to imperfect borrowers is justified :

historically , residential mortgages have been very safe investments . The
difficulty is that unless primary market lenders apply the flexibility in a
nondiscriminatory manner , minority applicants will not benefit to the same
degree as white applicants . The results of this study suggest that for the
same imperfections whites seem to enjoy a general presumption of
creditworthiness that black and Hispanic applicants do not , and that lenders
seem to be more willing to overlook flaws for white applicants than for
minority applicants .
The preponderance of flawed applicants and the significant discretion
accorded lenders have important implications for the efficacy of bank
examinations for compliance with the fair lending laws.

Since the bulk of

applications contain some flaws , most denials will appear legitimate by some

objective standard . Moreover , this study found that denied black/Hispanic
applications on average have poorer objective qualifications than denied white
applications ; that is , as measured by the median value , denied minorities had

lower income and wealth , higher obligation and loan-to-value ratios , and worse
credit histories than denied whites .

If these patterns hold true elsewhere , a

systematic bias in mortgage lending is very difficult to document at the
institution level , particularly when the number of minority applications is
small , as it is in the vast majority of institutions .

It becomes apparent

only when many applications are aggregated . As the supervisory agencies
themselves have already recognized , under existing examination procedures ,
examiners can be expected to uncover only the most flagrant abuses .
1. The Boston Area and the Boston Fed's 1989 Study of Mortgage Lending

Boston is the eighth largest metropolitan statistical area in the

nation , with a population in 1990 of 2.9 million . ' The area comprises more
than 100 politically distinct cities and towns . The largest of these
communities is the City of Boston , with a population of 574,000 .

Boston is an

old city with long-established neighborhoods , many of which are defined along

ethnic and racial lines . The communities surrounding the City of Boston were
also founded many years ago and their development has taken varied paths .
Some are lightly populated , almost exclusively residential communities .

Others function as small cities in their own right , as well as suburbs to the
City of Boston .

Boston is actually considered a primary metropolitan statistical area
( PMSA ) , meaning that it falls within an even larger agglomeration called a
consolidated metropolitan statistical area ( CMSA) . The Boston CMSA is the
seventh largest in the nation and stretches north into New Hampshire .

About 15 percent of the Boston area population is minority ( Table 1 ) .

As can be seen from the map , the minority population , especially the black
population , is concentrated in the City of Boston and surrounding communities .
Seventy percent of blacks live in the City , where they make up 24 percent of
the population . Within the City , blacks also tend to be very concentrated ;
many live in neighborhoods where more than 50 percent of the population is

black . The Hispanic population tends to live in the area's smaller cities as
well as in the City of Boston .

Both blacks and Hispanics are underrepresented

in the more residential , suburban communities.

Many of the more rural

communities are almost entirely white .

properties with two to four units . Single-unit properties are especially
scarce , and two- to four-unit properties are most common in the City of Boston
and some of the small cities . This pattern may have some bearing on mortgage

lending decisions , because evaluating an application to purchase a property
with more than one unit requires an assessment of the stream of rental income
that will be generated by the additional units .

In 1989 , the Federal Reserve Bank of Boston examined the pattern of
mortgage lending in the City of Boston and concluded that housing and mortgage
credit markets were functioning in a way that hurt black neighborhoods
( Bradbury , Case , and Dunham 1989) .

The number of mortgage originations

relative to the owner - occupied housing stock was 24 percent lower in black
neighborhoods than in white neighborhoods , after taking account of economic

variables such as income , wealth , and other factors . The study , however ,
? The results were consistent with some earlier studies that have found
evidence of redlining ( Avery and Buynak 1981 ; Dedman and others 1988 ; Gabriel
and Rosenthal 1991) . Three other studies , however , found no conclusive

1
Table

Primary
Area
Statistical
Metropolitan
Boston
the
of
Characteristics

Total

Race
by
Distribution
Percent

Population

Hispanic
Black

White

Area

PMSA
Boston

Boston
of
City

City
Central
in
Not

85.0

6.8

4.5

3.7

574.3

59.0

23.8

10.8

6.4

299.9

80.0

7.1

7.5

5.4

2.2

2.7

1,996.5

93.2

1.9

Not
of
origin
.Hispanic
9

Boston
of
City
borders
Cambridge
.Othe
nly
ynn
ramingham
CFWaltham
",Land
ambridge
Housing
.and
Population
of
Census
Bureau
of
,1Source
Census
the
:U990
.S.

Cityo
Central
Other

2,870.7

Brockton
MSA
hite
WNon
Percentage
%
2.5
Under

%
25

5.0
–
%2.5

Miles
5.0

%
10

%
50
Over

could not distinguish between discrimination in the housing market and
discrimination in the mortgage market . From the available data , it was not
possible to sort out the precise role played by lenders , as opposed to buyers ,
sellers , developers , realtors , appraisers , insurers , and others .

Thus , a

possible interpretation of the earlier study was that fewer mortgages were
made in black neighborhoods because people in black neighborhoods did not buy
houses as frequently as residents of white neighborhoods and therefore did not
apply for as many mortgages.
The results of this study do not suffer from this ambiguity .

analyzing the location of mortgage loans , this study explores the factors
affecting the decision to approve or deny mortgage applications .

In other

words, it bypasses the contention that blacks and Hispanics never enter the

doors of financial institutions and looks at what happens to individuals after
they are inside the institution and actually apply for a mortgage loan .

Such

a study is possible because amendments to HMDA in 1989 required that lenders

report not only the location of loans actually made but also the sex , race ,
and income of individual applicants and whether the application was approved

or denied.

Thus , 1990 was the first year for which information was

evidence that redlining had been practiced by lenders ( Benston , Horsky , and
Weingartner 1978 ; Canner , Gabriel , and Woolley 1991 ; Schafer and Ladd 1981 ) .

The different results from these studies appear to depend on the definition of
redlining used by the researcher .

Studies that characterized redlining in

terms of the amount of lending in a particular area were more likely to find
evidence of redlining . Others that looked at differences in the terms of
mortgage loans across neighborhoods found no conclusive evidence of redlining .

" The Home Mortgage Disclosure Act was enacted in 1975 in response to
concerns voiced by community activists that banks had demarcated areas in
cities where they were unwilling to make mortgage loans . The legislation
required that banks report the number of mortgage loans made by location of
property . These data , however , were never particularly useful in evaluating
banks ' performance , since standards were not available against which to
evaluate bank lending patterns nor was information available on individual
applicants .

available about the applicant as well as the property and about applications
that were denied as well as approved .

The new data changed the focus of

concern from " redlining , " that is , differential treatment by lenders based on
location of a property , to discrimination , that is , differential treatment of
applicants based on race or other personal , rather than economic ,
characteristics .

II . The Mortgage Lending Decision
In order to determine whether race plays a role in the lending decision ,

it is necessary first to account for all the economic factors that might bear
on the financial institution's decision .

If relevant economic variables are

not considered and they vary across racial groups , then a rational and

legitimate decision to deny a mortgage may appear to be based on race .

For

“ Although HMDA did not provide information on mortgage applications until
1990 , three major studies of applications data were conducted in the late
1970s . In 1977 , the Comptroller of the Currency and the Federal Deposit
Insurance Corporation sponsored a nationwide survey to determine what economic
characteristics were important in bank lending decisions and whether race or
sex entered into the determination ( Black , Schweitzer, and Mandell 1978 ) .

Based on an analysis of roughly 5,000 completed returns , the researchers found
that race played a statistically significant , although not particularly large ,
role in the lending decision .

97
example, if minority applicants have poorer credit records than whites ,
minorities will be rejected at a higher rate than whites .

If credit

information is not included in the analysis , the higher minority denial rate

would appear to be discrimination even if race were never considered by the
lender .

The only way to determine whether lenders ' decisions are influenced

by race is to include in a model all the economic variables that are available

to the lender and that might cause a loan to be denied , and then test to see
whether race is still a significant and important factor in the decision .
The Mortgage Application Process

The mortgage application and approval procedure is complex and far from
mechanical .

It generally consists of three steps - a quick review of the

application for viability , verification of the information and an appraisal of
the property , and an evaluation of the numbers and consideration of any
" compensating factors . "
An applicant who has decided to purchase a property selects a lender ,

based on proximity , attractiveness of rates and fees , or some other factor ,
and fills out a standard loan application form , such as Fannie Mae Form 1003 .
This can be done at the lender's site , by mail or via telephone , or by a
mortgage broker at the applicant's home .

The information contained on the

application is used by the intake person or the loan officer to make an
immediate decision as to the ultimate viability of the loan .

If the loan does

not appear viable , the lender may make its credit decision at that time and

deny the application . This initial review process saves some borrowers
application fees , but also represents the first level of discretion in the

process .
SThis paragraph describes the appropriate form of an initial review ,
which involves the completion of an application and an explicit denial or
encouragement by the lender . Examiners , however , are very concerned about the

If the lender believes that the applicant has a reasonable chance of
approval , the process enters a more comprehensive stage .

The lender attempts

to verify the information to ensure that the applicant has the financial
ability and inclination to repay the loan , and sufficient liquid funds for a

down payment and closing costs . Verification of employment provides some
assurance about both the adequacy of the income and the likelihood of

continuation of the current employment. A credit history report may provide
some information about the applicant's commitment to paying debts .

verification of bank deposits indicates whether liquid assets are sufficient ;
this step also provides some information about whether a gift , grant , or loan ,
rather than savings , serves as the down payment .

a hard look at the numbers , such as the ratios of monthly housing expense to

income and total obligations to income . These ratios are important indicators
of the ability to sell the mortgage in the secondary market . Secondary market
purchasers , such as Fannie Mae and Freddie Mac , use 28 percent and 36 percent ,
respectively , as maximum guidelines for these ratios , but these are
guidelines , and subject to considerable discretion on the part of the lender .

Assuming the application is still viable , the lender will proceed with an
appraisal and calculate the loan-to-value ratio . The secondary market uses 80
percent as a threshold for loan to value , but with private mortgage insurance
higher ratios are permitted.

At this point , the lender is in a position to approve or deny the loan .
If the credit history is clean , the applicant has a good supply of cash , all

prevalence of informal pre - screening where applicants are discouraged from

even filing a formal application or are not provided with the adverse action
notice , which is required by law when the informal process is pursued to the
point where the lender , in fact , makes a credit decision .

99
the debt and loan-to-value ratios are within the guidelines , and the property
is a single- family home in a desirable neighborhood , the decision is

relatively easy and , indeed , the application could probably be analyzed and
approved by a computer .

However , few ( less than 20 percent ) borrowers are

without blemish and , therefore , lenders are left considerable room for
subjectivity and discretion .

" compensating factors . "

To offset negatives , lenders can use a host of

For example , to compensate for high debt-to- income

ratios , lenders might note a large down payment , a good record of carrying

high housing expenses , a strong propensity to save and a high level of liquid
assets , and an excellent potential for future earnings based on education and

training .

Similarly , to compensate for credit history problems , lenders might

be willing to accept favorable letters from creditors , extenuating
circumstances such as an adverse judgment in a civil suit , or simply prior

life circumstances that have changed for the better .

In other words , many

flawed loan applications can be brought to a viable status and even made
eligible for sale in the secondary market .
A Model of Mortgage Lending

The information gathered and analyzed in the mortgage application
process can be used to model the mortgage lending decision . Because little is
known about the relationship between applicant characteristics and actual loan
performance , any model must by necessity explain what lenders actually
consider when making their decisions rather than what they ought to consider .

institution .

This goal requires that financial institutions attempt to

minimize the probability and costs of default associated with each mortgage

100
loan.6

This means that the probability of a lender denying a mortgage

application P ( D) is a function of the applicant's ability to carry the loan
( F ) , the risks of default ( R ) , the potential loss associated with default and
foreclosure ( L ) , and the terms of the loan ( T ) .

Although these factors are

listed separately , they are all interrelated ; for example , an applicant's

ability to carry a loan depends on the terms of the loan . If the lender's
judgment is influenced by the race or other personal characteristics of the
applicant ( C ) , that will also affect the likelihood of denial .

That is ,

P ( D) = f( F , R , L , T , C ) .

about the applicant - namely , income .

Income alone actually has less

explanatory power than one might expect , because lower- income borrowers
usually buy lower-priced homes . Moreover , as the discussion above suggests ,
many other variables affect the mortgage lending decision .

Thus , the Federal

Reserve Bank of Boston attempted to augment the 1990 HMDA report by gathering
information on 38 additional variables . These variables were selected on the
basis of numerous conversations with lenders , underwriters , and others
familiar with the lending process .

Most of the variables come from standard

loan application forms; several are taken from credit reports and a few from
lenders ' worksheets . The following is a brief summary of the major groupings
of variables .

Maximizing expected profit requires maximizing the difference between
the return on mortgage lending and the cost of funds to the lender .

In the

case of home mortgages, however, applications are usually either rejected or
accepted at the market interest rate .

Given expectations of inflation , the

market rate should generate a profit on loans that fulfill monthly payment
commitments .

Thus , the primary task facing the lender is avoiding default and

any associated losses . Even if the lender sells the loan on the secondary
market , default remains a concern , as the purchaser can return the loan to the
originator . At a minimum , secondary market buyers will not continue to buy
from lenders whose loans frequently default .

101

Ability of applicant to support loan . The original HMDA data did not
include information on two financial concepts - obligation ratios and wealth -

that could have considerable bearing on the applicant's ability to carry and
repay the mortgage loan .

" Obligation ratios , " which measure proposed housing

expenses relative to income and total debt payment obligations relative to
income, indicate whether the applicant can afford the mortgage more clearly
than income alone .

In addition , because the secondary market has established

guidelines for these ratios and because today most mortgages are sold in the
secondary market , lenders must be concerned about how the obligation ratios
affect the loans ' marketability .

Economists contend that wealth may also be important to the lender's
decision , since substantial wealth can make debt repayment easy even when
income is low and obligation ratios are high .

Not only can wealthy

individuals spend down their wealth , but also liquid assets can be a cushion
that prevents a temporary job loss or other income disruption from resulting
in a mortgage default .

Bankers and other lenders who were consulted said ,

however , that the available wealth information is not very reliable , and , for

this reason , they tend to place little weight on wealth , with the exception of
verifiable liquid assets .

Nevertheless , information was collected on total

assets and total liabilities , as well as liquid assets .

Risk of default. Two groups of variables - one relating to applicants '
reliability as borrowers and one pertaining to the stability of the

applicants ' income - were collected in order to capture the possibility that
the applicants ' circumstances might change and their commitment or ability to

repay the loan might decline .

Reliability of Borrower :

Lenders state that they place considerable

weight on applicants ' credit histories in judging their commitment to meeting

102

mortgage obligations . The contention is that past behavior may signal
creditworthiness in the future ; some people may be more responsible about
credit obligations than others and , therefore , less likely to default .

Loan

underwriters tend to view certain elements of the credit report as more
important than others .

For example , failure to meet previous mortgage

commitments is said to be viewed more seriously than a late credit card
payment .

Likewise, public record of default , foreclosure, or bankruptcy is

considered especially damaging to the borrower .

This study constructed a

concise outline of the prospective borrower's past creditor relationships that

provides substantial detail about different credit categories .

and skills or labor market discrimination , minorities are concentrated in jobs
that have a higher risk of unemployment , then unstable incomes could be the

reason for denials that appear to be attributable to differential treatment in
the lending decision .

Only by explicitly including a variable representing

A more sophisticated approach is also being investigated , which builds
on the job clustering work by Gittleman and Howe11 ( 1992) and the information

on individual spells of unemployment, given age , seniority , education level ,
and experience, from the University of Michigan's Panel Study of Income
Dynamics. The simpler approach adopted for this study , which uses 1989

unemployment rates in the Boston area for the major industrial groups , does ,
however , capture the concept and also has the advantage of incorporating the
local unemployment situation .

103

the probability of becoming unemployed is it possible to distinguish

discrimination in the mortgage market from effects related to race in the rest
of the economy .
Similarly , the earnings of the self-employed are thought to be more
variable than the earnings of those employed by others .
future income increases the riskiness of the loan .

Increased variance of

Thus , whether or not the

applicant is self-employed may bear on his ability to get a mortgage loan .
Potential default loss .

While credit history and employment stability

provide information about the possibility of default , several other variables
collected provide some indication of the magnitude of the loss should default

and foreclosure occur . These variables include the loan - to- value ratio , the
availability of private mortgage insurance , and neighborhood characteristics

that might affect the stability of the value of the mortgaged property .

calculated to measure the borrower's equity in the property .

Loan- to- value

ratios are poten'.ially important indicators of both the risk of default and

the magnitude of a potential loss in the event of foreclosure . The more
equity borrowers have in their properties , the less likely that declining
property values will cause them to abandon their homes to the lender . A
larger cushion also protects lenders from loss .

Private Mortgage Insurance : Since some of the loss associated with
default can be absorbed by insurers of mortgage loans , the survey collected
information on whether applicants applied for private mortgage insurance and

whether their application was approved or denied . To the extent that an
applicant applies for and receives private mortgage insurance , the potential
loss to the lending institution is reduced . More important , the secondary

104
market will not accept a mortgage loan that has a loan-to-value ratio in
excess of 80 percent without private mortgage insurance protection . Thus , any
applicant with a high loan- to-value ratio who is refused private mortgage
insurance is likely to be denied the loan .

As will be discussed later , the

fact that the insurers are basing their decisions on the same factors as the
lenders makes it difficult to determine the appropriate treatment of private
mortgage insurance in a model of mortgage lending .
Stability of Value :

Because of a variety of neighborhood features ,

inner-city properties are often thought to carry a higher risk of capital loss

than properties in other areas . While the appraised value should reflect
expectations that the property will rise or decline in value , it may not
capture the uncertainties surrounding these expectations .

Risk-averse lenders

will avoid loans with the same expected probability and costs of default but

higher variability of potential losses . As a result , lenders could be
economically motivated to avoid investing in areas that are perceived to be
risky .

Some researchers have included a separate variable for each Census tract
in their analysis to standardize for neighborhood characteristics .

approach has serious drawbacks when minorities are heavily concentrated in a
few Census tracts because the racial composition of the tract as well as the

race of the applicant may be relevant in the lending decision . A better
approach is to estimate directly the risk associated with the value of
property in different tracts .

For this study , the measure adopted was the

ratio of rent to the value of the rental housing stock in the Census tract
where the property is located , which can be calculated from Census data .

compensate investors for the higher risk , the same amount of capital invested

105

in an area with greater potential for loss should generate a higher stream of
earnings .
Loan characteristics . In order to isolate the effect of race on the
Tending decision , it is necessary to hold constant the characteristics of the
loan .

secured additional information on the duration of the loan , for example 15
years or 30 years ; whether the interest rate was fixed or adjustable ; and
whether the application was made under a program designed for low- income
individuals .

The survey also asked whether the property was a single-family

home , a condominium , or a building with two to four units .
Personal characteristics. The original HMDA data included information

on the sex and race of the applicant and co- applicant . The follow- up survey
requested data on age , marital status , and the number of dependents .

Age

could be an indicator of future earnings potential , as earnings tend to rise
with age over the average person's working life .

Similarly , lenders could be

interested in the number of dependents , because the more dependents for any
given level of income, the less money the applicant is likely to have
available to carry the loan .
In summary, the questions in the follow- up survey were designed to

secure all the financial , employment, and demographic information that lenders
may include in their determination to approve or deny a loan application .

8In the Boston metropolitan area in 1990 only 4 percent of all home
purchase applications ( only 4.5 percent of applications by blacks and 3.5
percent of applications by Hispanics ) were for government-backed mortgages .
Thus , the conventional mortgage represented the norm in Boston for blacks ,
Hispanics , and whites .

106

III . Survey Design and Results
It may be helpful to say a few words about how the sample was designed
and how the data were collected before looking at the results .

Because the

high denial rates for minorities prompted the survey and because only 1,200
blacks and Hispanics applied for mortgages in Boston in 1990 , the goal was to
collect information on every black and Hispanic applicant .

A sample of 3,300

whites was chosen to identify those characteristics that result in rejections

when race is not a factor ; this information provides a base against which to
assess the extent to which race contributes to the high rejection rate for
minority applicants .

To determine the cause of rejections among whites

requires that the sample include a sufficient number of white rejections ;

since the white rejection rate is only 11 percent , a large number of white
applicants was required .

the HMDA data that the institution had originally submitted for all its black
and Hispanic applicants and for the random sample of white applicants selected

by the Federal Reserve Bank of Boston.

For each applicant , 38 additional

pieces of information were requested . ( The survey questions are presented in

' The sample of applications by whites was selected randomly rather than
matched with black and Hispanic applications by institution or key borrower
characteristics , because matching would have required prejudging the causes of
rejection and precluded an evaluation of the role that the variables used in

the matching process played in determining rejection rates .

107

Appendix A. )

The completed forms were returned to the Federal Reserve Bank of

Boston for analysis .
Final Sample

A high degree of cooperation by lenders and considerable follow -up

resulted in a very high response to the survey, as can be seen in Table 2.10
The largest part of the divergence between the survey as designed and the
responses submitted by the institutions was caused by the closing of some

banks that had been significant lenders in 1990. A second source of
difference was that lenders , in the process of providing additional data ,
checked their earlier entries and made corrections .

In one of the more

notable examples , 51 applications that a suburban bank had coded as Hispanic

on its original HMDA submission were found to be white .

Some institutions

were simply unable to locate all their loan files .
The survey response was further refined to derive a sample of completed

applications for conventional loans for the acquisition of residential
property .

This required eliminating any application that , upon review , was

for refinancing as opposed to home purchase or for the acquisition of

nonresidential as opposed to residential property , and any application with
missing data for one of the key variables .

In addition , the decision was made

to exclude applications that were withdrawn .

10 The institutions participating in the survey were requested to keep

track of the expenses they incurred in supplying the information . Only
sixteen of the 131 institutions responded with estimates of the hours devoted
to the survey or with dollar expenditure figures . According to these

estimates , the time required to supply all the information for a single loan
averaged about an hour and the dollar cost averaged $ 30 per loan , a figure
generally consistent with the hourly estimate . These costs are probably
indicative of those experienced by the other lenders participating in the

survey. Applying these estimates to the entire sample indicates that
approximately 4,500 hours were expended in complying with this survey request
and that the total dollar cost was $ 135,000 .

Table
2

,Boston
MSA
HMDA
Reports
1990
with
Original
Sample
Final
Comparison
of
H
/ ispanic
Black

White

Total
of
Number

Applications

Source

Reports
HMDA
Original
Design
Survey

Final
Sample

16,019

11.0

1,210

30.7

4,443

3,300

11.0

1,143

30.4

4,153

3,123

11.4

1,013

27.6

3,062

2,340

10.3

722

28.1

108

Response
Survey

18,838

than
bapplicants
.,owhite
Hispanic
lack
"Irncludes
other
races
of

closing
of
bsurvey
the
of
ecause
(4fsurvey
)design
,443
short
alls
response
,153
:Tsome
Note
he
files
,aof
nd
loan
find
to
inability
lenders
some
tbanks
1990
in
he
significant
been
had
that

response
)(4of
,153
3corrections
sample
ffinal
survey
the
alls
,062
.Tshort
submissions
earlier
to
he
for
refinancings
)o(2were
00
r
withdrawals
,smissing
ome
32
ome
had
6because
data
18
loans
some
purchase
,ahad
mortgages
some
nd
lenders
home
as
coded
originally
(24hat
property
)tnonresidential
Hispanic
).(1,obneither
white
to
be7rlack
applicants
proved

109

Some experts have suggested that withdrawals may be hidden rejections .
That is , in the process of verifying an application , the lender could
encourage the applicant to withdraw rather than be rejected .

applicants might withdraw for a host of other reasons .

However ,

In particular , the

property might fail an inspection report or the buyer might simply get cold
feet .

Withdrawals accounted for roughly 8 percent of both black/Hispanic and

white applications .

An examination of the pattern of withdrawals in the

sample revealed , at most , a weak link to race or creditworthiness . Since
retaining withdrawals in the study would have complicated the econometric
presentation that follows and produced uninteresting results, they are not
included in the sample . Despite the reduction in the number of applicants in
the final sample , the pattern of denial rates is fairly close to that reported
in the original HMDA data .

The pattern of lending by type of institution is also very similar to
that reported for the original HMDA data .

In both cases , applications are

split relatively evenly between depository institutions and mortgage
companies ; this is true for blacks / Hispanics as well as for whites ( Table 3 ) .

Values of Key Variables

The values of key variables collected in the follow- up survey are

presented in Table 4 for black/ Hispanic applicants and white applicants , both
approved and denied .

of variables . )

( Appendix Table Al presents values for the complete list

These data and all subsequent analyses combine applications by

blacks and Hispanics . Both blacks and Hispanics had substantially higher
denial rates than whites and the number of applications by Hispanics was too

small to analyze separately . Moreover , statistical tests confirmed that the

110

Table 3. Institutions Providing Mortgage Loans and Denial Rates , Final Sample

Total Applications

White Applications B/H Applications

Percent

Institution

Banks , Thrifts , and
Credit Unions

1,638

14.0

1,265

9.6

373

28.6

Mortgage Companies

1,424

15.1

1,075

11.1

349

27.5

318

27.7

25.8

722

28.1

Subsidiaries
Independents
Total

1,297

3,062

15.3

979

12.6

11.3
8.3

14.5

2,340

10.3

111

Table 4
Key Characteristics of Mortgage Applicants , by Race and Loan Disposition
White

Approved

Variable

Denied

Black /Hispanic
Approved
Denied

Ability to Support Loan
26.0

26.6

26.0

28.0

14.6

38.9

23.4

51.5

3.2
1.5

3.2
7.4

90.0
8.9

329

Liquid Assets ( $ )
Risk of Default

Percent with Poor Credit History
Probability of Unemployment
Percent Self- Employed
Potential Default Loss

Loan/Appraised Value ( percent)
Rent/Value in Tract ( percent)

77.3
4.6

83.1
4.9

85.0
7.3

21.6

17.1
75.0

42.2

26.6

1.3

82.5

Percent Applied for Private Mortgage

Loan Characteristics
Percent Purchasing Two- to Four- Family
24.8
60.6

34.4
69.6

16.1

91.1
40.6

91.3
40.3

35.0
53.2
39.9

36.0
53.7
52.6

36.0

1.7
68.6
85.9

18.3
62.8
83.3

12.6

34.0
63.0
37.6

Personal Characteristics
Age
Percent Married

Percent with Dependents

55.0
52.2

Median value .

bpoor credit defined as having more than two late mortgage payments or delinquent consumer
credit histories ( more than 60 days past due ) or bankruptcies or other public record
defaults .
Base is those applying for private mortgage insurance .

See Appendix Table Al for complete list of variables .

112

independent variables affected the probability of denial for the two groups
similarly .
The data show that black and Hispanic applicants in the Boston area
differ from white applicants in a number of ways .

These differences tend to

support arguments that the higher denial rates experienced by minorities are
attributable , at least in part , to financial characteristics , credit

histories , and other economic factors . As reported in other surveys , black
and Hispanic applicants have considerably less net wealth and liquid assets

than whites .

Black and Hispanic applicants also tend to have poorer credit

histories than whites .
Blacks and Hispanics in Boston are substantially more likely than whites
to be purchasing a two- to four-family home .

The higher proportion of two- to

four - family homes among denied applicants , for whites as well as for blacks
and Hispanics , suggests that lenders perceive more risk associated with
financing the purchase of such properties . Blacks and Hispanics also make
lower down payments and have higher loan-to-value ratios than whites .

Since

the secondary market will not accept a mortgage with a loan - to -value ratio in
excess of 80 percent without mortgage insurance , minorities apply more

frequently for private mortgage insurance .

similar . Supporting the view that obligation ratios rather than incomes are
the critical variable is the fact that the median income of white applicants

whose loans were approved was virtually the same as the median income of
applicants whose loans were denied ; in the case of minority applicants , the
median income of denied applicants actually slightly exceeded the median
income of those whose loans were approved .

113

IV .

The Role of Race in the Mortgage Lending Decision
While the data in Table 4 suggest that financial and other differences

between black/ Hispanic and white applicants account for a large part of the
disparity in mortgage denial rates , determining whether race plays an

independent role , and how great a role , requires statistical techniques that
hold these characteristics constant . This can be done by estimating an
equation which makes the probability of being denied a mortgage loan a
function of obligation ratios , wealth variables , credit histories , and other

factors thought to affect the mortgage decision .

Race is then added to the

equation to determine whether it has any independent effects after the other
factors have been taken into account .

Regression Results

equations is presented in Appendix B , and it confirms the stability of the
results . 11

The first column of Table 5 reports the coefficient associated with each
variable .

The " t - statistic " in parentheses indicates the statistical

significance of the coefficient ; a t- statistic in excess of 2 means that the

coefficient is statistically significant .

With the exception of wealth , all

" As discussed earlier , little is known about the link between applicant
characteristics and loan performance; thus , the results describe what lenders
actually consider in their decision to approve or deny a loan , but these are
not necessarily the factors that would provide the best predictions of
repayment or default .

114

Table 5

Determinants of Probability of Denial of Mortgage Loan Application

Impact of Variable on
Coefficient
( t-Statistic )

Variable

Probability of Denial

-6.61

Constant

( -17.0 )
Ability to Support Loan

Housing Expense / Income

.47

33.9

( 3.2 )

Total Debt Payments / Income

33.0
4.5

Net Wealth

Risk of Default
.33
( 9.8 )

Consumer Credit History

37.2

11.4

Mortgage Credit History
Public Record History

113.7

Probability of Unemployment

11.4

Self - Employed

35.1

Potential Default Loss

.58

Loan /Appraised Value

11.5

( 3.2 )

Denied Private Mortgage Insurance

596.0

Rent / Value in Tract

9.3

Loan Characteristics

Purchasing Two- to Four- Family Home

.58

42.4

( 3.6 )
Personal Characteristics

Race

.68

56.0

( 5.0 )
Number of Observations

3062

Percent of Correct Predictions
For variables entered as 0 or 1 ( see the notes to this table ) , the increase
in the probability of denial associated with the variable . For continuous
variables , the increase in the probability of denial associated with a change
in the variable equal to one standard deviation .
The number of applicants with a probability of denial greater than 50 percent
who were denied , plus the number of applicants with a probability of approval
greater than 50 percent who were approved , as a percent of the total sample .
27

115

Rotas to tables
Der Variabl . Definitions :

Housing Expenso / Income
0

Total Debt Payments / Incom .

il groater than .30 ,
othomiso

value of question +46

Not Wealth

value of question +36 loss question +38

Consumer Credit

•

•
.

Mortgage Credit
•

if no " slow pav " account ( code I in question +43 )

If one or two slow pay accounts ( code 2 )

il noro than two slow pay accounts ( code 3 )

. 1 if no late payments ( cod . 1 in question # 42 )
2 if no payment history ( cod . 0 )

Public Rocord

Probability of Unemployment

Solt - Imployed

18 any public record of credit problems ( codes 1 , 2 , 3 , 4 in question

1989 Massachusetts unemployment rate for applicant's industry
1
0

Loan /Appraised Value

if self - employed
otherwise

value of loan amount divided by question #50

Percent Denied Private Mortgage Insurance

derived from question +53

Rant / alu , in Tract

rental income divided by estimate of value of rental property from
Census

Two to Four - Family Homes

Race

if applicant was black or Hispanic ,

otherwise

1 € purchasing a single - family or a condo ,
if purchasing a two to four - family home

Means and Standard Deviations :
Variable

Total Debt Payments / Income
Net Wealth ( $ )

Mean

Standard Deviation

33.46

11.26

230 , 160

979,245
1.70

Consumer Credit History

2.18

Mortgage Credit Bistory

1.75

. 53

Probability of Unemployment

3.82

2.07

Loan / Appraised Value

.77

.33

Rent /Value in Tract

.09

.23

116

the variables in the equation have a statistically significant impact on the
probability of denial .
The importance of the variables to the denial decision cannot be

interpreted solely from the t- statistics or from the coefficients themselves ,
but rather depends on the values of the variables in the equation . Thus , the
second column presents a measure of the impact of each variable on the
probability of denial .

For variables that have values of 0 or 1 , such as

self - employed , the figures in the second column represent the increase in the
probability of denial associated with having that particular characteristic .
That is , the probability of denial increases 35 percent for a person who is

self -employed . "

19.6 percent .

Since the average denial rate for the sample as a whole is

For continuous variables , such as the total obligation ratio ,

" ? Logit regressions are particularly suited to modelling discrete
outcomes , such as approval or denial . However , the resulting equations are
nonlinear and , therefore , calculating the impact of changes in variables is

more complicated than in the more familiar ordinary least squares and other
linear regression forms. In deriving the impact values reported in Table 5 ,
the first step is to determine the probability of denial in the absence of a

particular characteristic , such as being self - employed. This requires
determining for each non - self - employed applicant the probability of denial
based on the coefficients of the equation reported in Table 5. These
estimated probabilities for each applicant are then averaged to get a single

figure for the group . The second step is to add to each non - self- employed
applicant's probability of denial the impact of being self - employed ( the
coefficient 0.52 multiplied by 1 ) . These new probabilities are averaged. The
figure reported in the second column is the percent difference between the
average probability of denial for the non - self- employed with the self

employment effect and the probability for the non - self - employed without it .

117

the figures in the second column represent the increase in the probability of
denial associated with a one standard deviation change in that variable .

is , if the total obligation ratio rises 11 percentage points ( one standard
deviation ) , the probability of denial increases by 33 percent .

application denied .

Because the two obligation ratios tend to move together ,

that is , an applicant with a high housing expense ratio generally also has a

high ratio of total debt payments to income , it is difficult to sort out
precisely the relative importance of the two ratios . Suffice it to say that
these measures are crucial to the lending decision . As discussed above , one
standard deviation increase in the total obligation ratio raises the

probability of denial by 33 percent .

do not appear to affect the probability of denial , although they are cited in

secondary market guidelines as a compensating factor and are frequently
mentioned by lenders as an important consideration . The answer may be that
liquid assets are frequently used for the down payment and therefore their
effect is captured by the loan-to-value ratio .

Pre- screening may also exclude

13An equation was also estimated including income , liquid assets , and the
ratio of base to total income as alternative measures of the applicant's
ability to carry a loan .

None of these variables has a statistically

significant effect on the probability of being denied ; the results can be
found in Appendix Table Bl .

118

Risk of default. Credit information was categorized by the severity of

the problem in the consumer , mortgage, and public records areas ; the precise
definitions can be found in the notes to Table 5 . The results show clearly
that an increase in credit problems raises the probability of having the loan
denied .

A problem in the public records area , such as a bankruptcy , raises

the probability of denial 114 percent."

Thus , if an applicant with average

characteristics of the sample had a bankruptcy , this person's probability of

denial would roughly double from 14.5 percent to 31.0 percent .

denial .

Self- employment has by far the larger effect , however , raising the

probability of denial by 35 percent.15
Potential default loss . A high loan-to-value ratio raises the
probability of denial , but the effect is relatively small .

This result occurs

because virtually all applicants with loan-to- value ratios over 80 percent

must secure private mortgage insurance . Thus , as shown in Table 5 , the denial
of private mortgage insurance virtually precludes attaining a mortgage .

should be noted , however , that very few applicants were turned down for
private mortgage insurance .

The large impact , therefore , means that those who

were turned down were very unlikely to get a mortgage , not that denial of

" An alternative characterization of credit history , which treats the
credit information as individual dummies rather than as semi- continuous

variables , is presented in Appendix Table B2 . The results are fully
consistent with those in Table 5 .

119

private mortgage insurance was the most important reason to be denied a
mortgage loan .

The appropriate way to treat private mortgage insurance was a difficult
decision , because these insurers consider the same information provided the
financial institutions .

Thus , in one sense , they could be considered simply

another lender and the mortgage insurance variable omitted from the equation .
On the other hand , insurers could be viewed as outside the direct lending
market , and , to the extent that their denials fell disproportionately on
minorities , excluding a variable representing denial of mortgage insurance
from the equation would ascribe to lenders differential treatment occurring
elsewhere in the system .

For this reason , the denial of mortgage insurance

was included in the equation .
Since the treatment of private mortgage insurance is controversial , it
should be noted that excluding private mortgage insurance from the equation
has little impact on the coefficients of the other variables ; the exception ,
not unexpectedly , is the loan - to-value ratio , which takes on somewhat greater
importance in the absence of private mortgage insurance ( Appendix Table B4 ) .

Similarly , estimating the equation excluding those applicants who were denied
private mortgage insurance has little impact on the basic results ; again the
exception is the loan-to -value ratio . 16
Finally , the theoretical construct to standardize for the riskiness of
the neighborhood in which the property was located entered the equation with

the expected sign and was statistically significant . That is , the greater the
rent - to-value ratio , which attempts to measure the variability of housing
16 In terms of the determinants of private mortgage insurance itself ,
nearly all the variables included in the mortgage loan decision equation ,
including race , appear to be relevant . The effect of race disappears ,

however , with the addition of information about the racial composition of the
tract in which the applicant is purchasing the property ( Appendix Table B5 ) .

V
120

value from tract to tract , the greater the likelihood the applicant will be

denied a mortgage loan . " ?
dummy variable for each of the more than 500 tracts in the sample - the
ultimate exercise in controlling for neighborhood characteristics .
inclusion of these additional variables has a modest impact on most of the
other coefficients in the original equation ; the exception is the coefficient
on race , which increases ( Appendix Table B9 ) .
Loan characteristics . The loan characteristic that turned out to be
important is whether the applicant was applying for a mortgage for a two- to

four-family home . 19 Financial institutions clearly are less willing to make

" ? Equations were also estimated with several alternative indicators of
the risk of loss arising from the property's location ( Appendix Table B6 ) ;

these include vacancy rates , the appreciation in housing values , and a dummy
for tracts with more than 30 percent minority population . These variables do
not alter the basic equation appreciably . It appears that although blacks and
Hispanics tend to reside in minority areas, they are not being denied
mortgages because of where they live . Minorities living in white areas are
also denied mortgages at higher rates .

121

loans on two- to four- family housing that involves rental arrangements.

The

positive coefficient says that if the property is a multi -unit dwelling , the
probability of denial rises 42 percent .
Personal characteristics . The only personal characteristic that appears

to enter into the loan denial decision is the race of the applicant.20

The

positive and statistically significant coefficient suggests that after

accounting for obligation ratios , wealth , credit histories , stability of the
applicants ' incomes , loan-to-value ratios , private mortgage insurance , and
neighborhood characteristics , the race of the applicant still plays a role in
the lender's decision to approve or deny the loan .

Thus , for an individual

with average white economic characteristics and minority race , the probability
of denial increases 56 percent .

Evaluation of the results. A logical question is " How good are these
results ? "

This question can be broken into four parts .

The first pertains to

the robustness of the results with regard to race ; the second pertains to the
broader issue of how much of the variability in approval and denial rates is
explained by the equation ; the third relates to whether the results can be
explained by variations in underwriting standards among lenders ; and the
fourth relates to the pervasiveness of the behavior captured in the equation .

With regard to the race variable , nearly every equation that was
estimated had virtually the same coefficient and degree of statistical
significance .

As shown by the supplementary equations reported in Appendix B ,

adding variables to the equation reported in Table 5 had little impact on the
coefficient of race or for that matter on most of the other coefficients in

20The age , sex , marital status , and number of dependents do not affect
the probability of having a loan application denied ( Appendix Table B11 ) .

60-893 0 - 92 - 5
O

122

the equation .21

In short , the effect of race on the probability of denying a

loan application was consistently positive , large , and statistically
significant.22
Robustness of the race coefficient in and of itself does not fully
answer the question of how much credibility should be given to these results .

If important variables that differed by race were missing from the analysis ,
the race variable could be picking up their effect .

Two responses address the

issue of omitted variables . First , the survey included every variable
mentioned as important in numerous conversations with lenders , underwriters ,

and examiners and no reviewer suggested any other economic factor that should
be included in the equation .
Second , the variables included in the equation do a good job of
explaining the decision to approve or deny .

Although no simple measure of

" goodness of fit " exists for equations that estimate the probability of an
action , the explanatory power of the equation can be assessed .

The first

column of Table 6 reports actual denial rates for applicants in the survey by
total obligation ratio ; that is , the denial rate for very good credits

2lVarious interaction terms were tested to examine whether a combination
of certain variables was essential to the mortgage lending decision .
Interaction between the loan-to-value ratio and the obligation ratios and
credit history variables , as well as the interplay between the obligation
ratios and the credit variables were all tested . Only the loan-to-value ratio
and consumer payments interaction term was statistically significant . The
importance of this variable , however , derived solely from its severe

collinearity with the consumer payments index ; the consumer payments variable
becomes insignificant when this interactive term is included, and the
correlation between the two variables is 0.9 . None of these interactive terms

affected the race coefficient or its statistical significance . Finally , some
non-linearity in the obligation ratios and the loan- to- value ratio was
examined , but it did not improve the fit of the equation or change any of the
results for the other variables .
22As shown in the correlation matrix ( Appendix Table B14 ) ,

multicollinearity between any two independent variables is not affecting the
results .

123

Table 6

Explanatory Power of the Regression Equation
Denial.Rates
Predicted from Equation Based on
Original
Key

Sample

Full
Model

HMDA Data

Variables

9.9

10.6

14.0

12.4

Between 36 percent and 40 percent

14.4

16.2

15.2

16.2

Greater than 40 percent

38.8

32.3

16.2

22.9

Actual

Total obligation Ratio

36 percent or lower

Equation includes race , sex , and income of the applicant and the loan amount.
Key variables add to the original HMDA data a dummy when the ratio of housing
expense to total income exceeds 30 percent , a measure of the applicant's
consumer payment credit history , and the applicant's loan-to-value ratio .

124

( obligation ratios 36 percent or lower ) is 9.9 percent and for poor credits
( obligation ratios in excess of 40 percent ) is 38.8 percent .

The second

column reports the denial rates predicted by the equation for each group .
the good credits, the equation performs remarkably well , predicting 10.6

percent compared with the actual of 9.9 percent . The results for the denial
rates for poor credits are also quite good , 32.3 percent compared to the
actual of 38.8 percent .

In order to have a better sense of how good the equation results are , it
is useful to compare the predictions with those that emerge from an equation

using only information from the original HMDA data - namely , race , sex , and
income of the applicant and loan amount.

As shown in the third column of

Table 6 , these four variables produce a flat distribution of predicted denial

rates , explaining none of the difference between good and poor credits .

other words, the additional variables included in the full model explain a lot
compared to the basic HMDA data . To provide just one more point of
comparison , the last column shows the predicted denial rates from an equation
-

that adds only three additional variables to the original HMDA data - a dummy
for a ratio of housing expense to total income in excess of 30 percent ,

consumer payment credit history, and loan-to-value ratio . This equation
begins to pick up some of the tilt in denial rates as applicants move from
poor to good credits , but a substantial gap remains between actual and
predicted rates .

Third , the question arises about the pervasiveness of the results . That
is , does the impact of race come from a single large institution operating in
a discriminatory manner or is the practice widespread?

To test whether race

was consistently an important factor in the mortgage lending decision , the
sample was divided into large lenders and small lenders .

Large lenders , which

125

accounted for only 5 percent of the institutions , received exactly 50 percent
of minority applications ; the other 50 percent of minority applications were
distributed among the remaining 95 percent of the institutions .
equations were then estimated for the two sub - samples .

Separate

The results indicate

that the model is stable across institutions of vastly different size , and
that race is an important explanatory factor in mortgage lending decisions
among both small and large lenders ( Table 7 ) .

In short , the results represent

a widespread phenomenon , not just the behavior of a single institution .

and property characteristics , the argument remains that minorities may be

treated the same as whites within any given institution , but may simply
frequent institutions with tougher lending standards .

To test this

hypothesis , a " tough " lender variable was added to the basic equation .

This

variable was constructed by estimating the equation for white applicants only
and including a separate dummy variable for each lender , and then designating
specific lenders as " tough " based on the coefficients of the lender dummies.

The inclusion of this variable , however , had virtually no effect on the
coefficients of the other variables and the variable itself was statistically
insignificant ( Appendix Table B12 ) . This result was not unexpected given that
most lenders conform to secondary market guidelines .

Including separate dummy

variables for all institutions in the sample alters the coefficients slightly ,
but does not change the basic results .
This assessment shows that the results presented in Table 5 merit
serious consideration .

The coefficient of the race variable is stable and

always statistically significant ; it is difficult to think of omitted
variables linked with race that could be biasing the race coefficient ; and the
overall equation does a very good job of explaining the variation in denial

126

Table 7

Determinants of Probability of Denial for Large Lenders and Small Lenders
Coefficient
Variable

-6.59

Constant

( 14.1 )

-7.53

( 9.6 )

Ability to Support Loan

Housing Expense / Income

.50
( 2.5 )

( 1.7 )

.36
( 7.7 )

.30
( 6.2 )

.39
( 2.0 )

( 2.9 )

.39

Total Debt Payments /Income
Net Wealth
Risk of Default

Consumer Credit History
Mortgage Credit History

Public Record History
Probability of Unemployment
Self - Employed
Potential Default Loss
Loan / Appraised Value

1.54

Denied Private Mortgage Insurance
Rent /Value in Tract

Loan Characteristics

Purchasing Two- to Four- Family Home

1.16

( 5.3 )

-.09

( 0.4 )

Personal Characteristics
.51

Race

( 2.6 )
1968

Number of Observations
Percent of Correct Predictions

.68

( 3.4 )
1094

The number of applicants with a probability of denial greater than 50 percent
who were denied, plus the number of applicants with a probability of approval

greater than 50 percent who were approved , as a percent of the total sample.

127

rates . Moreover , the equation is describing widespread behavior , not simply
that of a single large institution or of particular types of institutions , and
variation in lending standards does not appear to explain the results .
An Alternative Approach
Estimating an equation that includes an explicit measure for race is not
the only way to test whether race is an important factor in the mortgage

lending decision . An equally good alternative is to estimate an equation for
white applicants and then plug in the obligation ratios , loan- to- value ratio ,
credit history , and other values for each black/Hispanic applicant to
calculate that applicant's probability of denial .

The resulting discrepancy

between the actual minority denial rate and the estimated minority denial rate
based on the white equation can be interpreted as the effect of race on the
mortgage lending decision .

The equations estimated separately for white and black/Hispanic
applicants are reported in Appendix Table B13 and the results of estimating
the probability of denial based on the white equation are shown in Table 8 .
If blacks / Hispanics had their own characteristics , that is , high obligation
ratios , weaker credit histories , higher loan - to- value ratios , and less likely

to buy a single-family home , but were treated by lenders like whites , their
average denial rate would be 20.2 percent rather than the actual 28.1 percent
experienced by minority applicants .

In other words , economic , property and

neighborhood characteristics explain much of the higher minority denial rate ,
but 7.9 percentage points remain unexplained .

race on the lending decision , this amount can be added to the white denial
rate to estimate the racial impact starting from the white base .

That is , the

third line in Table 8 shows what the denial rate would have been for black and

128

Table 8

Probability of Black/Hispanic Denials Based on White Experience
Denial Rates

Characteristics and Experience

( percent )

Actual Denial Rate for Blacks /Hispanics in Sample

28.1

Denial Rate for Blacks/Hispanics with Black/Hispanic

20.2

Denial Rate for Blacks / Hispanics with White

18.2

Actual Denial Rate for Whites in Sample

10.3

Addendum :

Ratios of Black/ Hispanic to White Denial Rates

Actual ( 28.1 / 10.3 )
Based on Black/Hispanic Characteristics ( 28.1 / 20.2 )
Based on White Characteristics ( 18.2 / 10.3 )

2.7
1.4
1.8

129

Hispanic applicants if they had white obligation ratios , loan-to- value ratios ,
credit histories , and other characteristics but were treated by lenders like
minorities .

Thus , even if minorities had all the economic and property

characteristics of whites , they would have experienced a denial rate of 18.2
percent , 7.9 percentage points more than the actual white denial rate of 10.3 .
Some ambiguity arises when these various denial rates are used to
characterize the ratio of minority to white denial rates .

If the ratio is

calculated using black/Hispanic characteristics , the ratio is 1.4 to 1 ; if
white characteristics are used , the ratio is 1.8 to 1.

The 1.8 to 1 ratio is

the appropriate comparison with the 2.7 to 1 ratio of unadjusted denial rates ,
since both use the white experience as the base .
The important point , however , is that the ratios bracket the 56 percent
increase in the probability of denial for minority applicants reported in

Table 5. This confirmation of the earlier results lends additional support to
their credibility .

VI . Conclusions

This study has examined one avenue through which differential treatment
could affect minorities ' access to credit and opportunities for homeownership .
It found that black and Hispanic mortgage applicants in the Boston area were
more likely to be turned down than white applicants with similar
characteristics .

It is important to clarify the limited focus of this analysis ; it

abstracts from discrimination that may occur elsewhere in the economy . For
example , if minorities are subject to discrimination in education or labor

markets , they will have lower incomes and their applications may reflect
higher obligation ratios , greater loan-to-value , or poorer credit history .

130

Similarly , if blacks and Hispanics are discouraged from moving into
predominantly white areas , they will limit their search to neighborhoods
sanctioned for minorities .

These tend to be older central cities with high

density housing , such as two- to four- family homes .

Denial of a mortgage loan

application on the basis of either these economic or property characteristics

would not be considered discriminatory for the purposes of this study.
Even within the specific focus of conventional lenders , the reported

measure of the hurdles faced by minorities should be placed in perspective ;
differential treatment can occur at many stages in the lending process . For
example , minorities may be discouraged from even applying for a mortgage loan
as a result of a pre- screening process . Similarly , if white applicants are
more likely than minority applicants to be " coached " when filling out the

application , they will have stronger applications than similarly situated
minorities .

In this case , the ratios and other financial information in the

final application , which were the focus of this analysis , may themselves be
the product of differential treatment . This study does not explore the extent
to which coaching occurs , but rather focuses on the impact of race on lenders '
decisions regarding the final applications received from potential borrowers .

consider whether to deny or approve a mortgage loan application .

The impact

of race is substantially less than indicated by the original 1990 HMDA data ,
which showed that black and Hispanic applicants for mortgages in the Boston
metropolitan area in 1990 were turned down at a rate 2.7 times that for white
applicants . As it turns out , the higher denial rate for minorities in Boston
is accounted for , in large part , by their having higher loan- to -value ratios

and weaker credit histories than whites . They are also more likely to be
trying to purchase a two- to four-unit property rather than a single- family

131

home.

Nevertheless , after taking account of such factors , a substantial gap

remains .
A black or Hispanic applicant in the Boston area is roughly 60 percent
more likely to be denied a mortgage loan than a similarly situated white
applicant .

This means that 17 percent of black or Hispanic applicants instead

of 11 percent would be denied loans , even if they had the same obligation
ratios , credit history, loan to value , and property characteristics as white

applicants .

In short , the results indicate that a serious problem exists in

the market for mortgage loans , and lenders , community groups , and regulators
must work together to ensure that minorities are treated fairly .

132
Appendix A
Attachment 1

INSTRUCTIONS FOR COMPLETING LOAN / APPLICATION REGISTER ( LAR )
Our records indicate that your institution listed ( XX ) applications from
blacks and Hispanics in your 1990 HMDA Report ; all of their identification

numbers and basic HMDA information are reproduced in Attachment 4 , the
Loan/Application Register . As a control group , we have randomly selected ( XX )
white applicants ; the information for the whiteapplicants also appears in the

Register. Although this information is taken directly from your submissions,
it would be useful for you to check it for accuracy .
In addition , please review " Reasons for Denial " ( column 19) , and if you
have not already included the reasons , please enter that information at this
time .

The reasons should conform to Attachment 2 , Regulation B , Form C - 1

" Sample Notice of Action Taken and Statement of Reasons " ( Adverse Action
Notice ) . The reasons ( up to three) should be entered on the Register , from

left to right in the space provided .

A. Data from Residential Loan Application ( Fannie Mae Form 1003 ) , see sample
on Attachment 3 .
Note : Information for loan applications which were approved should come

133

27 : Self - employed ( y or N )

For the next four columns , sum applicant and co-applicant
information if separate statements were completed .

B. Data Relating to Credit History
Column 39 : List the number of commercial credit reports in the file

C. Obligation Ratios ( from Tender worksheets)
Column 45 : Debt - to - income ratio ( housing expense / income)
D. Loan Characteristics
Column 47 : Fixed or adjustable rate ( F or A )

134

E. Unverifiable Information

Column 56 : Type of information on the application which could not be

Underwriting Information
Column 57 : List total number of times application was reviewed by the

INSTRUCTIONS FOR COMPLETING COLUMNS #42-44
Enter the number that best describes the credit history ( from the commercial

credit report ) of the applicant ( s ) . Note that these column's should be completed
regardless of the loan disposition or your answer to #40 .
CREDIT HISTORY CODES - Mortgage Payments ( Column 42 ) :
0 - no mortgage payment history

1 - no late mortgage payments
2 - one or two late mortgage payments

0 - Insufficient credit history or references for determination
1 - no " slow pay " or delinquent accounts , but sufficient references for

CREDIT HISTORY CODES - Public Records ( Column 44 ) :
0 - no public record defaults

1 - bankruptcy
2 - bankruptcy and charge -offs
3 - one or two charge- off ( s) , public record ( s ) , or collection action ( s ) ,

135

Appendix Table A

values of Variables Collected on Follow -Up Survey , Boston MSA
Loan Application Register
No.

Approved and Denied Applicants

Characteristic

Median number of units in property purchased

Median age of applicant

“ % 8A213.

White

36
14

54.1
72.3

7.4

1.8
n.a.

136

Appendix B

Alternative Specifications of the Probability

Alternative specifications of the probability that a mortgage

application will be denied are presented in this appendix .

The additional

variables are based on the model of mortgage lending outlined in the text and
the suggestions of experienced researchers in this field .

Ability to Support Loan

Table Bl compares the basic equation from Table 5 with an equation
incorporating additional measures of the applicant's ability to support the

loan . As can be seen , the coefficients of most of the basic variables are
affected only modestly by the addition of income , liquid assets , and the ratio

of base income to total income . The coefficient for race remains almost the

Risk of Default

137

probability of being denied a mortgage than borrowers with some late payments.
The finer breakdown of credit history does not alter the coefficients of the
other variables , including race .

Potential Default Loss
Private Mortgage Insurance. As discussed in the text , the appropriate
treatment of private mortgage insurance is unclear . If race enters into the

insurance decision , the inclusion of a variable representing the denial of
insurance will understate the difficulties that minorities face in securing
mortgages , since the effect of race on the ability to get insurance and ,
therefore , to get a mortgage would be subsumed in the mortgage insurance
variable. Accordingly , Table B4 shows the effect of omitting this variable .

Also shown is an equation in which all mortgage applicants who were denied
mortgage insurance are omitted from the sample . In both cases , the
coefficients for the other variables , including race , are similiar to those in

Table 5. These results suggest that the probability of denial facing minority
applicants is not substantially understated by including the mortgage

insurance variable in the basic equation .

138

and since the variables collected for this study were not gathered for this
purpose .

Location . Table B6 adds to the equation in Table 5 several indicators
of the risks of loss arising from the property's location . In the basic
equation , the riskiness of the neighborhood is represented by the ratio of
rental income to the value of the rental housing stock in the relevant census

Foreclosures . Some researchers have suggested the foreclosure rate as a

measure of neighborhood risk . This has considerable intuitive appeal , since
the lender's objective is to minimize the probability and costs of
foreclosure. The direction of causality is ambiguous , however . A high
foreclosure rate could be the result of lenders ' reluctance to make loans in a

neighborhood as well as a cause of such reluctance . Homeowners who fall
behind in their mortgage payments will not be able to get out from under their

troubles by selling their properties if prospective buyers cannot get loans .

Table B8 shows the pattern of foreclosures in the planning districts of
the City of Boston along with the racial composition of the districts . The
districts with the very highest foreclosure rates were predominantly white .
Foreclosure rates in predominantly minority areas ranged from high ( Mattapan )
to quite low ( South End ) .
Tract Dummy Variables. As a final test of whether the coefficient on

race might be representing lenders ' concerns about the location of the
property , a dummy variable was used to represent each of the more than 500
census tracts in which applicants were attempting to purchase homes .
a crude approach . It provides no indication of why lenders might deny

mortgages in a particular area , and if minorities tend to be concentrated in
particular neighborhoods it risks attributing rejections that are influenced

139

by the applicant's race to location . Nevertheless , as can be seen from Table
B9, the inclusion of dummy variables for each census tract actually increased
the coefficient for the race variable .

Loan Characteristics

Table B10 adds more loan characteristics to the basic equation . As
before, these do not change the coefficients for the basic variables. Whether
the rate was fixed or variable had no effect on the probability of denial .
The effect of longer loan terms also was not statistically significant .

The

presence of a gift or grant reduced the probability that the loan would be
denied, and the effect approached statistical significance .

Applications that were not made under special programs were denied more
frequently , but the effect was not statistically significant .

Many of these

special programs are offered by the Massachusetts Housing Finance Agency.
These are intended to encourage lending to lower - income and minority borrowers
and to first - time home buyers. Another large group consisted of First - Time
Homebuyer programs offered by various banks .
Personal Characteristics

Additional personal characteristics do not alter the basic results
( Table B11 ) .

The age , sex , and number of dependents of the applicant have no

significant effect on the probability of denial . The variable representing
marital status approached statistical significance , with applicants who were
not married facing a higher probability of being denied . a mortgage , other
things equal .
Lender Standards

The equations in Table B12 attempt to take account of differential

lender standards .

It has been suggested that black and Hispanic applicants go

disproportionately to institutions that have higher than average credit
standards and , therefore, higher denial rates for both whites and minorities .

This is a controversial hypothesis , since it implies that minority mortgage
applicants act against their own best interest; alternatively , the

institutions with higher denial rates may be more aggressive in soliciting
minority applications .

A " tough" lender variable was constructed by estimating the basic
equation with a dummy variable for each lender over the sample of white
applicants only . The coefficients of these dummies were then used to create a
dummy variable indicating that the lender had " tough " standards and the

140

equation was estimated over the entire sample of applicants .

As can be seen

from Table B12, the " tough" lender variable is not statistically significant
and does not alter the results . The inclusion of separate dummy variables for
each lender when the equation is estimated over the entire sample does reduce
the coefficient of the race variable ; but it remains large and statistically
significant .
Separate Equations for White and Minority Applicants

An implicit assumption underlying the equation in Table 5 is that
lenders treat white and minority applicants the same except for their race .
In other words , lenders accord the same weights to credit history, obligation
ratios , location risk , and all the other characteristics of white and minority

applicants. An alternative possibility is that lenders assess the
creditworthiness of minorities quite differently than they do that of whites ,

so credit history or obligation ratios are viewed differently if the applicant
is black or Hispanic .

Correlation Matrix

Table B14 presents a matrix showing the correlations among the variables
used in the basic equation . As can be seen , multicollinearity between any two
independent variables is not driving the results , because no two variables are
strongly correlated .

141

Appendix Table 81
Alternative Specifications of Probability of Mortgage Denial
Ability to support Loan

Basic Equation
Coefficient
( t - Statistic)

Variable

Constant

Coefficient
( t - Statistic)

-6.61
( -17.0)

-6.17
( -13.0 )

.47
( 3.1 )

.33

.33
( 9.8)

Ability to Support Loan

Housing Expense / Income
Total Debt Payments / Income

Net Wealth
Income

Liquid Assets
Base Income / Total Income

Risk of Default

Consumer Credit History

( 9.8 )

Mortgage Credit History
Public Record History
Probability of Unemployment
Self - Employed
Potential Default Loss

Loan /Appraised Value

.58

.60

( 3.2 )

.58
( 3.6 )

.57
( 3.4 )

Denied Private Mortgage Insurance
Rent /Value in Tract

Loan Characteristics

Two- to Four -Family Home
Personal characteristics

.68

Race

( 5.0)
Number of Observations
Percent Correct Predictions

3062

.70

( 5.1 )
3030

'See notes to Appendix Tables following Appendix Table B14 , for variable definitions and
sources .

The number of applicants with a probability of denial greater than 50 percent who were
denied , plus the number of applicants with a probability of approval greater than 50
percent who were approved , as a percent of the total sample.

142

Appendix Table B2

Alternative Specifications of Probability of Mortgage Denial
Risk of Default · Credit History
Basic Equation
Coefficient
( t - Statistic )

Variable

Constant

-6.61
( -17.0)

( -17.1 )

.47
( 3.1 )

( 3.0)

-6.04

Coefficient

( t-Statistic)
-6.68

( -16.4)

Ability to Support Loan
.66

.48

( 3.2)

Total Debt Payments / Income
Net Wealth

Risk of Default
.33
( 9.8)
.38
( 3.0)

.33
( 9.8 )

1.22
( 7.1 )

Consumer : More than Two Slow Accounts

Consumer : Delinquencies
Consumer : Serious Delinquencies
Mortgage: No History

.30
( 1.8)

Mortgage: One or Two Late
Mortgage : More than Two Late
.09
( .5 )

Mortgage : Prior History
Probability of Unemployment

.09

( 3.3)

( 3.2 )

.58
( 3.2)
4.70

( 3.2)

Self - Employed

Potential Default Loss
Loan / Appraised Value

Denied Private Mortgage Insurance

.60

.58
( 3.2)

( 9.6)

RentNalue in Tract
Loan Characteristics
.58

( 3.6)

.58
( 3.6)

Race

.68
( 5.0 )

.67
( 4.8)

.69
( 5.0 )

Number of Observations

3062

Two - to -Four- Family Home

.58

Personal Characteristics

Percent Correct Predictions

143

Appendix Table B3

Alternative specifications of Probability of Mortgage Denial
Risk of Default · Years on Job ; Co - signer
Basic Equation
Coefficient

( t - Statistic )

Variable

Constant

-6.61
( -17.0)

-6.62
( -16.3 )

Ability to Support Loan
.47

Housing Expense / Income

( 3.1 )

.44

( 2.9)

Total Debt Payments / Income
Net Wealth

Risk of Default
( 9.8 )

.33
( 9.9)

.58
( 3.2 )

.59
( 3.2 )

.58
( 3.6 )

.60
( 3.6)

.68
( 5.0)

.71
( 5.1 )

.33

Mortgage Credit History
Public Record History

Probability of Unemployment
Self- Employed
Years on Job

Presence of Co -signer

Potential Default Loss

Denied Private Mortgage Insurance

Personal Characteristics
Race

3062
89

Number of Observations
Percent Correct Predictions

2997

144

Appendix Table B4

Alternative Specifications of Probability of Mortgage Denial
Default Loss · Private Mortgage Insurance

Coefficient

Excluding
PMI Denials
Coefficient

( t - Statistic)

Basic Equation

Variable
Constant

-6.61

( -17.0)

-6.57
( -17.4 )

-6.61
( -16.9)

Ability to Support Loan
Housing Expense / Income

.47
( 3.1 )

.48

( 3.2 )

Total Debt Payments / Income
( 7.1 )

Risk of Default
.33
( 9.8 )

( 9.8 )
( 3.1 )

nennen

Net Wealth

.33
( 9.9 )

( 3.5 )
( 2.5 )

( 3.4 )

.62
( 3.2 )

( 3.1 )

.68
( 3.5 )

.58
( 3.6 )

( 4.2 )

( 3.6 )

.68
( 5.0 )

( 5.5 )

.58
( 3.2 )

Rent /Value in Tract

Loan Characteristics

Race

.59

.71

3062

Number of Observations

.69
( 5.0 )

2983

145

Appendix Table B5

Factors affecting Probability of Private Mortgage Insurance Denial
Coefficient

( t- Statistic )

-7.31

Constant

( -5.6 )

-7.30

( -5.6)

Ability to Support Loan
.44

.43

( 1.3)

.20
( 2.7)

Risk of Default

Loan /Appraised Value

1.53

1.72

Rent /Value in Tract

( 1.9)
-1.13
( -.8)

-1.78

( 2.1 )
( -1.0)

Minority Population Share
Loan Characteristics
.55

( 1.7)
.59

Race

( 2.0)

Number of Observations
Percent Correct Predictions

723

.52

( 1.6)
.34
( 1.0 )
723
90

146

Appendix Table B6
Alternative specifications of Probability of Mortgage Denial
Potential Default Loss - Tract Characteristics
Coefficient

( t-Statistic)

-6.61
( -17.0 )

Constant

-6.92
( -15.9)

Ability to Support Loan
.47

Housing Expense / Income

.47

( 3.1 )

( 3.0)

.33
( 9.8)

.34
( 9.7)

Total Debt Payments / Income
Net Wealth

Risk of Default
Consumer Credit History
Mortgage Credit History

Public Record History

Probability of Unemployment
Self - Employed

Potential Default Loss
Loan / Appraised Value

.58
( 3.2 )

.59
( 3.1 )

Denied Private Mortgage Insurance
Rent/Value in Tract

Housing Units Boarded Up
Housing Units Vacant
Housing Value Appreciation

.58

Race

Number of Observations
Percent Correct Predictions

.63

( 3.6)

( 3.7)

.68
( 5.0 )

.62
( 3.9)

3062

2788

147

Appendix Table 87

Alternative specifications of Probability of Mortgage Denial
Potential Default Loss • Foreclosure Rate
Basic Equation
Coefficient
( t -Statistic )

Variable

Constant

-6.61
( -17.0)

-6.66
( 17.1 )

.47
( 2.8)

.47
( 3.1 )

Ability to support Loon

Total Debt Payments / Income
Net Wealth

Risk of Default

.33

Consumer Credit History

.33

( 9.8)

.58
( 3.2 )

.58
( 3.6)

( 3.3 )

.68
( 5.0)

( 4.9 )

Mortgage Credit History
Public Record History
Probability of Unemployment

Self - Employed

Potential Default Loss
Loan / Appraised Value

Denied Private Mortgage Insurance

Rent /Value in Tract

Race

Number of Observations
Percent Correct Predictions

3062

.55

.67

3062
89

148

Appendix Table B8

Foreclosure' Rates and Racial Composition of city of Boston Planning Districts
Percent

Total
Foreclosures
as a Percent

Foreclosures
as e Percent

Planning District

Percent

Black and
Nispanic in
Population

East Boston

.37

.40

South Boston

.34

.37

1.9

Mattapan

.33

.37

94.5

Charlestown

.32

.35

2.1

18.9

Fenway /Kenmore

.24

.32

17.6

South Dorchester

.18

.23

46.7
36.8

North Dorchester

. 18

.16

Allston /Brighton

.18

15.5

Jamaica Plain

.17

.18

43.2

Roxbury

.16

.23

90.2

Back Bay /Beacon Hill

.14

.29

5.5

West Roxbury

.14

.15

3.2

South End

.13

.17

52.3

Central

.12

.13

7.0

Hyde Park

.07

.08

27.1

Roslindale

.05

18.1

city of Boston

.27

.22

34.3

' Foreclosures are for the years 1988 through 1990 .

All sellers are persons; commercial entities are excluded .
Source : Foreclosures were supplied by Banker & Tradesmen ; housing units
are from 1990 Census of Population and Housing .

149

Appendix Table B9
Alternative specifications of Probability of Mortgage Denial
Potential Default Loss - Tract Dummy Variables
Basic Equation
Coefficient

Variable

( t -Statistic)

-6.61
( -17.0 )

Constant

Ability to Support Loan
.47
( 3.1 )

.63
( 3.3)

.33
( 9.8)

( 8.3 )

Total Debt Payments / Income
Net Wealth

Risk of Default

Consumer Credit History

.47

Mortgage Credit History
Public Record History
Probability of Unemployment

.58

.81

( 3.2 )

( 2.5)

.58
( 3.6 )

( 2.6)

Rent /Value in Tract
Census Tract

Loan Characteristics

Race

.68

( 5.0 )
Number of Observations
Percent Correct Predictions

3062

.54

.93
( 4.1 )

3062
n.a."

* Constant is included in the dummy variables for the census tracts . These are not shown
because they are . so numerous .

' The large number of variables in this equation required a more powerful computer and the
regression package available did not calculate percent correct predictions.

150

Appendix Table B10
Alternative specifications of Probability of Mortgage Denial
Loon Characteristics
Basic Equation
Coefficient
( t - Statistic )

Variable

Constant

-6.61
( -17.0)

-6.57
( -11.8)

Ability to support Loan
Housing Expense /Income

.47

( 3.1 )

( 3.2)

Total Debt Payments / Income
Net Wealth

Risk of Default

Consumer Credit History

.33

( 9.8)
Mortgage Credit History

.33

( 9.8)

.35

( 3.0)
Public Record History

Probability of Unemployment

.09
( 3.3)

.08
( 3.0)

.58
( 3.2)

.62
( 3.4)

.58
( 3.6)

.61
( 3.7 )

.68
( 5.0)

( 5.2)

Self - Employed
Potential Default Loss

Loan Appraised Value
Denied Private Mortgage Insurance
RentNalue in Tract

Lom Characteristics
Two - to -Four - family Name
Fixed -Rate Loan

Not a Special Loan Program
Ten of Loan

Gift or Grant in Down Payment

Personl Characteristics
Race

Mumber of Observations

Percent Correct Predictions

3062

.73

3055

151
)
Appendix Table B11
Alternative Specifications of Probability of Mortgage Denial
Personal Characteristics
Basic Equation
Coefficient

Variable

( t - Statistic )

Constant

-6.61
( -17.0 )

-6.88
( -13.4 )

.47
( 3.1 )

.46
( 3.0 )

Ability to Support Loan

Total Debt Payments / Income
Net Wealth

Risk of Default
.33

( 9.8)

.33

( 10.0)

Mortgage Credit History
Public Record History

Probability of Unemployment
Self-Employed
Potential Default Loss
.58
( 3.2 )

Loan / Appraised Value

.63
( 3.3 )

Denied Private Mortgage Insurance
Rent /Value in Tract

Loan Characteristics
Two - to - Four - Family Home

.58

( 3.6 )

( 3.6)

.68
( 5.0 )

.65
( 4.7)

3062
89

3027

Personal Characteristics
Race

Age
Sex

Number of Dependents
Marital Status ( Not Married = 1 )

Number of Observations
Percent Correct Predictions

152

Appendix Table B12

Alternative specifications of Probability of Mortgage Denial
Lender Standards
Basic Equation
Coefficient
( t -Statistic )

Variable
Constant

-6.61

Coefficient

( t - Statistic )

-6.65
( 16.9 )

( -17.0 )
Ability to Support Loan
Housing Expense / Income

.47
( 3.1 )

.47

( 2.8)

Total Debt Payments / Income

Risk of Default
Consumer Credit History

.33
( 9.8 )

( 9.8 )

Mortgage Credit History
( 3.1 )

Public Record History
Probability of Unemployment

naturgement
system9

( 6.6)
Net Wealth

.39
( 10.0 )

.08

.09

( 3.3 )

( 2.7)

Self - Employed
( 2.8 )
Potential Default Loss

Loan /Appraised Value

.58
( 3.2 )

( 3.2 )

.67
( 3.2 )

.68
( 3.5 )

( 3.5 )

( 2.6)

( 3.6)

( 3.5 )

Denied Private Mortgage Insurance
Rent / Value in Tract

.56

Loan Characteristics
Two - to -Four - Family Home

.58

.64

( 3.6 )
Lender
Tough Lender

Lender Dummy

Personal Characteristics
Race

.68

( 5.0)

( 5.0)
Number of Observations

3062

.54

( 3.4 )

3061

Percent Correct Predictions
* Constant is included in the dummy variables for the lenders .

These are not shown because they are so numerous .

153

Appendix Table B13

Alternative specifications of Probability of Mortgage Denial
White
Coefficient
( t - Statistic )
Constant

Black and Hispanic

-6.61
( -17.0 )

-6.56
( -17.0 )

-6.22
( -14.6)

-7.33
( -7.6 )

.47
( 3.1 )
.04
( 6.6 )

.51
( 3.4 )
.05
( 6.6 )

.44
( 2.3 )
.04
( 4.9 )

.46
( 1.9 )

Ability to Support Loan

Housing Expense / Income
Total Debt Payments / Income
Net Wealth
Risk of Default
Consumer Credit History

.33
( 9.8 )

.35

( 10.6)

.32
( 7.5 )

.33

( 6.1 )

Mortgage Credit History

Public Record History
Probability of Unemployment
.65

Self - Employed

( 3.1 )
Potential Default Loss

Loan / Appraised Value

.79

.63
( 3.1 )

.56
( 2.9 )

( 1.2 )

.76
( 4.8 )

.78
( 3.4 )

( 1.7)

3062

2340

.68

Two - to - Four - Family Home

.58
( 3.6 )

.38

Personal Characteristics
.68

( 5.0 )

Number of Observations
Percent Correct Predictions

3062

60–893 0 - 92 - 6

722
81

B14
Table
Appendix
Matrix
Correlation

21
4

HEXP

DTOI

NETW

RACE

CONSPAY

MORTPAY

8
PUBREC

LTV

11
Rent

PMI

to
2
4

13
SELF

1.00

Race

Housing
IExpenses
/ ncome

Total
Debt
IPayments
/ ncome

.06

1.00

.07

.37

1.00

Wealth
Net
5

Consumer
Credit
History

.20

.01

.06

..03

1.00

History
Credit
Mortgage

.14

.06

..11

.15

.14

.05

1. 0

.01

.31

.07

..01

.03

..01

..02

.03

.01

1.00

.14

.04

.08

..07

.05

.12

.05

..01

.10

..02

.04

.03

.02

.05

.08

-.03

.07

.05

.07

.01

.15

..01

.01

..01

.07

.06

.04

.05

.07

.11

.09

1.00

.02

.12

..02

..05

.02

.16

..03

..02

.03

Public
Record
History

Probability
Unemployment
of
..05

Value
ALoan
/ ppraised
/Value
Rent
10
Tract
in

.23

:3062
Observations
of
Number

1.00

.001

1.00

.02

1.00

..0003

154

Denied
11
Private
Mortgage
Insurance
.10

1.00

155

Variable Definitions and sources
Question numbers refer to the questions listed in Appendix A.
Data from lenders ' MMOA reports were supplied by the lenders as
part of their normal Home Mortgage Disclosure Act filing.

Dependent Variable

: 1 if applicant was denied a mortgage

Housing Expense / income
Total Debt Payments /Income
Net Wealth
Income

Liquid Assets
Base Income / Total Income

Consumer Credit History

Mortgage Credit History

Public Record

Consumer : Insufficient History
Consumer : One or Two Slow Accounts

: 1

Consumer : More than Two Slow Accounts
Consumer : Delinquencies

: 1

Consumer : Serious Delinquencies

: 1

Mortgage : No History

: 1

Mortgage: One or Two Late

: 1

Mortgage : More than two Late

: 1

Geographic Profile of Employment and Unemployment, 1989
= 1 if applicant was self- employed

156

oan / Appraised Value

value of loan amount from original HNDA report

enied Private Mortgage Insurance

entNalue in Tract

rental income divided by value of rental
housing stock in census tract in which property
was located . Derived from U.S. Bureau of the
Census , 1990 Census of Population and Housing. Summary

ousing Units Boarded Up

percent of housing units in census tract in

Tape File 3 ( 1990 Census)
which property was located that were boarded up
Source: 1990 Census

pusing Units Vacant

percent of housing units in census tract in
which property was located that were vacant
Source: 1990 Census

pusing Value Appreciation

percent change in the median value of owner -occupied
housing between 1980 and 1990 in the census tract in which
in which the property was located
Source : Derived from 1990 Census and 1980 Census of

inority Population Share ( > 30 Percent )

preclosure Rate
ansus Tract Dummy Variable
: 1 if purchasing a two- to four- family home
ixed -Rate Loan

at a Special Loan Program
rm of Loan
ft or Grant in Down Payment
inder Dummy Variable
ice

= 1 if applicant was black or Hispanic,
= 0 otherwise ( lenders ' MMOA report)

mber of Dependents

157

References

Avery , Robert B. and Thomas M. Buynak . 1981. " Mortgage Redlining : Some New

Benston , George J. , Don Horsky, and H. Martin Weingartner . 1978. " An

Black , Harold , Robert L. Schweitzer , and Lewis Mandell . 1978. " Discrimination

Bradbury , Katharine L. , Karl E. Case , and Constance R. Dunham . 1989 .

Canner , Glenn B. 1982. " Redlining : Research and Federal Legislative Response . "

Canner, Glenn , Stuart A. Gabriel , and J. Michael Woolley . 1991. " Race ,

Dedman , Bill and others . 1988. " The Color of Money . " A Compendium . The
Federal National Mortgage Association . 1992. Fannie Mae Guides, Vol. 1
Gabriel , Stuart A. and Stuart S. Rosenthal . 1991. " Credit Rationing , Race and

Galster , George C. 1977. " A Bid- Rent Analysis of Housing Market

158
King , Thomas A. 1979. " Redlining : A Critical Review of the Literature with

1980. " Discrimination in Mortgage Lending : A Study of Three Cities . "
Federal Home Loan Bank Board , Working Paper No. 91.
Schafer , Robert and Helen F. Ladd . 1981. Discrimination in Mortgage Lending .

Yinger , John . 1986. " Measuring Racial Discrimination with Fair Housing Audits :

159

ACORN

TAKE
The

MONEY AND

THE

Siphoning

Deposits

RUN :

from Minority

June 4 ,

1992

Association of Community Organizations for Reform Now
Organizing & Support Center: 1024 Elysian Fields Avenue, New Orleans, Louisiana 70117 • 504-943-0044 FAX 504-944-7078

160

Summary
Analysis of bank deposit and home mortgage lending data in 14
major American cities reveals that the nation's publicly chartered
depository institutions reinvest the deposits of white communities

in predominantly white neighborhoods , yet disinvest , or siphon ,
the neighborhood deposit base of minority communities .

Nationally , the study revealed that , for every dollar on
deposit in predominantly minority neighborhoods , about 4 cents was
loaned for mortgages in those same neighborhoods in 1989. By

contrast , for every dollar on deposit in predominantly white
neighborhoods , nearly 8 cents are reinvested in those same
neighborhoods. This result is particularly striking given the

spate of branch closures in minority neighborhoods over the past
decade , leaving many residents of predominantly minority

communities no option but to do their banking elsewhere.

In light of the serious problems of credit availability in
minority communities made clear by this and other studies , ACORN

has called on leaders in the banking industry to participate in a
" Neighborhood /Bank Summit " next month in New York city to explore
productive ways to promote community reinvestment .

161

Introduction

This study measures the home mortgage investment of deposits
held in insured financial institutions in neighborhoods of

different racial and ethnic compositions in 14 major American
cities .

In the analysis , a " loan -to - deposit " ratio is computed for
neighborhoods of differing racial and income compositions. A loan
to-deposit ratio permits comparisons of depository institution
investment in residential mortgages on the basis of the
neighborhood characteristics of the deposit base of the examined
depository institutions .

162

FINDINGS

This study has three principal findings :

( 1) For the 14 cities under study combined, only 4 cents for
every dollar of deposits held in minority neighborhoods was loaned
by depository institutions in those same neighborhoods for
mortgages or other housing related credit . By contrast; nearly 8
cents for every dollar in deposits in predominantly non -minority
neighborhoods was loaned in those neighborhoods by depository
institutions for mortgages and other housing-related credit .

( 2 ) Wide disparities remained when comparing neighborhoods of
the same income , but different racial profiles . Nationwide , only

( 1)

Loan - To - Deposit

Ratios,

By_Race of Neighborhood

163

gentrification .
1

Other integrated neighborhoods , by contrast ,

s , perhaps reflecting
minority neighborhood
receeive lesse loans than ht
hborhood

whit , middl - class flig

from a neig

For example , Brooklyn has the highest loan- to-deposit ratio
in minority neighborhoods of any city under study . However , this
may simply reflect on housing market conditions in the city of New
York , so we cannot conclude from that figure alone that credit
availability in minority neighborhoods is greater in New York than

elsewhere .

In fact , the disparity between neighborhoods of

different racial profiles is least in Philadelphia , as illustrated
below .
TABLE 1 : LOAN - TO - DEPOSIT RATIOS ,
NEIGHBORHOODS
CITY

< 25%

BROOKLYN
CHICAGO
DALLAS
DETROITA
NEW ORLEANS
PHILADELPHIA
ST. LOUIS

17.3 %

WASH , D.C.

MIN

7.6 %

7.5 %
**

BY RACIAL COMPOSITION OF

> 7596 MIN

11.8%
2.4 %
1.3 %
0.39%

DISPARITY

1.5
3.2
5.8
0.0
14.5
0.72

14.5 %
3.6 %
4.2 %
10.2%

ADetroit has very few predominantly white tracts
CITY
BOSTON
KANSAS CITY
LITTLE ROCK
MILWAUKEE

< 25%

MIN

> 25%

12.3 %
5.9 %
9.8%

MINN-ST PAUL
PHOENIX

MIN

DISPARITY

164
( 2)

Loan - to - Deposit Ratios for middle - Income

Neighborhoods , By Race
The differential between minority and white neighborhoods
actually increases when comparing middle - income neighborhoods of
different racial profiles . It was not possible to compare low
and moderate - income neighborhoods of differing racial

compositions, because in most cities there are few or no low- or
moderate - income white census tracts . Similarly , comparison of
high - income neighborhoods of different racial compositions is made
difficult by the paucity of high - income minority tracts in most
cities .
TABLE

LOAN- TO- DEPOSIT

NEIGHBORHOODS ,
CITY

BROOKLYN
CHICAGO
DALLAS
DETROITA
NEW ORLEANSA

RATIOS

MIDDLE - INCOME

BY RACIAL COMPOSITION OF NEIGHBORHOOD
< 25 %

MIN

7.0 %

> 75 %

MIN

DISPARITY

10.3

8.0 %

4.8 %
2.1 %
.6 %

PHILADELPHIA
ST. LOUISA

3.1 %

1.2 %

2.6

WASH, D.C.

20.1 %

3.3 %

6.1

CITY

< 25 %

BOSTON
KANSAS CITY
LITTLE ROCK

9.6 %
7.8 %

.4 %
.6 %

MILWAUKEE
MINN- ST PAULA

3.7 %

.4 %

9.3

PHOENIX

3.3 %

1.5 %

2.2

21.7 %

MIN

> 25 %

1.5
13.3

MIN

DISPARITY

AToo few middle -income minority or integrated tracts in these cities.
( 3)

The Case of Philadelphia

Philadelphia was the only city studied in which the loan - to
deposit ratios in predominantly minority neighborhoods actually
exceeded that for predominantly non -minority neighborhoods .
There are several possible explanations of this data :

un
5

--a severely depleted branch network in predominantly
minority neighborhoods, thus understating the actual dollar amount
of deposits held by residents of these neighborhoods .

165

Another possible explanation for the data is that Philadelphia
banks are performing much better in predominantly minority markets
than are their peers in other cities . If this were to be the
case , this might demonstrate that reversing the patterns uncovered
by this study is indeed possible .

There is substantial anecdotal , and some statistical information
to support the claim that the Philadelphia data in fact represents
better performance by area depository institutions . A recent New
York Times article featured Philadelphia as the pioneering city
where community groups and banks have successfully increased the
flow of mortgage credit to inner-city neighborhoods .
Indeed , in 1989 , four of the largest banks in Philadelphia
offered special low - income mortgage programs tailored to minority
communities , with relaxed underwriting criteria . Continental Bank
and Fidelity Bank have community Reinvestment Act ( CRA ) agreements
with Pennsylvania ACORN , and , working with ACORN's loan counseling
I

program in Philadelphia , originated 311 home loans , worth over $ 7
million in mortgages to low- and moderate - income minority
homebuyers in 1989 .

Two other major banks , Mellon and CoreStates offered low- income

mortgage programs which resulted in 253 loans worth $ 5 million .
Together these loans accounted for 21.2 % of the dollar volume of
loans extended in predominantly minority neighborhoods .

loans are not included , the loan-to-deposit ratio in predominantly
ratio for predominantly non-minority neighborhoods .

minority neighborhoods drops to parity with the loan to deposit
While it is impossible to determine with certainty the
explanation for the Philadelphia data , this study provides

encouraging evidence that banks and community groups can indeed
turn the numbers around .

Conclusion

The fact that depository institutions lend substantially more
dollars in non -minority than in minority neighborhoods is in
itself not surprising . However , the fact that the loan- to-deposit

ratio in minority neighborhoods is substantially less than that in
predominantly white neighborhoods speaks to a differential
commitment of the industry as a whole to extending credit in all

the communities they are chartered to serve , including minority
neighborhoods .

166

APPENDIX :
Data

METHODOLOGY

Sources

This study is based on June 30 , 1990 branch deposit data

obtained from the FDIC's publication , Operating Banks and
Branches , which includes deposit data for every branch of

commercial banks and savings banks insured by the Bank Insurance
Fund ( BIF ) . Also included is data obtained from the Office of
Thrift Supervision's Survey of Branch Deposits, also as of June

30 , 1990 , which contains analagous data for thrifts now insured by
the Savings Association Insurance Fund ( SAIF ] .

The survey covers

the period June , 1989 - June , 1990 .

Deposit

Information

The total dollar number of deposits at the branches of
insured depositories in every central city under study was
included , including IPC deposits , public deposits , and all other
deposits .

However , deposits held at branches in non-residential

areas , or areas that were determined to be primarily downtown
commercial centers were excluded . Many institutions hold a
substantial portion of their total deposits in a given city at a

main branch , usually in the downtown area. Since such deposits
cannot be presumed to have been placed primarily by neighborhood
residents , deposits at these branches are not considered .
Categories

of Analysis

Each branch was coded for the income and racial profile of
the immediate area surrounding it , using publicly available census
data . For 8 cities , three racial categories were used :
predominantly minority ( > 75% of the residents were African

American , Asian American, American Indian , or Latino ) ; integrated
( between 26-74 % of neighborhood residents are minority ) ; and white
( < 25 of neighborhood residents are minority ) .

These cities were :

Brooklyn , Chicago , Dallas , Detroit , New Orleans , Philadelphia , St.
Louis , and Washington , D.C.

167

minority ( > 26-100 % minority) and predominantly non -minority ( < 25 %

minority) . These cities were : Kansas City, Mõ ; Little Rock , AR;
Milwaukee , WI ; Minneapolis and St. Paul , MN ; and Phoenix , Az .

Predominantly minority : > 75 % minority
Integrated : 26-74 % minority
Predominantly non -minority : < 25 % minority

Again , for the six cities with an insufficient number of
neighborhoods where residents comprise more than 75 % of the
population , only two racial categories were employed : " Combined

integrated and minority " and " Predominantly non -minority . "
Data Limitations &

Methodological

Issues

Deposit Data

The study was constructed given existing data limitations
concerning the source of deposits held by depository institutions .

Insured depository institutions do not report information on
the home address of individual depositors. This study uses the

only available proxy of geographic deposit location to measure the
deposit base in neighborhoods of different racial and income
profiles . That proxy is the geographic location of bank branches

where deposits are booked by the insured depository institution .

168
predominantly non -minority neighborhoods .

Because of the paucity of bank branches in many minority ,
urban areas , more minority individuals are likely to bank at
branches in commercial or downtown areas , where they work , thereby

understating the deposit base in predominantly minority
neighborhoods .
HMDA Data

HMDA data is the only available data with which to assess
capital flows by neighborhoods of different racial and income
compositions- : Other lending , consumer and commercial, obviously
occurs but , financial institutions are only required to report
home mortgage lending on a geographic basis .
This study did not consider the mortgage lending of
institutions with assets of under $ 10 million , because such

institutions are currently exempt from HMDA . In addition , mortgage
banks , were not included in the study , because they do not take

deposits . While the inclusion of lending by mortgage banks would
undoubtedly have raised the overall loan - to -deposit ratio , it is
doubtful that the inclusion of either of these types of lenders
would significantly alter the relationship between lending in

predominantly minority and predominantly white neighborhoods .

NOTE :

Statistical data and modeling prepared by Drs . William

Milczarski and Steve Johnston , Assistant Professors of Urban
Planning at Hunter College , CUNY .

LOF
: 4OEPOSIT
OAN
,1-DTTABLE
RATIOS
KACE
B
CITIES
Y
NEIGHBORHOOD
Boston

Brooklyn

Chicago

Dalbs

Kansas
C.

Detroit

MLittle
| ilwaukee
Rock

/STP
Minn

.
Ort
New

.
Philadel

Phoenix

D
,C
Wasth

Louis
St

>75NORMY
%M
$Loans

Dseposits
Ratio
/Deposit
Loan

8130
0.5596

1984
0.9696

151
0

878
1.0396

1098

447
2.6896

25
2758

647
2478
26.1196

0.9196

5.0196

121

2.9996

180
3015
5.9796

4042

270
7549
3.5896

298
3457
8.6296

142
3348
4.2496

356
1.9796

31140
0.4
596

718

341
3636

702
3576

1477

34
1253

0.8496

0.9496

0.5496

Too

5.8996

19.6396

1221

8061

2.7396

612
9.8096

188
3193

115
1484
7.7596

944

1.8096

2.7196
!

3001

237
2202
10.7696

2:
TABLE

Brooklyn Chicago

Dalbs

STP
Milwaukee Min
Rock
Little

Kansas
C

Detroit

Philadel
.

.
Orl
New

Phoenix

Louis
St

,DC
Wash

MOOLE
INCOME

131

96
12301
7.8096

163
5.5296

95
2577

1578
3.6996 S.5196

1136
0.3596

171
5585
3.0696

931
2826
3.2996

235
0.00961

1185
1.1096

2427
1.1196

406

1.4896

78
3208
2.4396

406
)
1.7296

SS
262
)
20.9996

28
1050
2.6796

169

OF
TNEIGHBORHOOD
1-DLOAN
,BRATIOS
CITIES
INCOME
AND
RACE
OEPOSIT
Y4
Boston

170

TAKING IT TO THE BANK :

Poverty, Race, and Credit in Los Angeles
A Report to the City of Los Angeles

September 1991

Prepared by the Western Centeron Law and Poverty

Gary Dymski, Assistant Professor
Departmat of Economics
University of Southes California

Jobo Veitch , Assistant Professor
Departmut of Economics
University of Soutbos Culornia

Michelle White , Executive Director

Jab Bonatag Canprus
of Southern Caloria

171

ACINOWLEDGEMENTS

Masy Dom individuals have contributed to this study thus the thru mother listed
on the title page. Fint and foremost we would be to acbaowledge the trale nuncarch

asistance of Nicole Rivas, valor at the University of Southern California . Mary Bur
dick , Director of the Waters Cantar on Low and Poverty, provided @emplary editorial
advice. Also havaluable u tuared mistants were Aathony Dahm , Betty Daha, Con .

stanjin Panis, and Lin Monique White. The wall busidus surveys won orchestrated

by Sharon Murray, Anthony Doroho, and Lape Garcia Daaie Kinsey of the Western
Centes provided typing wad word procening rapport.

172
CONTENTS

PART I

Overview and Context of the Study

EXECUTIVE SUMMARY AND OVERVIEW
A. PECUTIVE SUMMARY

B. MAJOR FINDINGS ON RESDENTIAL LENDING
C. MAJOR FINDINGS ON BANKING SERVICES ....

10 .

D. MAJOR FINDINGS ON ECONOMIC DEVELOPMENT

11.

E. FINDINGS ON LENDONO POR
12 .

15 .

16 .
17 .
10 .

CHAPTER 1. Contexts of this study :
A. INTRODUCTION

20 .

B. MORTGAGE LENDING IN URBAN COMMUNITIES :
20 .

24 .

29 .

35 .

CHAPTER 3. The design and conduct of this study
CO .

A. INTRODUCTION
B. THE SCOPE AND CONDUCT OF THE STUDY

40 .

C. A CAVEAT ON ASSESSING BANKING NEEDS
4 .

173

PART

Residential Lending

CHAPTER 8. Patterns of residential loan dows In Los Angeles
A INTRODUCTION
B. BIGELIGHTS OF CEAPTER 3

80 .
82.

C. DESCRIPTION OF THE DATA:
83 .

D. LOAN FLOWS BY INSTITUTION TYPE

62.

L. LOAN FLOWS BY INCOME QUINTILES

64 .

F. LOAN FLOWS AND
.. 72.
76.

77 .

CHAPTER 4. Race, residential lending, and become
A. INTRODUCTION

79.

B. EIGELIGHTS OF THE ANALYSIS

80 .

C. PATTERNS OF RESIDENTLAL SEGREGATION

... 81 .
85.

87.
94.

CHAPTER 5. Ranking Residential Lending Performance
A. INTRODUCTION
B. EIGELIGHTS OF THE ANALYSIS

96 .

99.

C. THE COHORT RANKING METHOD ... ,

100 .

D. ON -GOING MONITORING

109 .

CHAPTER 6. Race, residential lending, and income:
A. INTRODUCTION

B. TITLE TRANSFER DATA
C. AGGREGATE TITLE TRANSFER LENDING PATTERNS

D. MEDLAN TRANSACTION CHARACTERISTICS

112 .

.. 113.
114 .

174

PART DO

Financial Service Lending for

Economic Development and Affordable Housing
CEAPTER 7. Bankdng markets and transactions services
123.
124 .

128.

128 .
130 .

131 .

133.

138.

139.

141.

CHAPTER 8. Lending for economic development in Los Angeles
142 .

A. INTRODUCTION

B. HIGHLIGHTS OF THE ANALYSIS

143 .

C. DESCRIPTION OF THE DATA

144 .

D. THE SCOPE OF CREDIT SUPPLY
.. 146.
E. THE SCOPE OF CREDIT SUPPLY
147 .

F. LENDERS' CREDIT DECISIONS

149.

G. PERCEPTIONS FROM SMALL BUSINESS MEETINGS

152 .

CHAPTER 9. Credit and affordable housing development:
A. INTRODUCTION

165 .

B. BIGELIGHTS OF THE ANALYSIS

157.

C. DESCRIPTION OF THE DATA
D. LENDERS, MULTIFAMILY HOUSING

189.

160 .

E. CREDIT FLOWS TO DEVELOPERS
165 .

175

CHAPTER 9. Credit and affordable housing development:
168 .

171.

174

CHAPTER 10. Ranking bank services and
A. INTRODUCTION ........

177 .

B. TWO APPROACHES TO RANKING :
178 .

182 .
183 .

185 .
187.

PART IV
Conclusion

CHAPTER 11. Policy Recommendations
A. INTRODUCTION

...

192 .

B. A LINKED BANKING PROGRAM

194 .

C. INNOVATIVE IDEAS FOR REINVESTMENT .....

204 .

D. RECOMMENDATIONS INVOLVING CHANGES
209 .

176

CONTENTS

APPENDICES TO THE STUDY
CHAPTER 8. Patterns of residential loan Bows in Los Angeles
Appendiks SA

DMDA mortgage data summary tables
Appade 8B

Ceasu Irast level demographic data by Counci District.

CHAPTER 4. Residential lan ding, race , and income
Appender
Complete set of EIMDA dats , 1981-89

CHAPTER 5. Ranking schemes for Briancial bastitutions
Appendix b - A to Chapter 3

CHAPTER 6. Residential lending, race , and income
Appendix to Chapter 6

CHAPTER 7. Banking markets and banking service in Los Angeles
Appendixs 7 - A
Summary Tables for Data drama from Financial Lostitutions'

177

Appedts T - B

Summary Tables for Data drawn from Financial Institutions'
Responses to the 1990-91 Bank Survey

Appendis FC
List of Financial Lastitutions returning 1989 CRA Statements
Appadk T - D

List of Financial Institutions noturning the 1990-91 Bank Survey
Appede T

The 1990-91 Bank Sarvey Questionnaire

CHAPTER 8. Lending for economic development in Los Angeles
Appendix to Chapter 8

CHAPTER 9. Credit and affordable housing development
Appendiks -A

Summary Tables for Data drawn from Responses to
1990-91 Developer Survey

Appendik AB
The 1990-91 Sarvey Questionnaire for Bousing Developers

CHAPTER 10. Ranking bank services and non - residential lending
Appendir

Suggested Application Form for Financial Institutions
under a Linked Banking Program

vü

178

PART I

Overview and Contexts of the Study

179

TAKING IT TO THE BANK :

Poverty, Race, and Credit in Los Angeles
Executive Summary and Overview
A. EXECUTIVE SUMMARY
The City of Los Angeles commissioned this report on the role of financial institutions
in lower income and minority commmnities, seeking ways to encourage these institutions to

be more socially responsible. In particular, at the time the study was commissioned, Los
Angeles was considering linked deposit program , ander which the City would deposit
its funds with financial institutions providing better financial services to low income and
minority communities.

Financial institutions we often viewed as a repository for the wealth of the rich,
particularly in the wake of the " go- go" financial markets of the 1980s. But the behavior
of financial institutions also plays a crucial role in the health of all neighborhoods and
communities within their market veas .

Credit is less available and bank branches and services are scarcer in neighborhoods

with many low and moderate income residents than in upper income neighborhoods; and
credit is less available and bank branches are scarcer in neighborhoods with many minor
ity residents. The lives of working people, small business owners , and the poor in such

underserved neighborhoods are all profoundly affected. Low - income residents face onerous
charges for financial transactions and heightened personal insecurity because of the absence
of bank branches and of lifeline banking services. Moderate and middle- income residents

are denied a fair chance to own their own homes or their own businesses. Housing and
business development is stymied because of severe financing constraints.

La preparing this report, we first reviewed available information on the magnitude of
the housing crisis in Los Angeles, and examined previous studies of financial institution
1

180

lending in U.S. cities. We also looked into the legal responsibilities of financial institutions

toward low bacome and minority residents in the areas they serve. Federal law requires
that banks and savings and loun associations ( thrifts) assess and respond to the credit
Deeds of all the communities within their market areas . In addition , financial institutions
whose lending practices discriminate against African -American , Latino, Asian Pacific , or
other minority residents are breaking the law .

We looked at the relationship between financial institutions and the City's lower in
come and minority residents from several perspectives . First, we camined the level of
residential lending over the 1981-89 period across the City of Los Angeles. Given the legal

obligations of financial institutions to serve lower income and minority neighborhoods, we
examined in particular whether the level of residential lending varied according to the
income level or minority status of residents .' Ou findings at this stage were disturbing:
:

banks and savings and loon associations make fewer and smaller residential loans in lower

income and minority neighborhoods than in the City at large. This finding suggested that
the needs of residents in lower income and minority areas for home purchase and repair
were aot being met.
rever , it seemed possible, and even likely, that the pattern of lower lending in

minority neighborhoods could be explained simply by povesty — that is, lenders do not make
loans in minority neighborhoods because residents there are too poor to meet reasonable

*The term " hrift" is used asa synonym for" savings and loan association throughout
this report. The terms " bank " and " banking ' are sometimes used generically to refer to
depository financial institutions and their financial practices.
' The term " lower income" is used throughout this report to describe low and moderate

income weas of the City. Most often, it denotes the approximately 45%

tracts in the City which havelowmedian incomes.The term

of the census

minority communityor

sinority neighborhood" is used to denote areas in which a large proportion of residents

we Africao- American , Latino, and / or Asian -Pacific. These terms are given more precise
definitions us Decessary in various parts of the report.

• We use the terms " community," Deighborhood ," and " area . In fact, the geographic
uspects of the analysis in this report bave been conducted usingcensus tracts. Census tracts

we,

181

creditworthiness requirements. Sadly, apon cumination we found that this was not the
cuse . We discovered that banks and savings and loan associations make fewer and smaller
loons in African -American and Latino neighborhoods than in white neighborhoods in which
residents have comparable incomes.

We then turned to three other uspects of financial institutions' behavior in the City
of Los Angeles : their provision of banking services; their role in facilitating economic

development, particularly for small businesses; and their role in financing the development
of housing that is affordable for lower income residents.

Lovestigating banking services, we learned that financial institutions are reluctant
to maintain branches in lower income and minority communities. Further, some bank

practices are especially onerous for the poor about 14 % of Los Angeles banking institutions
do not allow non -depositors to cash government benefits checks, and fewer than 50 % offer

low -cost checking accounts to the non -elderly poor.
Our study of economic development found financial institutions reluctant to provide

some kinds of financing for small business, particularly in disadvantaged neighborhoods.
We also uncovered problems in the financing of rental housing that lower income people
could afford - few lenders were willing to provide predevelopment or permanent financing
for multifamily housing development. In sum , financial institutions do not provide adequate

banking services, economic development lending, or affordable housing financing for lower
income and minority communities.

After analyzing financial institution behavior, we developed some suggestions for rank
ing financial institutions on their success at meeting the credit and banking service needs

of lower income and minority neighborhoods. These suggestions encompasses two distinct
elements: first, lenders' performance in residential lending, relative to that of similiar in

stitutions; second, lenders' responsiveness in banking services, economic development, and
affordable housing development.

This study evaluates financial institutions' performance in residential lending, but
3

182

does not attempt to measure their performance in the other three areas . This partial

implementation is due to limitations in our data on Los Angeles lenders. With few ex
ceptions, all commercial banks, thrifts, and credit unions submit detailed data on their
residential lending to the federal government annually. Thus, all 220 institutions submit

ting residential lending data to the federal government were ranked on theis residential
lending to lower income and minority communities . By contrast, date on the second set
of topics are not systematically collected; for this study, we had only partial responses

on lenders' provision of banking services, economic development lending, or lending for
affordable housing construction .

This report concludes with a series of recommendations for the City government of
Los Angeles. The City is encouraged to take the lead in encouraging financial institutions
to meet the banking and credit needs of all City residents. Twelve recommendations are

made about the steps the City can take in program development, oversight, and advocacy.
B. MAJOR FINDINGS ON RESIDENTIAL LENDING
1. The Housing Crisis and Residential Lending Needs

The City's housing crisis is visible to anyone who walks the streets , reads the paper ,

or looks for a place to live. More than 25% of Los Angeles ' families spend more than half
their incomes for rent. For African - Americans and Latinos the situation is even grimmer:

more than 50 % of these households spend more than half their incomes for shelter. And
housing

For poor people, this statistic means real human suffering. A poor family whose adult
members work for the minimum wage may spend half its income for rent - and have to

live in a garage in a run -down neighborhood. After housing costs, little money is left for
food, transportion, or health care . This family is one injury or illness away from life on

183

the streets. How common is this scenario ? According to one Los Angeles Times study,
more than 40,000 Los Angeles families live in garages .

Mamy families and individuals are forced to live on the streets, in their cars or in

temporary shelters, because of the high cost of shelter. The number of homeless in Los
Angeles has been conservatively estimated at more than 30,000 individuals or 10,000 fam
ilies. African -Americans and Latinos are more likely than whites to be homeless. The
fastest growing category is families with children.

nia residents can afford to buy a home, compared to 45% nationwide. The situation in
Los Angeles is even worse. The average home price of $ 224,000 in Los Angeles — which
compares to a bational average price of $ 90,000 — implies impossibly high down-payment

and mortgage levels for even most middle -income households. By 1991, only 13.7% of all
households in Los Angeles could afford to purchase a home.

available to purchasers of homes whose mortage amount is $ 124,000 or less. This ceiling
is only about half the median cost of a house in Los Angeles; thus, few City homes qualify
for these programs. However, in South Central Los Angeles, the average home price is
$ 127,000. In this largely minority community, government backed loans do bold out hope,
if financial institutions would make a sizable volume of residential credit available.

The fact that most Angelenos, and particularly members of the City's minority popu
lation, caanot own a bome means more than the failure of the American dream . It means
that the traditional avenue by which lower income Americans accumulate wealth - through
the equity in their homes - is closed to many Los Angeles residents.

* This figure is taken from a comprehensive 1991 study of housing affordability by the
National Association of Homebuilders, based on 300,000 sales in 147 urban areas. Los
Angeles was the second least affordable city in the US, after San Francisco.
3

184

Public agencies, especially the City government of Los Angeles itself, bave invested

some resources in affordable housing, but public resources are very limited. Moreover, even
housing projects backed by public funds or tax credits usually require private sector insti
tutions to provide gap " financing. The need for housing among residents of lower income

and minority neighborhoods can be adequately met only if the institutions that tradition
ally lend for housing Deedscommercial banks and savings and loan associations will
extend credit to residents of these communities .

3. Residential Lending Patterns in Los Angeles
Financial institutions make more than 50,000 residential loans a year in the City of
Los Angeles. An analysis of computerized information on all those residential loans, for the

period 1981-89, forms the basis of our findings concerning residential lending. These loans

fall into four categories: loans for conventional mortgages on single-family homes; loans
for government-backed mortgages on single-family homes; loans for multifamily residences;
and loans for home improvements . A loan for purchase, repurchase, or repair is made on
about 5%

of the housing stock annually.

Almost all of this residential lending is done by savings and loan associations ( thrifts )

and mortgage companies. Commercial banks, however, make most loans for home improve
ments .

In the 1981-89 period, most residential loans by dollar volume- more than 82%

have been made for the purchase of a single family residence. About 6% have been made
for multifamily residences. Another 9 % have been made for home repair. Only 2 % of
all residential loans in the City have been in the government- backed loan programs

FHA, FmHA, and VA . The loans least frequently made in Los Angeles are those in the
government- backed programs aimed at low income and first -time home buyers. In 1989,
* Data
must be reported for these

185

fewer than 0.2 % of residential loans were government-backed.

Moreover, very few institutions make funds available for first- time bome buyers, con
centrating instead on homeowners who are trading op . Fewer than 5 % of the financial

institutions that make home loans and file Community Reinvestment Act ( CRA) state
ments report making loans to first-time home buyers .

multifamily loans.
S. Residential Loap Flows and Income Levels

Commercial banks, thrifts , and credit unions all make fewer residential loans as the

income of a community's residents declines. Between 1981 and 1989, almost 3 loans were

made in the City's highest income neighborhoods for every one ( 1) loan in a lower income
neighborhood. There are more residential buildings in upper income than in lower income

communities, but this ratio of eligible residential buildings is approximately two-toone.
Appronimately 2.50 loons are made per building in upper income communities for every
one ( 1) loon per building in low income neighborhoods. In dollar terms, almost $ 4 has been
loaned for residential credit in apper income areas for every $ 1 loaned in lower income

areas, on a per-building basis.
The provisions ofthe CRA require that reporting institutions list all the types of loan
that they provide. In 175 CRA statements made available to us for analysis in this study,
less than 3% indicated that credit is extended to first-time home buyers.
7

60-893 0 - 92 - 7
-

186
Single - Family Home Loans. The disparity in lending flows between upper and lower
income communities is even greater for single-family residential loans. In the 1981-89
period, financial institutions made more than 2.5 single -family residential loans per building
in upper income areas for every one much loon in lower income communities. Loan size

also increases uniformly with increases in median income.
Home Improvement Loans. More than twice as many home improvement loans are
made per building in upper income commmnities as in lower income communities. More

than $ 5 per building was loaned in upper income neighborhoods for each dollar loaned in
lower income neighborhoods for home improvement.

4. Residential Loan Flows and Racial Concentration
The lower the percentage of minority residents in a neighborhood , the more loans are
made for all residential categories. More than 2.25 loans are made per building in pre

dominately white communities for every loan to communities with significant minority
populations. Further, the lower the percentage of minority residents in a neighborhood,

the larger the loans made. Nearly 56 is loaned per building in predominately white Deigh
borhoods for every $ 1 per building in minority neighborhoods.
Single-Family Home Loans. The lending gap between white and minority communi
ties is even greater when single -family home loans are isolated. Lenders make nearly s

single -family home purchase loans per building in white neighborhoods for each such loan
per building in minority neighborhoods. And they lend almost $ 8 for single-family home
purchases in white areas for every dollar loaned in minority areus .

Home Improvement Loans. Home improvement loans follow the same pattern - more

loans and larger loans per building when lending for home improvement is compared in
predominantlywhite and predominately minority areas .
In sum , the lower a neighborhood's median income, the fewer and smaller the residen .

tial loans it receives. The higher the proportion of minority residents in a neighborhood, the

187

fewer and smaller the residential loans made. These differentials are not explained by the

number of residences, por are they explained by differences in reported crime levels - i.e,
by the perceived fiskiness of an area. The evidence does nuggest that mortgage bankers

play a major role in residential lending throughout the City, and especially in lower income
and minority areas. Data for morteuse bankers in Los Angeles suggests, further, that these
bankers do not attach more onerons terms on their loans than do banks or savings and loan

associations. But mortgage bankers' loan activity only partially offsets the lending pat
tern of banks and saving and loan associations. When all lenders are taken into account,

residential loans fall as income falls, and residential loans fall as minority population rises.
8. Racially Disparate Lending: Lower Credit Flows

Of course, it may seem possible, and even probable, that the disparity in loan flows
based on racial concentration could be due simply to the fact that areas with large propor

tions of minority residents are also areas with low median incomes. To test this possibility,
we compared the number and size of residential loans in largely white and largely minority
communities with comparable median incomes .

Financial institutions' failure to lend as much or as often in minority areas, when
it cannot be explained by income differentials, is termed racially disparate lending " in

this study." We found compelling evidence of widespread racially disparate lending in the

City of Los Angeles. Areas particularly affected by these racial disparities in lending
fows are highlighted on the map on the next page. The number of loans made, and
the amount loaned, is lower in minority neighborhoods than in white neighborhoods with

comparable median incomes. This pattern is particularly common in areas with many
African - American and Latino residents.

' This term is not synonymous with the term Sedlining." A stiffer standard of proof is
necessary to establish the existence of " redlining ; " as noted above, this termtypically refers

188

Whether lending institutions consciously avoid minority neighborhoods or have anin

tentionally discriminatory practices, the results are startling and unambiguous: For 1981
89, at least twice as many loans were made per building in low -minority areas than in mi
nority areas whose residents have comparable incomes. Nearly 85 was loaned per building
in low -minority communititus for every $ 1 loaned in minority communities with comparable
median incomes.

These findings are deeply troubling in light of the City's housing crisis. They imply

C. MAJOR FINDINGS ON BANKING SERVICES
As important as residential lending is to community health, residents in lower in
come and minority communities need other banking services as well. To gain insight into

lenders ' performance in the banking service area ( and also in their performance in lend
ing for economic development and for affordable housing ) , we collected information from
Beveral sources - lenders' CRA statements , a survey distributed to Los Angeles financial
institutions, a survey mailed to Los Angeles developers of affordable housing, and a survey

and several meetings with residents and business people in the City's enterprise zones .

Angeles
Los
of
City
Building
Residential
per
Loans

Loons
/1Bldgs
. 00

50.0
75.0
100.0

Miles
4
2
0

189

25.0

190

these simple seeds are not met in low income neighborhoods. Using 1989 data, we found
that bank and thrift branches are disproportionately located in upper income and low
minority areas of Los Angeles. Eseluding downtown and Mid -Wilshire branches, there
were only 1.3 branche per 10,000 residents of lower income areas, oerou 2.0 per 10,000

residents elsewhere in Los Angeles. The clorare of about 8% of branches in the past 3
years , much of it due to reorganisation within the thrift and banking industries, does not
bode well. Further opbeval is banking structure could signal even more branch closures .

always assessed on the elderly poor .

Low -income people who rely on government benefit checks are at a special disadvan
tage. Fewer than half of the City's financial institutions will allow general relief recipients
to " directly deposit ' their benefit checks into their accounts. Nearly 05 % of financial in .
stitutions in the City refuse to allow non -depositors to cash government benefit checks in
their branches. This, together with the paucity of bank branches in lower income neighbor
hoods, pushes government-benefits recipients into using check -cashing stores, whose rates
are as high as 10%

of the checks' face value. In addition , having to take the entire check

in cash makes these recipients particularly vulnerable to robbery and physical abuse.

D. MAJOR FINDINGS ON ECONOMIC DEVELOPMENT
Perhaps the outstanding finding concerning economic development lending is that
very little is known about lenders' activities in this area . Federal regulators do not collect
comprehensive annual information about loans to businesses and individuals, and few
Institutions have identified credit needs for economic development.

Some patterns did emerge in the data available to us, however. Some types of credit
11

191

are offered by only a few institutions , in particular, startup and expansion credit for
businesses, and mortgage credit for first time home buyers . The loans that are made
to businesses are typically almost 100 % collateralized . Further , lenders reported that

their most common reasons for denying credit to businesses are, in order of importance:
debt / income ratio; inadequate capitalization; insufficient collateral; and business' credit
history.

These tentative findings suggest two cautious conclusions. First, improved levels of
residential lending will have significant spillover benefits for entrepreneurial activity, since
this will increase the collateral in homes among owner -occupants. Second, it suggests that
innovative programs aimed at improved business planning, at business start-ups, and at
expanding capitalization could significantly enhance economic development in Deedy areas.

At present, only one in ten lenders have special credit programs for businesses in lower
income areas; one in five bave such programs foi minority -owned and /or women -owned
businesses. More such activities should be encouraged.
E. FINDINGS ON LENDING FOR

AFFORDABLE HOUSING DEVELOPMENT
Our inquiry into the relationship between housing construction and financing focused
on the experience of developers of affordable multifamily rental housing in Los Angeles.
Both financial institutions and experienced low income housing developers agree that de
velopers of affordable housing face formidable obstacles to financing their projects.
Developers overwhelmingly report that it is harder to get financing for affordable
housing projects that are to be located in minority or lower income communities, than when

they are planned for upper income or low -minority neighborhoods. However, the reluctance
of many aftuent communities to accept these projects is well known . Development of
these projects thus becomes doubly difficult, squeezed between lender and community
preferences .
12

192

Construction projects for affordable housing move through three stages of financing:

predevelopment and site acquisition, construction , and permanent financing. Public funds
play a key role in the development of affordable housing, especially housing built for very
low income tenants. But private sources play an important role in " zap " financing - that
is, filling the gap between total development cost and the cheap funding provided through
public prants and low - interest loans.

Very few banks or savings and loans will make loans for predevelopment. Almost
all affordable housing developers at this project stage have to rely on internal funds or

on monies from specialized sources, such as the California Equity Fund. This means
that new developers of affordable housing must start with substantial outside sources of
funds. Further, land acquisition, usually the most expensive cost item in affordable housing

projects, is the most difficult area for which to obtain credit from conventional lenders.
Financial institutions are generally more prepared to make construction loans. How
ever , very few Los Angeles banks and thrifts will provide permanent financing for these
projects - that is, the vast majority will not provide long -term mortgages to pay off de
velopment and construction costs . Not surprisingly, given these credit -market hurdles to

affordable housing construction, we found little evidence of recent, current, or planned
lender involvement in affordable housing projects. Only a handful of affordable units are

being financed by the more than sixty institutions that responded to our bank survey.

F. RATING FINANCIAL INSTITUTIONS' PERFORMANCE
1. A Ranking for Residential Lending
As requested by the City, we developed a system for ranking all financial institutions on

residential lending in lower income and minority communities. Institutions were grouped
by the size of their residential lending portfolios — that is, larger residential lenders were
compared to one another, and smaller residential lenders were compared to institutions of
13

193

comparable sise. Specifically, financial institutions were broken down into four groups.

areas:

• Lending in " Block Grant flow and moderate income) arcos. Lending in areas
officially designed as low and moderate income under the federal Community De

velopment Block Grant program . These neighborhoods comprise approcimately
45% of the City of Los Angeles.
• Lending in minority areas: Lending in census tracts with 71% or more minority

population. These areas constitute 40 % of the City.
• Lending in very low income areas. Lending in that 20 % of Los Angeles census
tracts with the lowest median income levels .
• Lending in arees with many African -American residents. Lending in census
tracts whose population is more than 30 % African -American . Theseareas con
stitute 20 % of the City.

• Lending in areas with manyLatino residents. Lending in census tracts whose
population is more than 51% Latino. These areas constitute 20% of the City.
. Lending in areas with many Asian Pacifie residents. Lending in census tracts

whosepopulation is more than 15% Asian-Pacific. These areas constitute 20% of
-the City.

The 220 financial institutions reporting residential lending information under the

Home Mortgage Disclosure Act in 1989 were ranked on the basis of these criteria. The
rankings of all institutions under this system are set out in the appendices to Chapter 5

of our full report.
2. Transaction Services, Economic Development Lending ,

Residential lending is the only banking activity for which federal regulators collect

detailed information annually from almost all financial institutions. Therefore, this is the
only area in which a ranking encompassing most Los Angeles lenders can be constructed.

194

City of Boston programs -both use such " comprehensive ranking systems to evaluate par
ticipating financial institutions.

G. RECOMMENDATIONS
Ow study concludes with a set of specific policy recommendations. First, and most
importantly, we recommend very broadly that the City of Los Angeles assume a leadership
role in encouraging reinvestment and socially responsible behavior by Los Angeles' financial
institutions.

This leadership role involves three elements, each embodied in several recommen
dations. First, the City can demonstrate this leadership by adopting and administering

a fair and effective linked banking program ( RECOMMENDATIONS 1-4 ) . Second, the
City should promote several innovative ideas for delivering credit and banking services

to underserved areas and residents of Los Angeles ( RECOMMENDATIONS 5-8) . The
City should encourage financial institutions in Los Angeles to experiment with these or
15

195

other innovative ways of achieving reinvestment goals. Third, the City should encourage
changes in the broader environment of credit and banking practices in Los Angeles ( REC
OMMENDATIONS 2-12) . Specifically, financial institutions should be urged to improve

their efforts to assess credit Doeds, and to increase the amount of information they make
publicly available. The City should seek statutory changes at the state level, and it should
encourage more vigorous federal regulation of lendes activities.

1. A Linked Banking Program
• RECOMMENDATION 1: The City of Los Angeles should imple

The City originally commissioned this study to determine whether a linked deposit

program is necessary and feasible, and whether it can be effective. The findings summa
rized above demonstrate that it is necessary. Its feasibility and effectiveness depend on

how it is implemented. The volume of deposits and City financial business awarded under
this program should be linked to institutions' scores on a four- fold ranking scale encom
passing residential lending, banking services, economic development lending, and lending

for affordable housing development. Points should be awarded in this ranking scale on
the basis of " socially responsible " reinvestment performance, using measures that are spe
cific, clearly defined , and standardized. Participating institutions should compete
" level playing field " —that is, lenders should be compared to others with comparable types
and volumes of business. The rankings of participating institutions should be publicized.

not strictly necessary, especially if the City's program emphasizes longer-term deposits.
16

196

Implementation of a linked banking program , instead of a Barrower linked deposit
program , would increase the number of participating institutions. Our bank survey found
that many institutions are not interested in serving as depositories for the City of Los
Angeles. However , a large number of lenders can be expected to particate in a program
that gives them access to a full range of the City's financial business such as payroll and
account processing, depository business, fees on special accounts, and fees on bond issues.
The data for ranking financial institutions are available from three sources : the Federal

government annually supplies a computer tape containing residential lending data compiled
under HMDA ; institutions' CRA statements , which must be reviewed annually; and a
special questionnaire for selected data items.

The administrative and technical demands of a linked banking system alone justify
the creation of a dedicated City office. We suggest, more broadly, that this office serve
:

also as an advocate for reinvestment by financial institutions in the City.
2. Innovative Ideas for Reinvestment
• RECOMMENDATION B : The City should work with lenders to

This second set of recommendations goes beyond the implementation of a linked bank

ing program . They are included both because they emerged in our discussions with City
17

197

officials, lenders, and residents , and because they bave promise us tools for reinvestment.

business .

Recommendation 7 sets out in one place several specific policies which , if universally
enacted , would greatly enhance the banking services available to very low income residents
of Los Angeles. Recommendation 8 responds to the concern on lenders' part that small
businesses, particularly in lower income areas, sometimes are poor credit risks because of
improper planning and administrative procedures . Rather than using this problem as an
excuse for not lending , we suggest that lenders overcome it by attacking it directly.

3. Changes in the broader environment
• RECOMMENDATION 9: The City should encourage financial insti

198

No matter how muccessful, a linked banking system and City initiatives can only

partially meet its reinvestment goals. The City will be more likely to achieve more equitable
credit and banking services in Los Angeles it changes occur in the broader environment

of banking practices and regulation. These last four recommendations call for the City of
Los Angeles to demonstrate its leadership by acting us w advocate for changes in lender
practices, federal enforcement, and state legislation. The City's efforts would be enhanced
If financial institutions could be induced to make deeper efforts to assess credit Reeds,
and to reveal detailed geographic lour information on launs in other areas than residential
lending. This would create a more complete picture of the gap between community need

and institutional response everywhere in the City. The City can seek these changes in its
relationship with financial institutions directly. However, it can also make efforts to affect
the regulatory and statutory environment within which Los Angeles' financial institutions

operate, since this sets the tone for lenders' behavior. Unfortunately, until very recently
regulatory officials have largely neglected enforcement of federal laws governing fair lending
and the disclosure of credit practices. Moreover, because credit is such an important issue
for revitalizing the local economy and local neighborhoods, it is appropriate for the City to
seek state legislation which will further enhance the status and importance of reinvestment
activities . Such legislation could give City officials and Los Angeles residents a greater stake
in determining which loan products and programs and which banking services best respond
to unmet local needs.

199

Chapter 1

Contexts of this study : Historical,

A. INTRODUCTION

This chapter places our analysis of financial institutions' behavior into its several
contexts - historical, legal, and socio- economic — thus setting the framework for the rest of

the chapters. First, section B provides a capsule history of banking and mortgage credit
practices as they pertain to lower income neighborhoods and to people of color.' The
current legal context of financial institution behavior is presented in section C. Section D
goes on to review contemporary social criticisms of banks' mortgage lending practices in
lower income areas. This section includes references to some of the important studies of

bank behavior in urban areas . Both " pro " and " con" studies are reviewed, including the
well-known " Color of Money " studies of Atlanta , Boston, and Detroit. Section E then sets
out the socio - economic context of this study, emphasizing the situation of the City of Los
Angeles.
B. MONTGAGE LENDING IN URBAN COMMUNITIES :

This study focuses on how well financial institutions are meeting financial needs in
lower income and high minority communities. The notion of the social responsibility of
financial institutions, in itself, has emerged only in the relatively recent past."

* As in the Executive Summary, the term " lower income” means both low income" and

“ moderate income," bere and in the remainder of the report. More precise definitions of
income are setforth and used when appropriate, particularly in Chapters 2-5.
? Lo the interest of brevity, this brief history is restricted to a discussion of the evolution of

residential lending markets. Many of the same events discussed herein helped to shape the
role of financial institutions in the economic development initiatives undertaken in lower
income and high minority communities.
20

200

largely bypassed the working class and the poor. The idea of long-term mortgages was not
well-developed. The mortgage market for real estate was largely a venue for underwriters

promising speculative returns to wealthy investors and trust fund managers . Further,
there was a housing affordability gap in the 1920's between working -class wages and the
supply price of housing. A combination of banking reform legislation in the 1930's, rising
wages for the working classes, and the emergence of a middle class put in place a Dew
system of housing finance.

Alongside this emerging ideal of a homeowner class, however, there also grew up an
ethic of bias against lower income households and against people of color. Even the Federal

Housing Administration, an agency whose substantial mortgage and insurance programs
encouraged home ownership , " until the mid -1960s ... consistently excluded minority and
lower-income familes and certain inner -city areas from FHA eligibility." Indeed , " from its

inception FHA set itself up as the protector of the all-white neighborhood." Other devel
opments in the housing market followed this lead. From the 1940's onward in Los Angeles,

suburban incorporations and homeowner associations were founded on the principles of
racial and class exclusion ."
Racial bias was even explicitly incorporated into the professional standards that guided
* Mortgage-based lending was actually illegal for national banks until 1913. Most mort
gages were long - term bonds floated by specialized mortgage companies and real-estate
underwriting firms. These were often affiliated with commercial banks under corporate

umbrellas. There was frequently self -dealing in these financing utangements. In the
Great Depression , somewhere between one-quarter and one-third of these bonds went into
default. See Chapter 3 of Alan Rabinowitz, The Real Estate Gamble: Lessons from 50
Years of Boom and Bust, Amacom , 1980.

* See Gail Radford , " New Building and Investment Patterns in 1920s Chicago ," mimeo ,
University of Cincinnati, December 1990 .

'Rachel G. Bratt, Rebuilding a low -Income Housing Policy, Temple University Press,
Philadelphia , 1989, Chapter 6.
Charles Abrams, Forbidden Neighbors, Harper and Brothers, New York , 1955. Calvin

Bradford writes that " Untilthe 1960s, and the aftermath of the riots , FHA was so re
strictivein its own lending that the term 'redlining' was coined to describe theinner-city
and minority areas defined on FHA maps as places where it would not lend" ( Never Call
Retreat: The Fight Against Lending Discrimination, Chicago, 1990: 1) .
' Mike Davis, City of Quartz: Ezcavating the Future in Los Angeles, Verso Press, New
York , 1990 , Chapter 3.
21

201

real estate practices. For decades, onderwriting experts contended that race was an appro
priate element in appraisals, because there were races whose presence reduced the value of
property. For example, Homer Hoyt, one of the early authorities of the appraisal industry,
ranked residents of English and German extraction as the most desirable neighbors, and

“ Negroes " and " Mexicans ' as least desirable . Such principles were incorporated into the
instructional materials of the National Association of Real Estate Brokers and, to the end
of the 1960's, of the American Institute of Real Estate Appraisers ."

for their services,

Social and political developments in the 1960's and 1970's brought substantial changes.

In particular, the urban riots of the 1960's and the Black and Chicano liberation movements

defined a new set of constraints and possibilities for social policy. The federal government
responded by passing new housing programs, implementing various urban program ini
tiatives, and making important changes in federal laws and administrative practices. For
example, the FHA reversed its practice of not making loans in minority areas. Numerous

Dew laws were passed; these are summarized in the next section . The social ferment of
this era gave rise to a community empowerment movement for which inequitable outcomes
Hoyt, One Hundred Years of Lond Values in Chicago. Chicago: University of Chicago
Press , 1933, 136. The FHA hired Hoyt to write its underwriting standards in 1932. An
other potable , Frederick Babcock , wrote that there is one difference in people, Damely

race, which can result invery rapid decline. Usually such declines can be partially avoided
by segregation and this device has always been incommon usage in the South where white
and negro populations have been separated " ( The Valuation of Real Estate, New York:
McGraw - Hill, 1931 , 91.
Calvin Bradford , Credit by Color: Mortgage Market Discrimination in Chicagoland,

Chicago Area Fair Housing Alliance, Chicago , 1990, 1 .
22

202

of bank lending policies are central concerns . Of particular concern is the occurrence of
Sedlining" -lower lending flows to minority individuals or high -minority areas which can
Dot be attributed to any characteristics that might affect creditworthiness. Throughout the

past two decades , the issues of equitable credit foms and banking services have remained
a touchstone of uban activism .

The Reagan era brought further changes. This administration's pro -market, Supply
side " orientation had important ramifications for mortgage and housing markets. Most

public housing subsidy and construction programs were either terminated or vastly re
duced in scope. The most important remaining housing program , the multi-faceted and

underfunded " Section 8" program , operates on the principle that the private sector should
supply housing; the government merely provides vouchers for those too poor to afford

market rents. Unfortunately, the lenders that had traditionally provided the credit for
housing construction and mortgages were poorly positioned to rise to this challenge. As
Reagan won election , the banking and thrift industries themselves were already entering a

decade of upheaval.
Tight monetary policy and deep recession in the early 1980's led to stratospheric
interest rates , credit- risk problems, and many bank and thrift insolvencies. Congress
passed financial

door to banking- industry consolidation, is on the table.

The decade- long crisis among banks and thrifts ( savings and loan associations) has had
an impact on banking and credit practices in lower income communities. Most importantly,
lending capacity has been lost in the residential credit markets for two reasons . First,

banks and thrifts bave been closed, merged, or otherwise reorganized. In principle, the
credit and services of closed institutions could be replaced by their purchasers; but in
fact, branch consolidations and moratoria on new lending have often accompanied these
23

203

reorganizations. Second, the privileged position of mortgage and residential lending in
U.S. credit markets bas been reduced . This reduction is largely due to the deregulation of
financial intermediaries, which bus caused increased bank dependence on financial markets

us sources of funds and us secondary -market customers.10

of an active, and sometimes hyperactive, market for the securitization , sale, and resale of
mortgage-backed securities . Banks and thrifts that previously held mortgages to maturity

Dow sell them off to Secondary lenders " financed by pools of mutual and retirement

funds. While this replenishes the lending capacity of primary lenders ( assuming they
remain solvent) , it also makes them increasingly dependent on financial market conditions
and on portfolio managers ' decisions."

C. THE LEGAL CONTEXT OF BANK LENDING PRACTICES

The last 30 years have also seen substantial changes in the statutory and regulatory
environment of residential lending. These changes have both been spurred by, and reflected,
the social environment and national agenda at the time of their passage. Two acts in

the areas of housing and lending are rooted in the civil rights movement associated with
sº See Anthony Downs, The Revolution in Real Estate Finance , The Brookings Institution ,

Washington, DC, 1985. Chapter 3 discusses the impact on Los Angeles residential lending
of lost lending capacity due to reorganizations and closures.

" One can conjecture that the increasing importance of thesecondary market in mortgage
finance will also heighten the importanceof the underwriting standardsused in professional

appraisals. Thus, someof the concerns about the criteria used to evaluate properties in
lower and /or in high minority areas may take on renewed importance.
24

204
Dr. Martin Luther King. The Fair Housing Act, first enacted in 1968 , made it illegal
to discriminate against any person to the terms, conditions or privileges of a residential

sale on the basis of the race of either the applicant, the residents of the building, or the
population in the rurrounding aru . The Equal Credit Opportunity Act, enacted in 1975,
made it illegal to practice racial discrimination in all credit transactions.

• Assess the credit needs of the entire community in which they are char

Under the CRA, lenders we periodically reviewed or " examined ” by their regula
tors. Further, a financial institution that wants to expand its operations ( for example, by

opening a new branch office) , must obtain the approval of its supervisory agency. This
approval depends, in part, on whether the institution in question has been meeting its
1912 USC Sec . 2902.
25

205

CRA obligations.

Several agencies were designated as responsible for enforcement nader these acts : the
Federal Reserve oversees state member banks, the Office of the Comptroller of the Currency
oversees national banks, the Federal Depository Insurance Corporation is responsible for

DOD -member and state mutual saving banks, and the Office of Thrift Supervision oversees
thrifts . Specifically, these agencies are charged with assessing the performance of lenders

under CRA . These agencies base their evaluation of institutional performance on analyses
of the extent to which each financial institution :
•

Ascertains community credit needs;

• Makes available and markets various loans for home mortgages and home im

• Engages in community economic development.

There are gaps between this broad language and the actual mechanisms for oversight
and evaluation that have been put into place, however. Regulators have not exercised their
option to require banks to seriously assess community credit Deeds. Regulators have also

failed to request information from lenders on the number and dollar volume of loans in
other lending areas besides residential finance. The paucity of information on economic
development lending reflects federal regulators' unwillingness to require lenders to publicly
release comprehensive data on business loans.

Another gap appears when the information that is collected is used to evaluate insti
tutional performance. Theoretically, regulators have the power to rate lenders as 'unsatis
factory " under these criteria. However , this negative rating has been little used. Between

1986 and 1989 , for example, regulators ranked as " unsatisfactory " only 3% of the lenders
26

206

they examined .'' Regulators have been somewhat more agressive in bandling lenders'

applications for expansion. Here, denials or delays in approving applications have resulted
in a number of agreements between lenders and community groups ."
The CRA's provisions that lenders seeking new branches must have satisfactory CRA
performance have similiarly made bank branching decisions a focal point for organizing
efforts.

The direction of federal oversight policy is currently in flux. On the one hand, the
thrift bailout bill of 1989 contained provisions which toughen reporting standards under

HIMDA : larger mortgage bankers will be required to report their residential lending data,
and data will now encompass some characteristics of applicants. Ratings will also now be
made public. However, amendments to the CRA Dow under consideration in the House
of Representatives would limit the coverage of the CRA to institutions with assets of $ 100
million or more; these amendments would also insulate institutions rated 'satisfactory "
or better from challenges, if these institutions seek to expand their operations within two
years of receiving this rating.

" Statement by the Leadership Conference on Civil Rights, Oversight Hearing on Govern
ment Check Cashing, Subcommittee on Consumer and Regulatory Affairs of the Housing

and Urban Affairs Committee, 101st Congress, first session, page 7 ( June6, 1989) ) .
" See, for example, • ' Blackmail' Making Banks Better Neighbors , Business Week, Au
gust 15, 1988 .

1. These changes may stiffen regulators ' standards in evaluating banks. According to a

preliminary studysponsored bythe Center for Community Change,89 % of lenders
27

207

legal and regulatory context for states and municipalities. This leads to the first of the
twelve recommendations of this study."

RECOMMENDATION : The City should intercede with
federal agencies to improve the enforcement of federallaws
on financial institutions' performance and reporting.
Various sections of this report will discuss specific examples of how improved reporting
and enforcement of federal laws could benefit the City's reinvestment goals.

vestment standards in their interstate banking legislation. For example, Minnesota requires

that out-of-state bank holding companies seeking to enter the state provide FIMDA-type
information on both residential and non - residential loans, deposit information by zip code,
CRA statements, and information on development loans” by category. " Development
loans" are loans in areas accorded a high social priority by the Minnesota legislature, and

specifically for lower income housing, female and minority -owned businesses , student ed
ucation, alternative energy and energy conservation , and for distressed areas and Indian
reservations.

208

plans for new capital investments, loans for businesses, and consumer and business services .
Out-of-state banks that acquire Maine banks must report annually on their performance
ander this criterion . At least four other states have followed Maine's lead , us has the
District of Columbia. The District's statute requires the creation of between 80 and 200
jobs for District residents for banks opening up offices there.

RECOMMENDATION : In conjunction with other cities
in California , the City should recommend and support

California state legislation to encourage improved lending
and banking services in lower income and minority areas .

D. CONTEMPORARY STUDIES OF BANK LENDING

Paralleling the surge of urban activism over banking policy in the 1970's was an
outpouring of research on racially disparate lending and discrimination. Initially, many
community organizations used the newly available HMDA data to construct direct compar
isons of loan iows in one area versus another. '* These studies became more sophisticated
over time: authors were more likely to be academically trained, and the studies themselves
drew more heavily on the literatures on mortgage and housing markets.

" The early history of HIMDA-based studies of redlining is detailed by Calvin Bradford
and the Urban -Suburban Investment Study Group in " Redlining and Disinvestment as a

Discriminatory Practice in Residential Mortgage Loans," Center for Urban Studies, Uni
versity of Illinois. Department of Housing and Urban Development,Office ofthe Assistant
Secretary for Fair Housing and Equal Opportunity, 1977.
29

209

lending is Shafer and Ladd's 1981 book, Discrimination in Mortgage Lending." These
authors centered their study on the occurrence of redlining, which they defined as " he
refusal to lend, or the granting of mortgages with less favorable terms, in certain areas

even though the expected yield and risk of loss are the same as they are for mortgages
panted in other areas." On the basis of applications data from California and New York
State, higher delinquency rates in 'redlined " areas were attributed entirely to borrower
characteristics. That is, redlining practices lacked an economic rationale.20 In a recent
study of racial discrimination, Gabriel and Rosenthal found, using data from the 1983
Survey of Consumer Finance, that " minority households are significantly less likely to
obtain conventional financing than whites, even after controlling for various proxies of

default risk.º31

210

with all other banks' refusing to lend there make it economically irrational for any given
bank to lend there. Spillover occur be that the cumulative decisions of all lenders about
credit transactions have an independent effect on economic values- Degative effect, in

this case which is not captured in the individual prices struck in credit-market trans
uctions in this wu . Unfortunately, if not conveniently, these two types of redlining us

characterized by these anthon - are indistinguishable in practice.

Several conclusions emerge from these and other pro ' and ' con ' studies. First, both
advocates and critics agree that discrimination against individuals and racially disparate

lending by geographic areas are different phenomena. Second, there is broad agreement
that discrimination against individuals has no economic rationale, but there is disagreement
about whether racially disparate lending does.

as A. Thomas King, Discrimination in Mortgage Lending: A Study of Three Cities, NYU
Graduate School ofBusiness Administration, Monograph Series in Finance and Economics,
Monograph 1980-4 .
" Appeals to this type of free market argument — the assertion that if u activity is prof

itable, someone will doit - are often made. For example , Robert Eisenbeis, Wachovia
Professor at the University of North Carolina, argued before theHouse Banking Commit
tee in Fall1990that if the thrift industry wereto disappear, commercial banksspecializing
in real-estate lendingwould rush into the mortgage- financing breech. Eisenbeissupported
his argument with 1970s data to the effect that a sample of commercial bankswith portfo
lios weighted heavily towardreal -estate lending wereno less profitable than other banks.
How compelling such arguments and such evidence are is notclear.
31

211

Third, it is difficult, ifnot impossible, to demonstrate to the satisfaction of all observers
that any given geographic pattern of differential agregate lending flows lacks any legiti

mate economic rationale. For example, Schafer and Ladd ( 1981) note that no community
based studies have identified redlining us they define it ( see above) , because they are unable
to control statistically for the nonredlining measures banks may legitimately use to make
DO or few loans in particular areas . These measures include lack of adequate demand,

few credit-worthy applicants, external risks, and decisions by autonomous entities ( like
insurance companies) .
Fourth , all residences and businesses in areas that experiences lower lending flows
because of their minority residents are subject to negative spillover effects: when fewer

or smaller loans are made in a given area, property turnover necessarily slows; absentee
property tenure and property abandonment are more likely; businesses may be less likely to
obtain loans, in turn , due to deteriorating neighborhood conditions. Urban neighborhoods
receiving systematically lower lending flows, in short, are more vulnerable to economic
decline, ceteris paribus, whatever other economic development policies or programs are

in force. When areas are denied what Shlay calls their " air share" of lending flows,
Seighborhood effect redlining ” ( Guttentag and Wachter) and overall economic decline
both ensue.36
Clearly, differential loan flows in different areas of cities, when these are observed,
are inextricably linked to a broader set of economic dynamics. Galster and Keeney have

shown how prejudice, housing and labor-market discrimination, segregation , and economic
disparities all interact and reinforce one another in the urban community.” Identifying
** Anne B. Shlay , " Maintaining the Divided City: Residential Lending Patterns in the
Baltimore SMSA , prepared forthe MarylandAlliance for Responsible

212

these links in a well-specified conceptual and empirical model is a task that has as yet
eluded researchers. Nonetheless, there is a prowing consensus that the very existence of
discriminatory or racially disparate lending patterns represents an urgent challenge for

public policy -makers.

The links between race, income, and aggregate lending flom received renewed at
tention with the publication of highly -publicised studies of Atlanta and Detroit in 1988 ,
followed by two studies of Boston in 1989. The Atlanta Constitution published its Pulitzer

Prize winning series, The Color of Money," on May 14, 1988. This study divided Atlanta
census tracts into categories based on median income and on the percentage of minority
residents."

It found a ratio of 5 loans in white neighborhoods for every I in a black

Deighborhood with a comparable median income level. Using application data voluntarily

submitted by two local thrifts, it found blacks much more likely to be rejected for mortgage
credit than whites, all else being equal.
In August 1988, the Detroit Free Press published its own series, " The Race for Money ."

This study followed a similiar research approach . In particular, it compared white and
black neighborhoods with similiar median incomes and with housing stocks of approxi
mately equal average age and median assessed value. This comparison revealed twice as
many loans per 1,000 homes in the white neighborhood as in the black, with a dispro
portionately small number in the latter being made by banks and thrifts. It also found a
worsening gap over time.

These studies were followed by two major reports about lending flows in Boston , both

* Specifically, tracts wereclassified according to the following percentages of overallAt
lanta median income: 0-70 % ; 70-86 % ; 87 % -102 % ; 103-122% ; and 122 % andabove . The

racial categorization was 0-20 % minority ( in Atlanta , this largelyequatedwith African
American ) ,20-80 % minority, uod 80-100 % minority. The analytical methods of this study
were developedand discussed at greater lengthin Stan Fitterman'sMaster's thesis for the

city planning department at theGeorgia Institute of Technology ( " Mortgage Redlining
in Metropolitan Atlanta," June 1988) . Fitterman's thesis also delves into the supply of
mortgages by mortgage companies, using title transfer data . Be finds that while mort
gagecompanies arevery activein making mortgages in low -incomewhite tracts,they are

relatively inactive in tracts with 80 % or more black residents; again, the activity ratio is
‫ام ة‬.
‫ا‬

213

released in 1989. In August 1989, the Federal Reserve Bank of Boston released a study

entitled, " Geographic Patterns of Mortgage Lending in Boston, 1982-87." * It found that
the percentage of black residents significantly affected loan flows even when a large number
of other variables are taken into account. Further, lending by mortgage companies was

not sufficient to overcome the pattern of racial bias in HIMDA -reported loans.

made our approach feasible.

* The authors were Katherine L. Bradbury , Karl E. Caseand Constance R. Dunhara.
Some 48,000 real estate transactions over the time period ofthe study were divided first

into Boston's 60 " neighborhood statisticalareas " ( NSA's) These NSA's were divided, in

tum , into four equal groupson the basis of median incomeandintofive racial categories.
The authors also separated mortgage companies' loans from those made by banks and
thrifts.

3. This study also used Boston's 60 NSAs as its point of reference for examining both
HMDA and title transfer data . Finn used just two categories for racial concentration:

neighborhoods with 70% or more minorities,and thosewith 70% or less. The NSA's were
grouped into three equal parts on the basis of median income.
34

214

E. SOCIO - ECONOMIC CONTEXT:
HOUSING AND ECONOMIC CRISIS
Several trends have converged in the 1980's to create a profound housing and economic
crisis in urban America. The crisis has descended with particular ferocity on the City of

Los Angeles. This section briefly sketches out the socio-economic context by providing
some background for the honsing and economic crisis both in the nation us a whole and in
Los Angeles itself.
1. Housing and Economic Crisis in America
Mamy American households are caught in a squeeze between their housing Deeds, their
incomes , and their housing costs. On the one hand, declining real wages have reduced the

average American household's purchasing power . The U.S. income distribution has been
" hollowing out :" an ever larger proportion of all households are at both the lower and the
opper ends of the income spectrum ." On the other hand, housing prices have been rising
more quickly than overall prices." As a result, households have been squeezed between
a declining ability to pay for housing and a rising cost of housing. In Los Angeles, for
example, the median real price of homes more than doubled between 1974 and 1987, while

the average real income of all homeowners grew just 20% in the same period.

real incomes of all renters stayed approximately constant."

According to a 1990 report

by the Economic Policy Institute, over a third of the nation's households are not able to
s0For an extensive discussion of U.S. trends in wages and income distribution , see Barty

Bluestone and Bennett Harrison , The Great U - Turn, Basic Books, New York , 1987.
in the same period. Both figures are nominal ( not inflation-adjusted ) ,and bothare drawn
from the California Department of Housing and Commercial Development.
» From Tables A - 3 and A-10 of The State of the Nation's Housing, 1990, Joint Center for
" In California , rents increased 348% between 1970 and 1989 , while income increased 247%

Housing Studies, Harvard University.
35

215
afford decent housing within budgetary limits established by the federal government." By
1985, three - quarters of poor households paid more than half their incomes in rent."

an extreme, the great press of demand for available housing causes some to slip over the
edge into homelessness .

Rental Affordability Crisis .

Since lower income households are typically renters , the crisis of housing affordability

bas taken the form of a crisis of rental costs. According to the Joint Center for Housing
Studies at Harvard , the proportion of renters living in 'adequate but affordable housing "
has declined from 41 % in 1974 to 35% in 1987; but the proportion of renters in " adequate
but not affordable housing " bas risen from 23% in 1974 to 44% in 1987. “ Affordable
housing" in this context means that renters are spending no more than 35% of their
disposable income on monthly rent."

There is also a racial dimension to the rental affordability crisis. Nationwide, a larger
absolute number of poor white households live in inadequate rental units than do minority

households. But while African American and Latino households represent 17% of all
American households in 1985, these two minorities accounted for 42%

of all households

living in substandard housing."
The lack of affordable rental housing has far reaching effects. One consequence has
been a rapid increase in the size of the homeless population , whose fastest growing com
ponent is families with children . Another consequence is that families are increasingly

renting smaller quarters und " doubling op ." Overcrowding is the inevitable result. In
" Michael Stone, One Third of the Nation, Economic Policy Institute, Washington, DC.
The long-standing federal criterion is that housing expenses should consume no more than
30-35 % of household disposable income.
" Ong , Widening Divide, Chapter 8.

" " Adequate ” .housing is defined on the basis of the presence or absence of plumbing,
heating, and mechanical subsystems, along with repair and upkeep.
" These statistics are drawn from P. Leopard , C. Dolbeare, and E. Lazere, A Place to
Call Home: The Crisis in Housing for the Poor, Center on Budget and Policy Priorities,
Washington DC, April 1989.
36

216

1980, a third of poor households lived in overcrowded rental units; when the 1990 Census
data are available, this percentage will be far higher.

Ownership Affordability Crisis.
The single family home is the largest single component of a typical household's wealth.

It is the traditional means by which lower income families have accusulated wealth over
their lifetimes. Indeed, the dramatic increase in home prices throughout the 1980's allowed
owners to reap large returns from their housing investment.

But rising housing prices also have a disturbing aspect: they imply declining opportu
nities to own homes among all but higher income households. This trend will, over time,
have profound implications for the pattern of wealth accumulation by Income class, and
hence for the extent of entrepreneurial activity in American society."

The increasing difficulty of owning homes also has a racial component. African Amer
icans and Latinos are less likely to own homes than whites . For example , in 1986 the

homeownership rate for African American households was 45% , versus 70 % for white
households. The implication is that minority households' wealth has fallen behind that of
whites because the former have not shared equally in the real estate appreciation of the last
decade. Nationwide in 1987, 17% of white households had the income and wealth necessary
to buy the typical starter home with a 20 % downpayment. Only 1.5% of African Amer

ican households could qualify for the same home. Even if the downpayment were 10% ,
only 3.5%

of African American households could qualify. Within the African American

population, the differences in wealth between homeowners ( 538,321 in 1986) and renters
( 5735 in 1986) is staggering.
2. Housing and Economic Crisis in Los Angeles

s' For example,

217

These trends for the nation have been magnified in Los Angeles and Southern Cali
fornia . The distribution of earnings was more unequal in Los Angeles than nationally in
1987: proportionately, Los Angeles has both more low -wage and more high -wage workers
than the rest of the nation. Housing prices and rents have been climbing faster in this

region than in the U.S. as a whole. Hence the wedge between the wealth of homeowners
and that of renters has grown even more pronounced in Los Angeles tban elsewhere.
There are two notable aspects of the disproportionate growth of low income house

holds. On the one hand, full-time year -round workers are more likely to be poor because
of declining real wages . In Los Angeles, full -time year-round workers below the poverty
line totalled 38,700 in 1979, 11% of the workforce. By 1987, this figure had risen to 55,000 ,
14% of the workforce." On the other hand, more of the poor are not in the formal labor
market.

Again, race plays an intrinsic role in this dynamic. In 1969, most members of the
working - age poverty population in Los Angeles were white; by 1987, this population was

three- quarters minority. Both low - income Latino and African American communities, in
particular, are experiencing profound social crises. These crises are somewhat different
in kind because of the divergent historical roots of these two communities in Southern

California. Together, a segmented labor market and steady migration have combined to
create a workforce which is steadily employed at insecure jobs with low wages and do
benefits. Latinos constitute a disproportionate share ( almost two -thirds) of Los Angeles '
pool of year-round full-time workers in poverty. The crisis of employment in the African
American community is especially acute. Members of this community constitute less than
one- twentieth of the year-round full-time workers in poverty, but nearly one- fifth of the
poor outside the formal labor force.

" These figures and those immediately below are drawn from Ong, The Widening Divide,
especially pages 136-7.
38

60-893 O - 92 - 8
-

218

Affordable Housing in Los Angeles .

reaction to continued growth has cut housing production; and ( 5 ) Federal housing program
monies that have been cut by 75 % since 1980.

" This is an extremely conservative estimate of the homeless population . The State De
partment of Social Services and Los Angeles County PublicServices estimate 160,000
homeless people in the County. This gency also estimates that approximately 30 % of
homeless adultunits, single or paired , have children .
* The ten least affordable cities in this study of 200,000 households were all located in Cal

ifornia. According to the California Association of Realtors, 10 % ofCaliforniahouseholds
could afford to purchase the median -priced home in 1989, compared to 45 % of households
nationwide.
39

219

Chapter 2
The design and conduct of this study
A. INTRODUCTION
In the midst of the economic environment described in Chapter 1, the City of Los

Angeles requested proposals in Spring 1989 on a study of the performance of financial
institutions in meeting the credit and banking needs of low income and high minority
Deighborhoods in the City of Los Angeles . This study was also to make recommendations
about the feasibility of a program that linked the awarding of Los Angeles' municipal de

posits to institutions with demonstrated performance in meeting those credit and banking
Deeds. The study began in January 1990 .

Section B sets out the scope of the study, as requested by the City, and our basic
approach in undertaking it. Section C discusses both the conceptual and logistical prob
lems in assessing credit Deeds and institutional performance, and how we handled these
problems. Section D enumerates our activities in conducting this study. The contents of

each of the following chapters is succinctly identified in Section E.

B. THE SCOPE AND CONDUCT OF THE STUDY

Ou contract provided that in conducting this study, we were to undertake the fol
lowing tasks:

• Analyze and rank lender performance on the basis of publicly -available

220
documentation required for information gathering;

• Make recommendations reguding adoption by the City of a linked de

In conducting this study, we begun by amassing information on the current environ
ment ofhousing and economic crisis in Los Angeles, and by gathering contemporary studies
about financial institutions' lending performance in other cities in the United States . This

raised many questions; in particular, in city after city it has been found that residential
lending flows vary systematically with income level and with the degree ofracial concentra
tion. We first used comprehensive data on residential lending flows to find whether these

trends also obtained for Los Angeles. We then extended the models developed nationally
to the case of residential lending ilows to Los Angeles. In so doing, we developed a method
for ascertaining differential lending flows more objective than the methods used in studies

of other cities. We looked first at the correlation between median income and residential
lending flows, and next at the correlation between racial concentration and residential
lending flows.

Our efforts in meeting the City's requests for information of other types - on banking
services, and on lenders' performance in the areas of economic development and affordable
housing development - were very different, because no comprehensive data and few prior
studies were available as guideposts. Both the data and the analytical methods had to
be developed from scratch . We pursued a two -track strategy for obtaining data . First,
we solicited CRA statements from financial institutions. Financial institutions are legally

obligated to produce these statements; so they are almost universally available. These
statements were found to offer only limited insights into banks' lending behavior and their
provision of banking services. Second, we designed four surveys to obtain data on credit
Deeds and financial institution behavior. Specifically , surveys were designed for financial
41

221

institutions thernselves, for developers of affordable housing in Los Angeles, for small
businesspeople in the City's five enterprise sones , and for residents of the City. We had
limited financial resources for this portion of the study. Thus, our research design was to
elicit responses selectively from informed participants in local credit markets. These survey
instruments are available for the City, and can be used in the future to gain information

from operators of small businesses, developers, banks, and residents, if the City puts an
information -gathering mechanism in place. As described in the chapters that follow , we

had varying degrees of success with our own efforts to obtain information using these four
survey instruments.
C. A CAVEAT ON ASSESSING BANKING NEEDS

This study evaluates lenders ' performance in meeting credit and banking needs in
the communities they serve, and particularly in lower income and high minority areas .
In principle, assessing lenders' performance relative to communities' credit and banking

Deeds would require interpreting four factors: ( 1) the type and scope of credit supplied by
lending institutions; ( 2) the range of banking services they supply; ( 3) the level ofcredit and
services provided relative to borrowers' and customers' needs, and those of the community

as a whole; and ( 4) whether the credit and services provided meets the special needs of
lower income areas and residents . An ideally designed study would measure credit and
banking Deeds independently of the observed levels of credit flows and of banking services .

This research design would be able to assess bank performance without prejudging whether
the markets for credit or for banking services were working well or poorly. The data we
assembled for this study - drawn from HMDA tapes, lenders CRA statements , and several

rurveys we designed — are sufficient to directly assess factors ( 1) and ( 2) ; but we can make
only cautious inferences about aspects of factors ( 3) and ( 1) .
Clearly, a comprehensive assessment of even a small geographic area would require a
42

222

tremendous amount of effort and information . To assess the availability and adequacy of

credit flows and banking services for households and businesses in a particular geographic
community, for example, one must then check whether the full range of banking services
and credit needed in each of these substantive areas is provided. Credit reeds for busi
Desses would include startup credit, working capital for operations, and apansion credit;
credit Deeds for households would include finance for home construction , mortgages , re

habilitation, and so on . One would need some independent measure of households' and
businesses ' Deeds in these various areas , with special information on the needs of lower
Income residents. "

This type o comprehensive assessment - to the full satisfaction of both community
activists and lending institutions

is well beyond the resources the City committed to this

study. Indeed , no study of financial institutional performance has even approached this
optimal research design .

This caveat does not make the assessment of institutional behavior in meeting com
munity needs impossible, but it does require careful framing of conclusions. The most

pertinent example occurs in Chapter 4. In that chapter, we demonstrate the existence of
racially disparate lending: that is, we find that lending flows vary inversely with racial con
centration even when median income is held constant. Note that the presence of racially
disparate lending — which this study does demonstrate does not demonstrate the exis

tence of Sedlining" in the sense of discrimination against individuals.

“ Using therhetoric ofeconomic theory, assessing credit Deed is akin to assessing the level
of demandinthe credit market. The difficultyis in measuring demandindependent of
supply: how does oneknow how much credit they would really need if creditwere more
freelyavailable , if in fact lenders have never made credit more freely available. Defining
demand requires being able to construct a credible counterfactual.
43

223

thove subareas with lower income residents. Federal regulator implementing the CRA
have not established any requirement that information from these needs assessments be
done consistently or to any standard, nor that findings concerning credit reeds be publicly
disclosed . We found lenders' CRA statements uniformly uninformative about credit needs

in lower income and minority areas.

As a practical matter, then , the CRA requirement concerning credit Deed assessment

was of no help in our study. The elected officials and civil servants of the City of Los
Angeles would, however , be better situated to evaluate financial institutions' performance

it better credit need assessments were done. This leads to the first recommendation of our
study.
RECOMMENDATION : The City should encourage finan

cial institutions to more seriously and more creatively as
sess the credit needs of lower income and high -minority

Deighborboods in their market areas.

D. A SUMMARY ROSTER OF STUDY ACTIVITIES
The specific activities we undertook in conducting this study are summarized briefly
here for convenience . Additional information on what data were obtained, how these were
obtained, and how these were analyzed is contained in Chapter 3 and 4 and in Chapters
6-9, in sections entitled " Description of the Data . "

224

• PC

225

markets. We established contact with a variety of experts in these areas, with
administrators of cities pursuing similiar studies, and with officials of a number
of Los Angeles financial institutions.

• Recommendations. Finally, we developed a set of recommendations on the basis

E. CONTENTS OF CHAPTERS 3 TO 11
This study of credit flows and banking services in the City of Los Angeles has three

broad components: ( 1) analysis of the provision of residential and other forms of credit,
and the availability of banking services, particularly in lower income and in high minority

areas; ( 2) methods for monitoring and evaluating the performance of financial institutions
with offices in Los Angeles; and ( 3) recommendations about how the City of Los Angeles
might implement policies to bring about a more equitable level of credit flows and of
banking services.

The chapters that follow present the methods, findings, and recommendations of this
study. Much of the data underlying our analysis are included in this report. It should

be noted that much of the specific data prepared for this study, and several subsidiary
sections in particular, the guide to the PC software program for analyzing HMDA
appear only in an appendix which is bound separately from this report on our study.
The organization of material in this report, by chapter, is as follows:
• Chapter 3. Acomprehensive analysis of aggregate patterns of residential lending

. Chapter 4. A comprehensive analysis of aggregate patterns of residential lending
according to the degree of racial concentration ,

226
• Chapter 6.

As analysis of 1989 title transfer data on home sales and loan

227

PART II
RESIDENTIAL LENDING
INTRODUCTION TO CHAPTERS 3-6

Part II focuses on the residential lending performance of financial institutions over
the City of Los Angeles. It uses a comprehensive dataset for the period 1981-89 for all

residential loans reported under the Home Mortgage Disclosure Act ( HMDA ) . The data
contains information on the types of loans made, their number and dollar amounts, and
which financial institutions made the loans by census tracts . There were almost 300,000
residential loans made in the City of Los Angeles in the period 1981-89. Consequently, the
emphasis of the next four chapters is primarily quantitative.

A ranking scheme is developed in Chapter 5 to enable the City to identify financial
institutions with residential lending behavior which is socially responsible. Each institution
48

228
us graded based on its 1989 lending performance over six wreas of special Deed within the

City. A discussion of how this ranking scheme can be wed to implement a linked deposit
or linked banking program is also contained in this chapter.

actions within the City during 1989. This title transfer data is especially useful for its

insight into the role of mortgage companies as residential lenders in the Los Angeles hous
ing market. Most mortgage lenders are too small to report under the Home Mortgage
Disclosure Act. In egregate, however , they form a significant part

229

Chapter 3

Patterns of residential loan flows in Los Angeles
A. INTRODUCTION
The Mayor's Blue Ribbon Committee for Affordable Housing aptly described the
City's housing crisis:

The Los Angeles housing crisis is one of affordability. Soaring costs of a decent

place to livecause a chain reaction resulting in the extreme in thousands of
and the disabled as well asfamilies on welfare; even middle class opportunities
homeless families and individuals .... The crisis affects low wage workers, seniors

for homeownership are virtually gone ."
Two elements combine to generate the housing affordability crisis. On the one hand,

the supply of housing does not keep pace with demand. On the other hand, those who
want housing are increasingly unable to obtain the means to buy it. The availability of
credit is an important aspect of both elements. The volume of new housing depends on
financing being available to developers, particularly developers of affordable housing. The

ability to acquire single family housing is also in large part dependant on prior access to
financing by home buyers.

This chapter assesses financial institutions' residential lending during the period 1981.
89 in Los Angeles, with particular emphasis on how single family and multifamily residen
tial lending varies with income levels or sacial composition of different areas of the City.

Do financial institutions provide ' adequate" credit for rental and owner-occupied housing
in the various areas of the City, as required by the Community Reinvestment Act ? This
is a particularly important question for low -income and high minority communities.
This chapter fummarizes standard, uniformly collected data on the residential loan

patterns of financial institutions in Los Angeles. First, an overview of residential loans
by type of loan and by type of financial institution over the period 1981-89 is provided
· Housing Los Angeles: Affordable Housing for the Future, City of Los Angeles, Blue Ribbon
Committee for Affordable Housing.
80

230

for the City. These residential loan flows are then shown to vary systematically with
w wea's income level and racial concentration. The cumination is based on the detailed

Home Mortgage Disclosure Act ( HMDA) lending data reported by a majority of residential
lender over the period 1981-1989 .

residential lending. Section C describes the organization of the loan data into a system of
tables which appear in Appendix 3 - A but whose results are featured throughout the text. It

also defines the terms used and explains how the census tracts in the City were divided into

five approximately equal poups, “ quintiles" , to characterize the relative median income,
proportion of minority residents, and incidence of crime for the census truets comprising
the City of Los Angeles. Sections D through G provide detailed analyses of the relationship
between HIMDA -reported loan flows and relative levels of income, race , and crime. The

chapter closes with a section underscoring our most troubling findings and a few of the
societal implications thereof.

231

B. HIGHLIGHTS OF CHAPTER 3

1. Loan flows over the City of Los Angeles
( Highlights of Tables 8.1)

• All Residential Loans, 1981-88 : Almost 250,000 loans were made in the City
617,000 residential housing properties.

• All Residential Loans, 1989: Roughly 50,000 loans totalling $ 10.7 billion were

• Market Share by Type of Loan :

1) Conventional single-family constitute the vast majority of residential loans
247,000 or 83%

• Marke : Share of Lenders: Savings and loan associations dominate almost

232

2. Loan Flows by Area Income Levels

( Highlights of Tables 8.3 and 3.3)
• Behavior of Types ofLenders : The three largest categories of lenders; Savings

• All Residential Loans , 1981-88 : Over 83,000 loans made between 1981-88
went to the highest 20 % income areas in the City versus 17,500 loans to the

lowest 20 % income areas. The ratio of loans made between the highest income

20 % and lowest income 20 % was almost 5 to 1, more than double the ratio of 2.25
to one for residential structures between these two areas .
• Income Patterns by Type of Loan :

1) Conventional single-family loans across all lending institutions are heavily

233

8. Loan flows by racial concentration

• Behavior of Types of Lenders The three largest categories of lenders; savings

. All Residential Loans, 1981-88: The lowest 20 % minority areas of the City

• Minority Patterns by Type of Loan :

1) Conventional single-family residential loans exhibit an even clearer lender

234

C. DESCRIPTION OF THE DATA : CHAPTERS 8-5

This chapter melds Home Mortgage Disclosure Act ( HMIDA) residential loan data and
U.S. Census demographic data to walyze the relationship between different characteris
tics of the City at the census tract level - median income, minorities us

proportion of

population , crimes per 1,000 residents and residential loan Rows. Our data sources and
the terms wed in the tables contained in this chapter's appendices are described below .
1. Coverage of Residential Lending Data
We reviewed HIMDA data collected for the City of Los Angeles over the.period 1981
1989 in this study. During this time period, residentiallenders were required to report loans

according to provisions contained in the Home Mortgage Disclosure Act of 1975. Virtually
all commercial banks, thrifts, credit unions, and other entities accepting deposits from the

public were required to collect on any loans secured by residential real estate, or for home
improvements, by census tract. Non -depository mortgage companies owned by depository

institutions were also subject to EMDA reporting requirements ."
Five distinct types of financial institutions appear in this chapter. Each type is regu
lated by a different agency. National banks are regulated by the Office of the Comptroller
of the Currency ( OCC) . State banks that are part of the Federal Reserve System are
regulated by the Federal Reserve and appear in the tables as FRB banks. Non -member

insured banks are regulated by the Federal Deposit Insurance Corporation ( FDIC ) and are
referred to as state banks. Savings and loan associations, previously insured under FSLIC

and regulated by the FHLBB , are now insured under the Savings Association Insurance

Fund ( SAIF ) which is under the jurisdiction of the FDIC . Thrifts we Dow regulated by the
* Certain exemptions were, however, set forth in the 1975 legislation . Institutions with
assetsof $ 10 million orless were exempt from reporting. Institutions werenotrequired
to report: ( 1) loans originated or purchased by trusteesor other fiduciaries; ( 2) loanson
unimproved land; and ( 3) refinancing loans not resulting in an increase in the outstanding
principal involving the same institution and borrower.
55

235

Office of Thrift Supervision ( OTS) .' Credit unions are monitored by the National Credit
Union Administration ( NCUA ) . Other mortgage lending institutions are termed exempt;

they report to the U.S. Department of Housing and Urban Development ( HUD ) .
7. Measures of Lour Flows In EMDA Data
While all loans recorded in the FIMDA data tapes are residence-based , there are four
distinct types of loans: ( 1) government-backed loans for single family residences ( SFR ) ,

encompassing those made under the FEA, FmHA ,and VA programs; ( 2) conventional
loans for single-family residences ; ( 3) home improvement loans for single-family residences;

and ( 1) loans for multifamily residential units. For the purposes of HMDA, single-family
residences are defined as buildings with between 1-4 units while multifamily dwellings

are defined as those with 3 or more units ." We also report separately the loan flows to
DOD -occupants ( which include both SFR and multifamily loans ) .

In the tables ofthis report thrifts are sometimes identified by FHLBB ( Federal Home
Loan Bank Board) andother times by OTS. This change in supervisory names occurred
during the course of the study.
* There is a suggestion of systematic under -reporting of multifamily loans ander FIMDA.

It appears that mixed use ( commercial/residential)

MAP
C3-1
: ity
of
Angeles
Los

/100
.Loons
Bldgs

25.0
50.0
75.0
100.0

Miles
420

236

237

all lenders over the period 1981-88. Dark areas represent areas with high number of loans
per 100 residential buildings, while light areas respresent those areas receiving relatively
few loans per 100 residential buiding. Finally, the average dollar value per loan ( 5) was
constructed, providing a rough indication of differences in real estate values .
8. Measures of Demographic Characteristics from Census Data
We used the following method to analyze demographic variables. Census tracts were

divided into five roughly equal groups, termed quintiles " . Each quintile represents 20 %
of the census tracts in the City of Los Angeles for each demographic variable of interest.
The City of Los Angeles contains 741 census tracts; so each quintile contains roughly 148

census tracts. We established groupings ofcensus tracts for the following variables: median

income, percent African American, percent Latino, percent Asian Pacific, extent of overall
racial concentration, and crime per 1,000 persons . Each such grouping is independent ofall
other groupings. For example, the income rank groupings ofcensus tracts have no necessary

relationship to the racial concentration groupings of tracts. Thus, the same census tract
can be placed in different quintiles for different variables. The correlations between income

quintiles and minority quintiles reflect the underlying demographic makeup of the City.
Appendix 3B contains Tables 3.6 - 3.10 which present demographic information at the

individual census tract level by Council District for the City of Los Angeles .
represents the entire stock of potential residential housing transactions within a census

tract. Our estimate is that there are 616,899 such residential housing units within the
City at the end of 1989.

* Note that average loan size is a crude proxy for average sale price. Whenthe mix of loan

types, the size of down payments ,andotherfeatures of mortgagetransactions varyacross
census tracts, then the relationship between average loan size and sale priceweakens.
A better and more direct estimateof sale price is obtained from the title transfer data
analyzed in Chapter 6.
.The terms minority and facial concentration " are used throughout this report to

denote the combined portion of the population that is African American and /orLatino
and /or AsianPacific. Strictly speaking,however,the formerterm has become a misnomer
as of the 1990 Census, since these three groups Dow account for a majority of Los Angeles
residents .

• For instance, of the 149 census tracts in the lowest income quintile, 90 are in the highest
57

238

4. Locome Measures drawn from Census Data
Using quintile groupings, avoids having to impose ad hoc definitions for high minor
ity areas or " lowest income areas. " Lowest income' census tracts in ou study we defined
as the 20 % of census tracts in the City of Los Angeles with the lowest income fractions.
The income fraction " for a census tract equals its median income, us reported in the 1980
Census, expressed as a percent of the median 1980 income in the Los Angeles-Long Beach

Metropolitan Statistical Area ( MSA ) . For each of the census tracts that constitute the
City of Los Angeles, the median income percent for 1980 was computed. All census tracts

were then rank -ordered from lowest to highest by median income, and divided into five ( 5)
groups- quintiles-

The lowest 20 % of census tracts have income fractions ranging from 0 % to 52% of the
LA-Long Beach median income. The next quintile's income fractions range from 53 % to
73%

of median income. The middle quintile's income fractions range from 74 % to 97% .

The Dext-highest quintile has income fractions ranging from 98 % to 127 % . The highest
quintiles have income fractions ranging from 128% to 395 % of the median Los Angeles
income. These quintiles are labeled, respectively, the 0, 1, 2, 3, and 4 quintiles in tables
appearing throughout this report andin its appendices. Map 3-2 presents a map of the
City of Los Angeles by income quintile. The groupings of census tracts with similar median
income levels suggests that there is considerable justification for speaking of " income areas"
within the City. It is also of interest to compare the pattern of income levels to the patterns
of lending presented in MAP 3-1. There is a close association between high income and
minority quintile while only four ( 1) census tracts are in the lowest minority quintile.

Similarly in the highest income quintile thereare148 censustracts. Of these,01 are in the
lowest minority quintile, and none are in the highest minority quintile. Clearlythere isan
overwhelmingly strong inverse relationship between incomeand minority percent within
the City of Los Angeles.

' There we 149 tracts in the lowest income quintile, 148 in the next highest, 143 in the
middle quintile, and then 153and 148 in the twohighest quintiles. The number of tracts

in each quintile differ slightly because the income- level break points did notoccur precisely
at 20%

intervals.
58

MAP
C3-2
: ity
of
Angeles
,Los
Quintiles
by
Income
Tracts
Census

Qulniile
Income

3:1 3
yt

%
73

4
7:2*
%1
8
:927
J
-

%3-14:SU
99
28
TAKT

Miles

239

%0:-52

240
high numbers of loans per 100 residential buildings.
1

6. Measures of Racial Concentration based on Census Data
The procedure of sorting census tracts into quintiles was followed for tracing the
59

241

African American , Latino, and Asian -Pacific populations in the City of Los Angeles. We

used 1986 figures for both overall and for minority populations.' Breakdowns for each of
these three minority groups are analyzed in the next chapter.

For each census tract, the ratio ofminority nudents to total population was calculated

on a percentage basis. Census tracts were then rank- ordered from lowest ( 0 % ) to highest

( 100 % ) . As with income, census tracts were divided into five approacimately equal groups.
The low minority quintile had between 0% and 18% minority residents, the highest 94 % to

100 % . The results for the various sub -groups appear in Table 3.6 . Map 3-3 presents a map
of the City of Los Angeles by overall minority quintile. The proupings of census tracts with
similar minority percentages suggests that there is considerable justification for speaking
of minority areas " within the City. It is also interesting to compare Maps 3-1 and 3-2

with Map 2-3. There is a close relationship between high minority areas and low income
areas across the City. Similarly, high minority areas and areas with low levels of loans per
100 residential buildings are closely related. The remainder of this chapter quantifies the
association between the three variables, income, race and loans, that is visible in the maps.
Each minority quintile represents approximately 20 % of the census tracts in the City,

but again there is no guarantee that population or the number of residential buildings is
similarly distributed across these quintiles. Chart 3-2 illustrates the distribution of these

two variables across the minority quintiles. Population again turns out to be approximately
uniformly distributed across the quintiles. The lowest 40% minority census tracts have 37%
of the City's population while the highest 40% minority census tracts have about 42% of

the population.

The disparities are once again greater across minority quintiles for the distribution

of residential structures as Chart 3-2 makes clear. The lowest 40% minority census tracts
in the City have 37% of the population but 48% of the City's residential structures. By
' The data for population by census tract in 1986 were obtained from Western Economic
Research and are estimatesobtained by updating 1980 census data with birth and death
statistics for the City of Los Angeles.
60

MAP
C3-3
: ity
of
Angeles
Los
Minority
Quintiles
by
Tracts
Census

Quinlile
Minority
%
15
-

OVE

1:16

%
35

3:26

%
70
-

7:31 .
101

9JX

%14
4
:900

Miles

242

243

CHART 3-1

Popolation

Percentages Across Lacone Quiadles
17.76 %

19.26 %

20.14 %
1
2
[ 03

23.18 %
19.65%

Residendal Bulldings
Percentages Across lacome Qulaules
26.58%
11.84%

00
1 :

12
16.12%

25.29 %

20.17%

244

CHART 3-2

Populados
Percentages Across Minority Quintiles
19.15%

18.29 %

23.76 %
2

18.64 %

20.16%

Residential Buildings
Percentages Across Minority Quioules
.

16.75%

15.66%

24.35 %
1

12
II 3

19.56%

23.67 %

245

comparison , the highest 40 % minority census tracts with 42% of the population have only
32% of the residential structures. Density is clearly higher in the higher minority census
tracts. There is, in fact, a close correlation between high income and low minority census
tracts across the City. Lower minority census tracts often have higher median income levels

and consist primarily ofowner -occupied single -family residences. On the other hand, higher

minority census tracts are most often lower bacome census tracts and so contain higher
Qumbers of multi- family rental dwellings .

6. Crime Measures from Police Data

The final set of quintile groups divide census tracts according to the amount of reported
crimes , based on data provided to us by the Los Angeles Police Department. The crime
measure was compiled by averaging together reported crimes in certain categories over
the 1986-89 period for each census tract and then dividing this figure by the aumber of

tract residents in 1986 times 1,000 . These crimes can be broadly separated into personal
crime ( homicide, rape, robbery, and assault) and property crime ( burglary, larceny, and
car theft ) . As with income and minority measures , census tracts were then rank -ordered
on the basis of the number of crimes per 1,000 residents. The results appear in Table 3.6.

The lowest-crime ( first) quintile ranged from 14 to 50 crimes per 1,000 residents, while the
highest ( fifth ) quintile contained tracts reporting 120 or more crimes per 1,000 residents.90

• Tables 4.1 through 4.4 in Appendix to Chapter 4 contain a breakdown of the number of
census tracts, population , and residential buildings by minority and income quintiles.
1. The highest quintile is purposely left vague because of anomalies in the data. For in
stance, the census tract with the highest crimes per 1,000 residents was Los Angeles LD
ternational Airport with a measure of 4,000. This fabulously high figure arisesbecause

almost nobody lives in the census tract which includes LAX but crime is high because of
the many travelers passing through the terminals.
61

246

D. LOAN FLOWS BY INSTITUTION TYPE :

Table 3.1 in Appendix 8 - A presents statistics for residential loans in the City of Los

Angeles in the period 1981-89; Table 3.1A angements Table 3.1 by showing percentages by
type of lows across financial institutions, both number of loups and dollar amounts, of total

loans. Chart 3-3 presents a percentage breakdown by type of loan for the period 1981-88
and the latest year 1989. This structure of prwenting both historical data and data for
the most recent year is followed in the remainder of the chapter .

Multi- family loans are the third most frequent category, 6% of loan activity. The least
frequent residential loans are single - family , government-backed loans ( FHA , FmHA , and
VA ) . The 5,302 of loans in this category encompasses just over 2%

of all reported loan

activity. The percentage share in 1989 is roughly similar to that for 1981-88 except that

conventional loans are even more important ( 81% by number) and government-backed
loans for first- time homebuyers are virtually non -existent ( 0.14 % .) .

" The totals at the bottom of each column do not add to 616,899 in every case because
theseinclude only the number of residential buildings in all census tracts for which each
type of lending institution made at least one loan over the 1981-88 period.
62

247

CHART 3-3

HIMDA Loans

Percentages By Type of Loan

75.98 %

FHA , VA SFR
Conventional SFR
Home Improve
O Multi- Family
0.80 %

2.31 %
20.92%

HMDA Loans
Percentages By Type of Loan

S of Losas, 1989
81.17%

FHA , VA SFR

Conventional SFR
Home Improve

O Multi - Family
0.14%

248

approximately 21% by dollar volume. As a consequence, the remaining three categories

of residential loan categories have lower dollar percentages than their corresponding share
of the total number of all loans . The most dramatic difference is recorded for home

improvement loans, typically of smaller average size. This category's share of dollars,

2.3 % , is just one-fourth of its 0 % share of total loans.

Table 3.1 also illustrates the type of financial institutions making each type of resi

dential loan. Government- backed loans are provided primarily by thrifts and FRB banks.
Together these two institutions account for about 84 % of all such loans. National banks

are next with 12.8 % ." Note that in terms of dollar flows, thrifts account for a substan
tially larger percentage ( 64.4 % ) than any other institution. National banks are second,
with approximately 25% of dollar flows, and FRB banks are third with only 2.94 % .

Loans as % " and " $ Loans as % * zows display percentages two
1? In Table 3.1A, the
different ways. The first column to the right of the margin calculates percentages on
the basis ofall activity in this ( first) column. For the remaining columns of every row ,
percentages are calculated on the basis of all activity in this two. The same methods of
calculating percentages are used in Tables 3.3A.
63

249

CHART 3-4

AMDA Loan Shares
By Fleancial lasutudos Type

• 2.17 %

OS& L's
W FRB Banks

State Banks
Nat'l Banks
2.89 %

Credit Unions

6.07 %

0.84 %

8.03%

HMDA Loan Shares
By Finansial Institution Type
# of Loans, 1989
79.90 %

S & L'S

FRB Banks
State Banks
Nat'l Banks
2.85 %

3.74%
0.76%

60-893 O - 92 - 9
-

12.75%

Credit Unions

250

followed distantly by national banks ( with approximately 6 % of both flow ) and state
banks ( with approacimately 3% ) . The dominance of savings and loan associations is even
more pronounced in the market for multifamily loans. These institutions comprise about
95 % of these loan flows, with only state banks ( at about 3% of loans and 2.4 % of dollar

dows) being remotely important contributor to the city -wide totals.
Home improvement loans are the only category of lending that savings and loan as

sociations do not dominate. National banks lead here, with 42.4 % of loans and 39.8 % of
dollar volume. Next are state banks, with 19.3 % of loans, followed closely by FRB banks

and thrifts, with 16.5% and 13.4 % , respectively, of loan flows. In dollar volume terms,
FRB banks and thrifts are more important than state banks.

E. LOAN FLOWS BY INCOME QUINTILES:

This section contrasts the residential lending performance of financial institutions in

low and low -moderate income census tracts in the City of Los Angeles over the 1981-89

period. Once again , the census tracts in the first income quintile ( 0) approximate govern
ment definitions of low income households, while the first two income quintiles approximate

governmental definitions of low to moderate income areas. We investigate the differential
treatment received by these census tracts compared to the remainder of the City.
The main finding is that low and moderate income areas receive systematically lower
levels of all types of residential lending by all types offinancial institutions covered under

the EMDA reporting regulations. It appears that contrary to their obligations under CRA
the entire spectrum of financial institutions has concentrated their activities outside of the
special needs low -moderate income areas in the City of Los Angeles.

Two tables, 3.2 and 3.3, contain data on loan flows by income quintiles. Tables 3.2 and
3.2A, and 3.3 und 3.3A present data at the most acgregated level for residential lenders
over the period 1981-89. The figures in these four tables are discussed bere in some depth.
64

251

Subsequent tables contain data for subuets of the information in these four tables ; the
results in these tables will be discussed more rccinctly.

Tables 3.2 and 3.2A are designed to be used together. Table 3.2 itself presents sum
mary data on loans by active institutions of different types, for areas whose residents have
different income levels. Table 3.2A presents some of the aume information reported in
Table 3.2 , but in percentile breakdowns. Whereas Table 3.2 per se shows, for each income
level, the actual sumber and dollar volume of loans per building, and the actual number of

residential buildings, Table 3.2A shows the percentage of each variable that is accounted
for in each income level. The same relationship obtains between Tables 3.3 and 3.3A .
Essentially Tables 3.2 and 3.3 present the scale of lending activities while Tables 3.2A
and 3.3A present the market share or loan portfolio sbares across income quintiles and
institution types .

1. Loan / Income Patterns for All Institutions
Chart 3-3 summarizes the aggregate data in Table 3.2A on the level of total loan flows
to the census tracts by income quintile for the period 1981-88 and 1989. The number of

residential buildings in each quintile increases as one moves from lowest income quintile ( 0)
to highest income quintile ( 4) as Doted in Chart 3-1. One, therefore, expects more loans
in the higher income tracts. The ratio of loans in the highest income quintile to those in
the lowest income quintile, however is almost 5 to 1. This is more than twice the ratio

of 2.3 for residential structures between these two quintiles. The number of loans per 100
residential buildings in Table 3.2 illustrates the gap in loan flows. The wealthiest 20 % of

census tracts received 51 loans per 100 residential buildings, an average annual turnover
of 6.5 percent of their residential structures. In contrast, the poorest 20 % of census tracts
received only 24 loans per 100 residential buildings, an average annual turnover of just 3
percent.

This differential in residential lending is made abundantly clear by Chart 3-5. The
65

252

CHART 3-5

All HMDA Loans
Percentages Across Lecome Quintles

7.14%

33.97%

BO
12.62 %
1

19.10 %
27.18%

All HMDA Loans
Percentages Across Income Quintiles
# of Loaps, 1989
30.69%
7.71 %

12.93%

28.34 %

20.12%

253

bigbest 40% income census tracts received almost 60 % of residential loans by number
over the entire period 1981-89, while the bottom 40% received less than 20 % of loans by
number. Even after adjusting for the slightly higher number of residential buildings in
higher income census tracts it is clear that overall, the higher the median income in a

census tract, the more loons made per residential structure, and the more dollars looned
per residential structure .
Table 3.2 provides suggestive evidence of a housing affordability crisis . In addition to

having a larger turnover rate, the highest income quintile has a larger average loan size
than the lowest income quintile ; $ 186,188 versus $ 107,516. Significantly, however, the ratio
between the two average loan sizes is 1.7, considerably less than the income ratio of 2.5
between these two quintiles ." This suggests a housing affordability gap , since residential

real estate values do not decline proportionately with income, but decline less quickly
than income. The more modest the incomes of the residents in a given area, the fewer the
number of residents able to afford residential real estate in that area ( much less elsewhere) .
Data for the middle income quintiles provides further evidence of the uneven relationship
between income and residential real values . For all institutions, the average loan was about
$ 135,000 for both the second and third quintiles ( 1 and 2) , but only $ 125,000 for the second

highest income quintile ( 3 ) ."

2. Loan / Income Patterns by Financial Institution Types

Tables 3.2 and 3.2A provide more detailed information on loan flows by type of finan

cial institution. The patterns of loan flows across income classes are the same for national
's The ratio between the income level of the most prosperous quintile- o tract and the least
prosperous quintile -4 tract is 2.5; this ratio, however, is between 3 and 4 for many tracts.

254

banks, thrifts, and state banks. Institutions in each category make more loons per building
w income rises, and have higher dollar flows per building w income rises .
With few exceptions, the average dollar value of residential loans is onevenly dis
tributed across income quintiles for each type of financial institution in the aggregate .

Both Savings and Loan Associations ' and State banks ' average loans are roughly the same
across the middle three income quintiles, but the lowest income quintile receives sub
stantially fewer loans, and the highest income quintile receives substantially more . The
pattern for Federal Reserve banks and credit unions is only slightly different. Federal Re

serve banks made more loans in the second highest income quintile ( 3) than in any other,
but their loans in this category are for a substantially smaller average amount than those
made elsewhere. These banks have concentrated their efforts in these areas most heavily
in home improvement loans . Credit unions' totals for loans per building vary the least

with income class, and do not register uniformly higher dollar flows per building in higher
income classes; this pattern , however, have little influence on overall loan flows because of
the insignificance of credit- union totals relative to other institutions ' loan totals.
The market structure of financial institutions nationally has changed drastically over
the period of our analysis here, 1981-89. We attempted to estimate the impact of these
changes on the Los Angeles residential lending market. Our estimate is that between 1981

1988, approximately 12% of residential lenders by loan dollar volume dropped out of the

Los Angeles market. Some of these lenders were closed by regulators, while others were
merged into other financial institutions, hence it is difficult to measure the true impact on
the availability of residential loans. We estimate that between 1988 and 1991, an additional
14%

of residential lenders were closed, merged , or taken over."

It appears that some of these institutions may have misreported home equity loans as
home
improveñent loans, which is impermissible under HMDA reporting guidelines. It
is difficult, however, to assess the magnitude or extent of this problem ,given the HMDA
data lone.

" The majority of takeovers involved purchases of insolvent savings and loan associations
by national or Federal Reserve System banks.
67

255

These failed, merged or taken over lenders do not appear to bave concentrated their
lending activities in low income quintiles. Therefore, market adjustments in recent years
have not significantly affected the distribution of loans between income quintiles across the

City. Indeed, loans by these now inactive lenders are skewed toward higher -income census
tracts. For example, in the lowest- income quintile, inactive institutions account for only
1,335 loans worth $ 10.4 million, adding just 1.83 loans per 100 residential buildings over
the eight years. By contrast, in the highest- income quintile, these institutions account for

9,480 loans worth $ 1.41 billion , adding 4.64 loans per 100 buildings. Almost this entire

difference — 7,614 loans worth $ 1.21 billion is due to inactive thrifts ."
3. Loan /Income Patterns for Different Loan Types
The remaining exhibits in Table 3.2 present lending trends across income quintiles by

different types of loan. Governmentally -backed loans ( FHA, FmHA, and VA) for first-time
homebuyers by income quintile are presented in Chart 3-6 which summarizes Tables 3.2G
and 3.2H. More than half of these loans over the period 1981-88 were made in tracts in the
second -highest income quintile. Dollar flows associated with these loans, however, were
weighted disproportionately toward lower income tracts. FRB banks made 39% of loans in

this category, just below the 45% share attributable to savings and loan associations. This
category of loan is almost non -existent in 1989, so that while the distribution over income
quintiles is more even in Chart 3-6 for 1989, it involves only 138 loans in representing less

than $ 15 million .
The " core " category of residential lending is the conventional single- family residential

loan . Conventional SFR loans by income quintile appear in Chart 3-7, summarizing Tables
3.21 and 3.2J. This category is dominated by thrifts over the entire period 1981-89, which
make 90 % of these loans. State commercial banks account for 5% of lending, other banks
for 4% , and credit unions for less than 1 % .
" ' For details of closures by type of institution refer to Table 3.2D .
68

256

CHART 3-6

FHA, FmHA, and VA SFR Loans
Percentages Across Iacone Quintiles

7.71%
12.83 %

12.71 %

51.21 %

III 3

15.54 %

FHA, FmHA, and VA SFR Loans
Percentages Across Income Quintiles
# of Loans, 1989
5.80 %

29.71 %
19.57 %

3
4

18.12 %
26.81 %

257

SingleCamily residential loan flows are heavily concentrated to upper-income census

tracts 1981-88 and 1989. For all active institutions in the 1981-88 period, there were 2.6
loans per building in the top income quintúe for every 1 loon per building in the bottom
quintile. This is an even greater imbalance in turnover rates than for total loans. Further,
average loan size increases uniformly with income. Thus, the flow of dollars per building was
6.6 times higher in the highest than in the lowest quintile tracta. Except for credit unions,
the figures for every category of lending institution reflect lending flows that rise with

income for both dollars -per-building and loans-per- 100 buildings. Chart 3-7 graphically
illustrates the larger share of loans going to the highest 40 % income census tracts ( over
60 % ) and the disproportionately low share going to the lowest 40 % income census tracts
( less than 20 % ) .
The data displayed here also provide indirect evidence on the housing affordability

crisis. The ratio of average house price, implied by the HMDA data, between the highest
and lowest income quintiles is 2.5, substantially more than indicated in Table 3.2B. This
is due primarily to the fact that the average SFR loan size is vastly lower for the lowest
income tracts ( $ 77,163, versus $ 105,406 in Table 3.2B) , but relatively constant for the
highest- income tracts ( $ 192,000, versus $ 182,400 in Table 3.2B) .' * Average SFR loan sizes
in the three middle- income tracts, however, are all clustered between $ 91,200 and $ 115,000,
figures substantially less than the average price recorded for the highest-income tracts. This
relatively " at " distribution of home prices does suggest an affordability crisis, as did the

evidence in Table 3.2B.19
Home improvement loan data are displayed in Tables 3.2K and 3.2L. Overall, both
1 This greater disparity is largely due to the fact that these data exclude multifamily loans,

which have much higher averagedollar volumes and are more evenly distributed than SFR
loans .

1. The ratio of average home prices for the second -highest and lowest income quintiles is
$ 115,036 / $ 77,163, or 1.49. By contrast, the ratio of incomes for residents in these two
quintiles is at least 1.88, and frequently higher. So as above, the distribution of home
prices. -as reflected in average loan amounts — is " fatter " than the distribution of average
incomes .
69

258

Chart 3-7
Conventional SFR Loans
Percentages Across Lacome Quintiles

6.20 %

36.63%

Jo
11.20 %

18.64 %

27.33 %

Conventional SFR LOUDS
Percentages Across locome Quintiles
# of Losos, 1989

31.98%

7.17%

Oo
11.91 %

19.12 %
29.22 %

259

the oumber and the dollar volumes of home improvement loans per residential building
rise steadily with the income quintile. There are 2.14 of these loans per building in the top
income level for every 1 in the bottom income level, and the low of dollars per building was
3.2 times higher in the former tracts compared to the latter. Savings and loan associations
do not dominate bome improvement lending; they account for only 15% of the number
of loans, and 21% of the dollar volume. This category is dominated by national banks,
which account for about 40% of loan activity, followed by FRB and state banks, which
together account for another 35 % . Credit unions have a small but appreciable share of
the home- improvement market ( 7 % of all loans made, and 5% of dollar volume) . Again ,
a relatively uniform bias toward upper -income tracts is found for all categories of lending
institution .

Multifamily rental building loans by income quintile appear in Chart 3-8, using infor
mation contained in Tables 3.2M and 3.2N . This category clearly bears on the question of
affordable rental housing for low and moderate income areas. Almost half these loans by

number over the entire period 1981-89 were made in the lowest 40% income census tracts
within the City. This is no surprise given the greater density observed in Chart 3-1 which
implied the majority of multifamily residential structures were to be found in the lower

income quintiles. Almost all multifamily lending activity was accounted for by savings

and loan associations ( 96 % of loans, 95% of dollar volume.) .' It is also unsurprising that
average loan size increases with the income quintile, from $ 307,279 in the poorest quintile

to $ 704,354 in the wealthiest quintile. While this association reverses that found in every
other loan category, it is important to recall that multifamily dwelling loans account for
only 6% of the total number of loans.

" Once again it is important to note that we suspect substantial under-reporting in this
area by financial institutions, which often have separate divisions whose operations fall
under the rubric of commercial, rather than residential, lending for projects of this type.
70

260

CHART 3-8
Multi - Family Dwelling Loans
Perceolages Across Lacome Quintles

6.91 %
16.40 %
17.64 %

Oo
1
2

27.51 %

31.54 %

Multi - Family Dwelling Lowas
Percentages Across locome Quintiles
# of Loans, 1989
14.74%

5.98%
17.83%

Oo
1

III 3

27.81 %

33.63%

261

4. Summary Data on Loan / Income Patterns
Charts 3-6 through 3-8 are particularly valuable for presenting summary data on lend
ing for each category of residential loun. There is a clear and consistent inverse relationship
between income quintiles and both conventional SFR loans and home improvement loans.

The advantage of wealthier quintiles over pooru ones is even more pronounced in dollar
flow terms than in numbers of loans. Government-backed loans aimed at first- time home
buyers are spread relatively evenly across quintiles in dollar terms, but are concentrated in
the second highest income quintile in terms of the numbers of loans. Multifamily loans are
concentrated primarily in the second and third quintiles ( accounting for 50 % of all such

loans) , and secondarily in the first and fourth quintiles ( 34 % ) . In dollas flow terms, most
financing in this category has gone to the second, third , and fourth income quintiles.
Tables 3.3 and 3.3A reveal systematic patterns in the flows of residential credit by in

come quintile. The lowest income quintile receives almost do conventional SFR or home im
provement credit but significant amounts of government-backed and multifamily dwelling
loans. The pattern for the second income quintile is about the same, with somewhat
larger flows in each loan category. The third quintile, accounting for 20 % of residential
buildings, receives the same proportions of government- backed and multifamily loans as

does the second quintile but its conventional SFR and home improvement loans are larger,
though still smaller than its 20 % share of residential structures. The second highest in
come quintile receives the largest share ( among quintiles) of government-backed loans, and
a large share of multifamily loans. This quintile's Rows of conventional SFR and home
improvement loans are higher in numeric terms than its share of the residential building
stock , but lower in dollar terms. Only the highest income quintile, then , receives flows
of conventional SFR and home improvement loans that unambiguously exceed its share
of Los Angeles' residential building stock : indeed, this quintile receives 50 % of the dollar

filows in these two loan categories while its share of residential buildings is only 26 % . Yet
its levels of government-backed and multifamily loans are comparable with those of the
71

262

first quintile, for which the flows of conventional SFR and home improvement loans are
almost non -existent.

This comparison gives some perspective on which areas and loan categories have been
adversely affected by institutional mergers and closures during the 1981-88 period. The
evidence in Table 3.3D indicates that the gap between active fostitutions' and all insti
tutions' loan flows is about 10 % in primeric terms, and 8% in dollar termos within the

period 1981-88. This gap is largest ( about 14 % in numeric terms, 18% in dollar terms) for
government-backed loans. It is next largest for conventional SFR loans. Home improve
ment and multifamily loans are little affected . Since 1988 a further 14 % of the residential

lenders by dollar volume have been subject to closure, merger , or takeover. How this has

F. LOAN FLOWS AND RACIAL CONCENTRATION :

This section contrasts the residential lending performance of financial institutions

across areas of the City of Los Angeles with different degrees of racial concentration. Racial
concentration is measured here by adding together, for each census tract, the populations
of African American , Latino, and Asian Pacific residents into the category minority."
Evidence for residents in each of these three groupings is presented in the next chapter.

* For government-backed loans, first income quintile census tracts have experienced the
largest decline, followed by the second and third quintiles. The evidence forconventional
SFR loans reveals, to the contrary , that this gap between active and all institutions is

somewhat larger for the third through fifth quintiles than for the bottom two .
72

263

spectrum of residential lenders have concentrated their activities outside of the special
Deeds, high minority areas within the City.

1. Overall Trends for Loans and Racial Concentration

Tables 3.4 and 3.4A in Appendix 3 - A provide data on 1981-89 HIMDA loan flows by
minority quintile across the city . The main result of this section is that the lower the

proportion of minority residents in a census tract, the more residential loans are made. In
assessing whether loan flows are uneven according to minority population, however, recall
there are more residential buildings in the three lowest minority quintiles than in the

two highest minority quintiles ( Chart 3-2 ) . Even after adjusting for the number of loans
per building, however, almost 2.5 loans per building were made in the lowest minority
quintile for each loan in the highest minority quintile. This pattern is consistent across
the intermediate minority quintiles as well. Chart 3-9 illustrates that the pattern of lower

loan flows to higher minority areas was roughly the same in 1989 as it was over the period
1981-88. The lowest 40% minority census tracts have received over half of all residential
loans by number over the period 1981-89. By contrast the highest 40% minority census
tracts receive only 20 % of residential loan flows, far lower than their share of the City's

residential structures.

Loans per building are significantly lower in the two highest minority quintiles - that
is, when minority residents become 73%

or more of the population. The dollar flows of

loans are more heavily weighted toward tracts with low proportions of minority residents.

There were $ 5.73 in residential loans per building in low minority quintile for every $ 1 in
the highest minority quintile.

264

CHART 3-9

All HMDA Loans
Percentages Across Mlaority Quintiles

9.26 %

12.80 %

30.62 %

1
2

20.67 %

II 3
4

26.65 %

All HMDA Loans
Percentages Across Minority Quintiles
# of Loans, 1989
10.11%
12.99%

Do
28.75%
2

20.65%
14

27.50 %

265

CHART 3-10

FHA, FmHA, and VA SFR Loans

Percentages Across Minority Quintiles
21.03%

7.13%
Do
9.66 %
13.54%
1
2

48.64 %

266

CHART 3-11

Conventional SFR Loras

Percentages Across Minority Quintiles

7.93%

4.05 %

15.66%

DO
1
2

III 3
46.22 %

267
1

Credit unions' lous flows are relatively independent of the minority composition of tracts .
However, this trend must be carefully interpreted, because credit unions have made so few
loans, and their average loan size is so small relative to that of other institutions.

3. Race /Loan Patterns by Type of Loan
Tables 3.4G - P present results for minority concentration by type of residential loan .
These tables reveal diverse patterns for different categories of loan .

Further , total dollar flows are higher for tracts with the highest proportions of minority
residents . Since there are fewer residential buildings in high minority tracts, these patterns

are even more pronounced on a per-building basis. The data for the various types of lending
institutions are remarkably uneven . Of the 5,302 governmentally -backed loans made over
the 1981-88 period, some 40% were made by FRB -chartered banks in the middle quintile
( where between 36% and 70% of residents are minorities) . Further, 20 % of these loans
went to the highest minority tracts from thrifts, which made a further 20 % of these loans

to the third and fourth quintile tracts. This left only 1,317 loans to be made outside of
these areas and institutions.

Chart 3-11, summarizing Tables 3.41 and 3.43, reveals an even greater pattern of
lender preference for low minority tracts when making conventional loans for single-family
residences than is evident in the data for overall loan flows. The lowest 20 % minority

census tracts received almost ball the number of conventional SFR loans made in the City
over the period 1981-88. By contrast, the highest 20 % minority census tracts received less
than 5%

of the conventional single - family residential loans during 1981-88, even though
74

268

they contain almost 17% of the City's residential structures. This pattern of precipitously
lower loans to higher minority areas is true for each category of residential lenders, except
credit unions.

The ratio of 2.73 loans to lowest minority quintile for each one made to the highest

minority quintile exceeds the ratio of 2.27 to 1 found across all loan categories. In terms
of dollars per building, this ratio is even higher with $ 7.85 in the lowest minority quintile
for every $ 1 in the highest minority quintile. This is more extreme than the $ 5.73 to $ 1
ratio between thes two quintiles for total loans.

buildings and for dollar volumes of these loans all reflect a preference for tracts with lower

proportions of minority residents.
Chart 3-12 , summarizing Tables 3.4M and 3.4N , provides information on multifamily

loans. As discussed above, these loans are fewer in number but larger in average size than
other loans. The data on these tables indicate two interesting patterns in multifamily loan
flows. First, these loans have disproportionately flowed to the third and fourth quintiles

that is, to veas with between 36 % and 93 % minority populations while the lowest number
of multifamily loans were made to the first ( lowest minority) quintile. Second , the average
loan size varies inversely with minority population. Indeed , the average multifamily loan
was $ 734,227 in the lowest minority quintile, but a third that, $ 227,823, in the highest
minority quintile. Since savings and loan associations have made virtually all multifamily
loans, the same trends obtain in that category.

269

CHART 3-12

Multi -Family Dwelling Loans
Percentages Across Minority Quindles

14.86%

13.88%

Do
26.29%
1

2
WII 3
17.95%
4

27.02 %

270

G. LOAN FLOWS AND CRIME QUINTILES:

Tables 3.5 and 3.5A contain data on loan flows broken out by crime quintiles. The
number and dollar volume of loans by crime quintile is of interest because the number of
aimes represents a crude but available index of the Siskiness of a census tract. One

might expect loan iows to vary inversely with crime quintiles as a reflection of this risk.
The previous two sections have illustrated the strong association between loan flows, on
the one hand, and income and minority status on the other. There is an overwhelmingly
strong connection among these factors in the data.

These results do not prove," however, that banks necessarily base their lending deci
sions on these two factors. There may be other factors affecting the level of loan flows such

as the level of demand for loans, the profitabilityof the loan, or the degree of riskiness of
the area. If these other economic factors are, in turn , highly correlated with either census
tracts' median income levels or proportions of minority residents, then the apparent corre

lation between loans and income or between loans and minority status - could simply be
a reiection of a correlation between loans and these other demand or risk -related factors.
It is difficult, however, to find variables that are available at the census tract level. Crime

statistics are investigated in these tables because they offer a useful proxy for one poten
tial economic rationale behind differential loan flows - the riskiness of the areas in which
residential loans are made.

The data in Tables 3.5 and 3.5A indicate some sensitivity of loan flows to crime
quintiles: the number of loans is uniformly lower , the higher the crime quintile, and the
first quintile scores well ahead of the fifth on most measures of loan flows. However, these
patterns do not obtain across all quintiles. Further, there is no evidence of any pattern for
average loan size. The relative insensitivity of loan flows to crime quintile is also found in

the data on total loan ilows, on 1989 loans considered separately, and on loan flows by type
of loan ( all of which are contained in Tables 3.5B - 3.5P , contained in the appendix to this
76

271

chapter ) . The weak correlation between crime quintiles and loan activity is, in a sense,
a non - finding, but it is an important clue ponetheless to the nature of residential lending

activity in Los Angeles. The relatively weak relationship between loan iows and crime
quintiles implies that loan flows are not extremely sensitive to crime. Thus, the stronger
correlations found between loan flows and income, and between loan flows and minority
status, are evidently not attributable to the indirect effect of crime on loan flows.

H. SUMMARY : SOCIETAL IMPLICATIONS
OF THE CORRELATION BETWEEN RACE , INCOME,

The evidence in this chapter bears on the affordable housing crisis in several important

ways. Government-backed loans do not meet the needs of the low and moderate income
areas simply because median home prices in the City of Los Angeles have consistently

exceeded the federal maximum loan amount under these programs. Consequently the
majority of residential mortgage loans are conventional; these have, in the past, been
largely originated by thrifts. Savings and loan associations representing at least 14% of
the 1981-88 totals have been closed or merged since 1988. Of the thrifts that have been
merged, most have done so with one of four large national banks. These large banks

have in the past concentrated on home improvement, rather than loans for the purchase
of residences . Taken together, this evidence raises the spectre of a declining supply of

residential lending in at least the immediate future.
The most troubling aspect of the empirical results is the aggregate behavior of lenders

in meeting the credit needs of special need areas and groups. Evidence from the HIMDA

residential lending data strongly indicates that, in the aggregate, all types of financial
institutions make systematically lower levels of residential loans of all types to low - income
and high minority census tracts within the City of Los Angeles. Whatever economic
77

272

rationale might be advanced for this pattern , it is clearly not consistent with the goals
that lenders are mandated to pursue under the Community Reinvestment Act.

* Analysis based on per-capita crime figures indicate that neither income por minority
populations are acting as proxies for general area " riskiness " , since loan flows do not vary

with per capita crime in any significant manner.
78

273

Chapter 4

Race, residential lending, and income
A. INTRODUCTION
This chapter analyzes the relationship between residential lending flows and race con
trolling for the influence of income. The results reported in the previous chapter established

that high minority census tracts receive far lower levels of residential lending than do census
tracts with fewer minorities. It was not possible, however, to determine whether these dif
ferentials resulted from discrimination based on applicants' income levels, or from racially

disparate lending even when income is held approximately constant. Lenders ' decisions
about whether to lend to a borrower depend on many factors. Race is just one possible
factor. There are others, including borrowers ' income levels, credit histories, employment

records, and education levels.
The main results of this chapter are strong and troubling. There is compelling evidence

of de facto racially disparate lending in the aggregate by residential lenders throughout the
City of Los Angeles. It appears that African American and Latino populations are primarily
affected. Asian Pacific minorities, as a whole, do not seem to be adversely impacted by
racially disparate lending.

by contrast, occurs at an individual level. Lending decisions to individuals depend on
non - economic variables, such as race, rather than economic variables. Racially disparate
lending and discrimination are related , but evidence of racially disparate lending need not
prove the existence of discrimination ."

* Note that the terms " racially disparate lending " and " redlining" are not identical. The
79

274

It is beyond the capacity of the data used here to determine conclusively whether
" discrimination on the basis of nice irrefutably exists in the City of Los Angeles . This
chapter presents, however, an analysis of the HMDA residential lending data that allows
a more nuanced understanding of the relationships between lending flows and ethnicity in

Los Angeles.

Our principal results are described in the highlights section which follows immediately.

Chapter 6 explores the same set of questions using a different data source — figures for
individual housing sales in the City of Los Angeles during 1989, us reported to the County

Assessor's Office for purposes of title transfer. These data provide a much more detailed
view of conditions and terms in the residential housing market.

B. HIGHLIGHTS OF THE ANALYSIS
1. Patterns of Residential Segregation by Race
• Low income areas in the City are almost always also high minority areas. Simi

latter term - two definitionsof which are provided in Chapter 2 -is an appropriate descrip

tion when it can be established that racial disparities in lending remain even when the
various other factors that might affect creditworthiness have been taken into account.
The aggregate data available in this study are clearly sufficient to establish the presence of

racially disparate lending patterns, but not to establish the presence of redlining as defined
in Chapter 2.

* The only conclusive way to demonstrate discrimination is to have two applicants, iden
tical in every economicaspect, apply to a lender for the same loan. If one applicant is
denied while the other is accepted, a bias based on the non -economic variables in which
the applicantsdiffer is proven . This difference may be the applicants' race or the racial
composition of the area in whichtheproperty is located. Tests for discrimination have
been conducted in cities which exhibit aggregate lending patterns similar to those for the

City of Los Angeles. The general outcome of these tests has been that discrimination " is
indeed presentwhen aggregate patternssuggest it. We knowof nosuch tests conducted
in the City of Los Angeles, however.
80

275

is a high degree of racial segregation in the City of Los Angeles.

• Loans, both numbers and dollars, chibit . precipitous decline for census tracts

8. Residential Lending, Income, and Race
• A review of the HIMDA data presents overwhelming evidence of racially disparate

C. PATTERNS OF RESIDENTIAL SEGREGATION

Before turning to patterns of residential lending, we first analyze the patterns of segre
gation , and their relation to income levels, within the three sub- populations of " minorities "
in Los Angeles. Charts 4-1 through 44 visually summarize the data in Tables 4-1 through

h respectively, depicting the extent of segregation across the City of Los Angeles by type
of minority and income levels. Each of these charts presents information on the percent

of the City, measured by the percent of census tracts, by both minority concentration and
income quintile. Each chart involves a grid with income quintiles along one axis and minor

ity quintile along the other. The height of the column rising out of each square of this grid
81

276

Chart 4-1

Overall Minority by lacome Quindles

14.00

12.00
10.00

8.00

of City

4: Highest Locome
3
2

Highest Minority: 4
Minority
Quindle

0. Lowest Locome
Locome
Quiotile

277

represents the percent of the City having this particular minority /income characteristic.

For example, in Chart 4-1 the lower -most square on the grid represents the percentage
of the City that is in both the highest minority quintile and the lowest income quintile.
The height of the column represents the percentage of the City having this characteristic,
approximately 12% of census tracts in the City '

Latino minority quintile census tracts appearing in the bottom two income quintiles; in
the highest income quintile there are no census tracts in the two highest Latino minority

* Each row of the grid will sum to roughly 20 % along either the minority or income quintile
axes. This is a consequence of the construction of Tables 4-1 through
in the corresponding Chart .

which is mirrored

* Even more starkly, 130 of the 149 low income census tracts are high minority, a staggering
87 %

of the quintile.

Similarly , 131 of the 148 census tracts in the highest income quintile are low minority,
fully 88.5% of the quintile.
82

278

Chart 4-2

Black Mlaority by Income Qaladles

-8.00

-7.00
-6.00

-5.00
-4.00

Lowest Minority : 0

ghest
2

Highest Minority 4

0: Lowest Locome

Black Minority
locome

Quintile

_
279

Chart 4-3

Hispanic Minority by Lecome Quintiles

12.00
10.00

8.00

6.00

of City
4.00

2.00

4: Highest locome
3

Highest Minority: 4
Hispanic Minority

0. Lowest Locome
Income
Quintile

280

Chart 4-4

Aslan Minority by Lacome Quintiles

12.00

10.00 -

8.00

6.00

of City
4.00
2.00

4: Highest Income
0.00

Lowest Minority: 0

Highest Minority: 4

0. Lowest Locome
Income
Quintile

Asian Minority

281

quintiles.

It is important to establish the relationship between African American and Latino
populations within each income quintile. For the three lowest income quintiles the majority
of the census tracts bave high overall minority populations, hence a lov African American
minority census tract is often associated with a high Latino population. This correlation
serves to explain certain patterns observed in the loan data. In particular, as Chapter

3 illustrated, loan flows are lower to high minority areas generally. Hence in the lower
income quintiles, both the highest African American ( low Latino) and the lowest African
American ( high Latino) minority census tracts will have low loan flows because both sets
of areas have high overall minority populations. The same is true for loan flows to the
highest and the lowest Latino population census tracts in the lowest three income quintiles

( bottom 60 % of census tracts) for the same reason .
In the top two income quintiles the majority of census tracts are low minority, hence

low African American or low Latino populations typically imply census tracts that are
primarily low minority. In the top two income quintiles we observe a different pattern: loan

flows are highest in the lowest minority ( primarily Caucasian) tracts, gradually decreasing

as the minority population percentage increases; loan flows then drop off sharply in the
highest minority census tracts ( both African American and Latino) . This pattern suggests
racial " ipping" , wherein some trigger level ofminority residents drastically reduces lenders'
willingness to make residential loans.

Tables 4-1 through 44 present overall patterns for each minority subgroup throughout
the City, independent of income levels. We first look at the pattern for African Ameri
cans. In the 20 % of census tracts with the lowest number of African Americans, African
Americans make up less than 0.5% of the population. In the next highest African Ameri
can quintile, African Americans constitute between 0.3-2 % of residents, and in the middle
African American minority quintile, just 2-4 % . The fourth quintile consists of tracts with
between 5% and 29 % African American residents, and the fifth quintile has a concentra
83

60-893 0 - 92 - 10

282

tion of African American residents between 30 % and 98 % . Some 60 % of census tracts in

Los Angeles therefore are almost completely segregated in terms of the African American
population

In the 40 % of census tracts in which African Americans constitute at least 5% of the
population, the degree of segregation is even greater if the presence of other minorities is
taken into account. For example, census tracts in the fourth African American minority
quintile have populations between 5% and 29 % of African Americans among residents, but
some 37 % of these tracts have overall populations that are at least 71% overall minority
population, and 53 % of them have overall populations that are 80 % or more minority.

Even tracts that are ethnically diverse - having some African American residents, but not
an overwhelming proportion - are, in the main , populated by significant percentage of all
three minority subgroupings. Census tracts with significant aumbers of African American
residents and a majority of low minority residents are extremely rare.
Table 4.3 presents the pattern for the Latino population. The quintile with the small

est proportion of Latino residents encompasses tracts with between 0 and 5 % Latinos .
This 5% figure was the breakpoint for the fourth quintile of African Americans, indicating
that Hispanics are not only a larger percentage of the overall population, but are also
spread more widely throughout the City the African Americans. The next three quintiles

all encompass census tracts in which most residents are non - Latino. Also, these census
tracts are likely to have a majority of low minority residents. The fifth quintile, however,

is composed entirely of census tracts having a majority of Latino residents — 51 % to 100 % .
Finally, the dispersion of the Asian Pacific population in Los Angeles ( Table 4.4) shows
that the vast majority of census tracts have either insignificant ( 0 to 3% for the first two

quintiles) or relatively small ( 4 to 14 % ) number of Asian Pacific residents. In part, this
happens because there are relatively fewer Asian Pacific residents than African Americans
and Latinos. This pattern is also caused by the fact that Asian Pacific residents are thinly
spread across most of the census tracts in Los Angeles. Census tracts which do have larger
84

283

proportions of Asian Pacific residents, however, also tend to be primarily populated by
other minority residents. For & ample, consider the highest Asian Pacific quintile, which
encompasses tracte with between 15% und 97 % Asian Pacific residents. Just over three
quarters of the tracts in this quintile have overall minority populations in excess of 50 % .

For the next highest Asian Pacific quintile ( 7 to 14 % ) , about 44 % of the census tracts have
more than 50 % minority residents.

The overarching conclusion here is that once the members of any minority sub-group
become significant in a census tract, that cenous tract js likely to be populated primarily by
minority residents . There is therefore a high degree of racial segregation in the City of Los
Angeles. Minorities are more likely to live in an area populated primarily by minorities ,
while low minoritys are more likely to live in areas that are primarily low minority. The
remaining sections do not seek to address the causes of this segregation, but rather the

effects that this racial segregation has had on the patterns of residential lending.
D. RESIDENTIAL LENDING AND RACIAL CONCENTRATION

This subsection reviews data for each of the three sub-groups within the “ minority "
category. The data discussed appear on Tables 4.6-4.8, each of which encompasses two
pages. Table 4.5, also included here, contains information about all minorities in the
same format. Once again we rely on charts to present this information, chart Bumbers 4-5
through 4-8 corresponding to the same numbered tables. These charts presents information
on the number of loans per 100 residential structures across the various types of minority
quintiles. They also show the composition of these loan flows by the four types of residential
loans.

Chart 4-6 presents data for lending by African American minority quintiles. As noted
in the preceding section, the first three quintiles have very few African American residents

( 1 in 25 or less) . The fourth quintile is somewhat diverse, primarily because of the presence
85

284

Chart 4-5
Loan Flows by Minority Quiddle

50.00
45.00
40.00
35.00

30.00

# of Loans
per 100 Res. 25.00
Bldgs.
20.00
15.00

10.00 -

5.00

0.00
0

Minority Quintile

FHA Or VA SFR

Conventional SFR

Home

O Multi-Family

285

Chart 4-6
Lord Flows by Black Minority Quintile

45.00

40.00
35.00

30.00

of Lous 25.00
per 100 Res.
Bldgs .

20.00
15.00

10.00

5.00

0.00
1

FHA or VA SFR

Conventional SFR

Home

Improvement

O Multi- Family

286

of other minorities. The fifth quintile is composed largely of African American -dominated

tracts. Interestingly, the overall data suggest that residential lending flows ( particularly
on & per building basis ) are essentially flat for census tracts in the first four African
American minority quintües, but exhibit a precipitous decline for census tracts in the highest
African American minority quintie.

This overall pattern is due to the nature of flows for conventional SFR financing. In
the period 1981-88, between 334 and 369 loans were made per 100 buildings in the first four

African American minority quintiles, but only 186 in the highest African American quintile.
In dollar- flow terms, the difference between the first four African American quintiles and
the highest quintile is even more pronounced: 3-to 1: or worse. For home improvement
loans, the number of loans do not reflect this precipitous decline, but the dollar volume

pattern does. Multi-family loans diverge from this pattern , however, because we suspect
that these loans are under- reported it is not clear whether this divergence is real or simply

the result of the under-reporting. In numerical and dollar terms, most loans in this category
are made in the third and fourth quintiles. Government-backed loans are, in dollar terms,
most prevalent in the fifth African American quintile. These government-backed loans,
however, still constitute an insignificant amount compared to conventional mortgage loans .

Latino quintile data are presented in Chart 4-7, summarizing the results of Table 4.7.
As already noted , the Latino population is more uniformly spread across census tracts

than are African Americans. When Latinos constitute a minority in a census tract, that
tract is more likely to have more than 50 % low minority residents than it is to have more
than 50 % minority residents. Perhaps reflecting these patterns, the residential lending
data for Latino quintiles do not show the precipitous decline that data for the highest

African American minority census tracts do. Instead, the number and dollar volume of
overall loans declines smoothly from the lowest to the highest concentrations of Latinos in
the population

Conventional SFR loans also exhibit this smooth pattern of decline. For example, the
86

287

Chart 4-7

Loan Flows by Hispanic Midority Quindle

30.00
45.00

40.00
35.00 -

30.00 # of Loans

per 100 Res . 25.00
Bldgs .
20.00

15.00

10.00
5.00
0.00
0

FHA or VA SFR

Conventional SFR

Home

Improvement

O Multi -Family

288

first quintile receives 1.86 such loans for every 1 in the fifth quintile, and $ 4.23 for every
$ 1 in the fifth quintile. The data for home improvement loans provides weaker evidence

of this smooth inverse relationship between Latino quintiles and lending flows. The only
trend evident in the data for government- backed loans is that the fourth Latino quintile
receives the lion's numerical share. The data for multi-family loans also yield no definitive

trends. It is clear that, in general, the higher the percentage of Latinos in e census tract,
the lower the level of loan flows, both in numbers and in dollars, to that census tract. This
is consistent with the pattern established for overall minority percentages and loan iows.

areas .

E. RESIDENTIAL LENDING , INCOME, AND

The above results suggest that lending flows decline when there is a large concentration
of African American residents, or an increasing percentage of Latinos in a census tract.

One explanation for this phenomenon might be that there are more Asian Pacific banks
in the City relative to other minority-owned lenders, hence the AsianPacific community
is better able to offset any differential in loan flows.
87

289

Chart 4-8

Loan Flows by Asian Minority Quintile

40
35

# of Loans 25
per 100 Res.
Bldgs.

kilku

S
0

Asian Minority Quintile

FHA or VA SFR

Conventional SFR

Home

Improvement

D Multi- Family

290

This pattern is suggestive of racially disparate lending, but is not definitive proof. There
is clear evidence of lower numbers and dollars of loans to areas where Latinos, African
Americans, and Asians are found in high concentrations than to low minority areas. It is

not within the province of this study to offer definitive proof of racial redlining, if racial
redlining is strictly defined us lower loan flows based solely on the racial characteristics of
an applicant or of a neighborhood , and not based on any other factors linked to borrower
creditworthiness. It is possible, however, to provide results that separate the effects of
income from the effects of race on the residential lending decision.
It is important to determine exactly what constitutes compelling evidence of racially
disparate lending in the context of our data . First, one must correct for differences in the

scale of potential" loans across census tracts. High income, low minority census tracts
are primarily single family residential. Therefore , they may have higher absolute levels
of numbers and dollars of loans, simply because there are more buildings on which loans
can be made. We, therefore, examined numbers and dollars of loans per 100 residential
buildings in assessing differential loan flows.

It is also crucial to isolate the effects of economic variables, us opposed to racial vari
ables, on the lending decision . This is accomplished by confining analysis to census tracts

within a given income quintile, and then investigating the relationship between loan flows
and racial characteristics within the income quintile. Income is being held approximately
constant, so loan flow differentials reflect only differences in the racial composition of cen
sus tracts . If racially disparate lending is not an important feature of the data , loan flows

per 100 buildings should be roughly constant as minority population changes across census
tracts with similar median income levels.
Our analysis was complicated by the excessive polarization of high minority popu
lations within the lowest income census tracts, and of low minority populations within

the highest income census tracts. There are almost no lowest incomenow minority census
.88

291

tracts, and similarly almost no highest income high minority census tracts .' As a result it

is impossible to examine how loan flows vary with race in the lowest and highest income
quintile census tracts. The majority of the analysis is therefore confined to the relation

between loan flows and nice across census tracts belonging to the middle three income
quintiles, since only in these census tracts is it possible to get a diversity of racial compo
sition .

The remainder of this section investigates the relationship between loan flows and

racial concentration, with income held roughly constant. We first examine the pattern for
all minorities taken together, and then analyse the patterns for African American , Latino,

and Asian Pacific concentrations in turn .
1. Lending Flows and Overall Racial Concentration

Charts 4-9 through 4-12, summarize data contained in Tables 4-9 through 4-12, on
lending flows by minority quintiles, with income held constant ( in quintile terms) . The lay

out of these charts is similar to that of Charts 4-1 through 44, presenting information by
both minority and income quintile simultaneously. The charts differ in the variable illus
trated , however. Charts 4-9 through 4-12 present the number of loans per 100 residential

structures across minority and income quintiles. This measure corrects for differences in
the number of residential structures across census tracts, yielding a measure which allows
comparison of loan flows to different areas . The height of the columns appearing on the
grid in each chart represent the number of loans per 100 buildings. Areas receiving large
numbers of loans will be disproportionately higher than areas receiving low numbers of

' For instance , there are census tracts that are both in the highest two Latino minority
quintiles and in the highest income quintile. Correspondingly there are only 25 census
tracts whichare in both thelowest income quintile and lowest two Latinominority quintiles.
Of these25 tracts, most have a high African American minority, and there are only 4 lowest

income/lowest: minority census tracts in the City's entire 741 census tracts. The problem
ofexcessive polarization prevails primarily for African- American and Latino minorities.
Asian Pacific minorities havea qualitatively differentdistribution, high Asian Pacific tracts
comprising a relatively small percentage of lowest income quintile tracts ( 16% , or 24 of

149) , but otherwise uniformly dispersed throughout the other four income quintiles.
89

292

Chart 4-9
Loan Flows by Lacome and Overall Minority Quladles

250.00
200.00
150.00

# of Loans
100.00

per 100 Res.

50.00

4: Highest Income
0.00

Lowest Minority:0

2
2
3

Highest Minority, 4

0. Lowest Locome

Locome
Minority

Quintile

293

loans. For example, in Chart 4-9 the lower -most square on the grid represents the num
ber of loans per 100 residential structures that went to census tracts in both the highest
minority quintile and the lowest income quintile.'

across the middle three income quintiles corresponding to numbers 1, 2, and 3 on the

minority quintile axis in the chart. There are larger numbers of census tracts in each
of these areas , avoiding the problem encountered in the lowest income quintiles. After
controlling for income levels, loan flows are lower in both number and dollar terms per
100 residential structures, the higher the minority population percentage in a census tract.
Within any of the three middle income quintiles, the lowest minority census tracts receive
at least twice the number of loans per building and almost five times the dollar amounts
per building as are received by the highest minority tracts. This result is consistent across
census tracts in each of the middle three income quintiles.

*It is no longer the case that each row of the grid will sum to roughly 20 % along either the
minority or income quintile exes , since loan flows rather than percentages are presented in
these charts.

• There is evidencethat the Assessor Parcel information derived from the LUPAMS tape
provided by the Planning Department is seriously out-of-date in this fast-growing area,
suggesting the loans per building figures even higher than reflected.

* For example, for the third income quintile, the loansper building in the first minority
quintile are 45 per 100 residential buildings, and in the fifth minority quintile, 24.6 ; a ratio
of 1.83 to 1. In dollar terms, the discrepancy is again greater. Dollars per building are
90

294

The HMDA data presents overwhelming evidence of racially disparate lending in vr .
eas by residential lenders as a whole. Whether this avoidance of high minority areas is
conscious, or the result of unintentionally discriminatory practices, the results are un
ambiguous. The higher the overall minority population, the lower the level of residential
lending flows per building going to that tract even after holding income levels constant.
The phenomenon of lower loan flows is particularly acute for census tracts in the highest
minority quintile. If the pattern of lower lour flows to higher minority areas is the result
of:

then the pattern is the result of a violation of both the Fair Housing Act and the Equal
Credit Opportunity Act.
2. Lending Flows and African American Quintiles

The disturbing conclusion that race dominates income in residential lending also holds
for certain minority sub -categories. Lending flows by African American quintiles , holding
income quintiles fixed, are presented in Chart 4.10 and Table 4.10. The precipitous decline
in loan flows to census tract with high African American populations discussed previously
is evident here. We have already noted that the African American population in the first
three African American quintiles is small ( less than 5% ) , because many African Americans
live in the fifth African American quintile. The eighty percent of the census tracts in
the City with African American populations of less than 29 % have qualitatively different
experience with lending flows than do the twenty percent of the census tracts in which the
$ 91,395 in the first minority quintile , but only $ 17,977 in the fifth , a ratio of $ 5.08 to $ 1 .
91

295

Chart 4-10
Loun Flows by Lacome and Black Minority Quintiles

60.00
50.00
40.00

* of Loans

per 100 Res.

30.00

20.00

10.00

Lowest Minority:00

4: Highest Income
3
2
3

Highest Minority: 4
Black Minority

0. Lowest Income
Locome
Quintile

296

African American population is higher than 30 % .

Loan flows are uniformly lower, in both aumber and dollars per building, from census
tracts in the fourth African American minority quintile to tracts in the fifth ( highest

concentration) African American quintile. Lending flows fall from 37 loans per building
in the fourth African American quintile to 26 per building in the fifth African American

quintile. The drop in dollars per building is even larger, from $ 71,000 per building in the
fourth quintile to $ 24,000 per building in the fifth.
A second interesting pattern reflects the relationship between African American and

Latino populations in each income quintile. In the second income quintile, the numbers
and dollar volumes of loans are notably lower in the first and fifth African American quin

tiles. This " tent-shape" pattern in lending over African American minority census tracts is
repeated in the third and fourth income quintile. Both numbers and dollars per building in
these income quintiles for the fifth African American quintile are notably lower than those
for the preceding African American quintiles while the data for the first ( lowest proportion

African American ) quintile are next lowest. In the fifth income quintile, as above, no pat
tern is evident. Recall that in the three lowest income quintiles there are considerably more
high overall minority census tracts than low minority tracts. In particular, low African

American minority in these income quintiles implies a high Latino minority population .
Hence the " rent- shape " -low loan flows to the lowest and highest African American quin
tiles, but higher flows in the middle three African American quintiles — is a consequence of
loan flows reacting to higher overall minority populations. Loan flow dips in the highest
African American quintile in response to a high African American population; it dips in
the lowest African American quintile in response to a high Latino population . This pattern

disappears in the highest income quintile simply because there are no high Latino census

tracts in this quintile. This is further evidence for 'racial tipping' based on overall minority
populations rather than any of the individual racial sub-groups. As the composition of
the City has changed, residential lending has shifted away of from central areas of the
92

297

City that are primarily occupied by minorities towards the San Fernando Valley with its
predominantly low minority areas.

The results presented in Chart 4.11 for loan flows over Latino minority populations
holding income constant are roughly similar to the results for the African American mi
nority population. The " tent-shape " in loan flows arises once again in the middle three
income quintiles. There are low levels of loans to both highest and lowest Latino census
tracts, with higher levels to census tracts in the middle three Latino minority quintiles.

This is a consequence of the segregation of minorities in Los Angeles. In the lowest three
income quintiles both highest and lowest Latino tracts have overall high minority popula

tions ( Latino in the highest and African American in the lowest) . Consequently loan flows
are lower in these tracts even after controlling for income, reflecting the results for overall
minority population .

298

Chant 4-11

Loan Flows by Lacome and Hispanic Minority Quintiles

-60.00
50.00

-40.00
30.00

Lowest Minority: 0

Y
3

Highest Minority: 4
Hispanic Minority

0 : Lowest Income

299

4. Lending Flows and Asian Pacific Quintiles

Loan flows by income quintile across the Asian Pacific population is presented in
Chart 4.12. There is little or no evidence that lending flows acrow Asian Pacific quintiles
are uneven after holding income quintiles constant. This conclusion is consistent for all
five income quintiles. Loan flows do not decline systematically across census tracts with
different levels of Asian Pacific populations in any of the five income quintiles. Indeed,
in two cases the levels of flows increase as the Asian Pacific quintiles increase . These
data do not suggest that individual Asian Pacific loan applicants face no discrimination in
residential credit markets; as discussed above, the analysis here pertains only to aggregate
geographic patterns. The data shown in Table 12 disclose no explanation for the systematic

difference between Asian Pacific quintiles and the quintiles for African Americans and

Latinos. One can conjecture that there are a sufficient number of Asian Pacific -controlled
financial institutions so that avoidance by other lenders does not impact systematically on
the Asian Pacific community. Another factor may be the smaller number of Asian Pacific
residents, as a share of the population in Los Angeles. In any event, racially disparate
lending is not as important a problem for the Asian Pacific community, when measured by
the quintuie method, as it is for the African American and Latino communities.
F. RACLAL DISPARITIES IN LOS ANGELES :

We have used a " quintile " approach to income and race to show that racially disparate
lending - defined here as significantly lower loan flows into high-minority areas than low
minority areas, holding area income approximately constant - occurs in Los Angeles. The

basic idea of this method is to hold income approximately constant at the census tract
level, and then to compare the number and dollar volume of residential loans ( detrended by
the number of residential buildings) . No method of determining racial biases in lending can
94

300

Chart 4-12

Loans Flows by lacome and Asian Minority Quintiles

60.00

50.00
40.00

# of Loans

4: Highest Income

2
3

Highest Minority: 4
Asiao Minority

O : Lowest Income
Locome
Quindle

301

claim to be definitive; but the ' quintile " method developed for this study is analytically
neutral, and hence findings based upon it are very suggestive.

of the study.

11 The title transfer data examined in Chapter 6 to this chapter come to the same conclu
sions as are drawn here for HMDA data.
95

302

Chapter 5
Ranking Residential Lending Performance
A. INTRODUCTION
The two previous chapters demonstrated that financial institutions in Los Angeles
make substantially lower levels of residential loans to low income areas and to high minority
areas . The HMDA data also strongly muggest that residential lenders practice racially
disparate lending, that is, they make significantly fewer loans to high minority areas even

when income differentials are adjusted for.

This chapter establishes a ranking scheme that identifies individual financial institu
tions whose residential lending behavior deviates from median industry behavior in both
desirable and undesirable ways. It evaluates individual lenders by comparing them to the

behavior of the median residential lender. Median lender behavior represents the average
lending profile for comparable areas of the City. It is used here as a benchmark even though
previous chapters have suggested that this median behavior is itself deficient in lending to
low income and /or high minority areas of the City. Nonetheless, median lender behavior

provides an objective standard by which all financial institutions can be evaluated, with
each assigned a letter grade. The City can then decide what grade constitutes " satisfac
tory ” lender performance, given the clearly unsatisfactory behavior of the industry as a
whole .

This ranking scheme, or an alternative scheme incorporating more dimensions of lender

behavior ( discussed in Chapter 10 below ) , can be used by the City to channel its deposits
toward financial institutions that achieve high grades for lending areas of particular im

portance. This type of program is called a Linked Deposit Program ( LDP) . The reasoning
behind a linked deposit program is straightforward: preference is given in allocating pub
lic deposits to lenders whose activities are more heavily concentrated in specific areas and
activities deemed to be publicly valuable. A linked deposit program bas two effects: it
96

303

directly channels funds to the lenders most likely to use them in desirable ways; and it
establishes incentives for other lenders to shift towards more socially desirable activities,
so as to qualify in the future for benefits under the program .

RECOMMENDATION : The City of Los Angeles should
implement a linked deposit program .

Studies of other successful and not so successful - linked deposit programs have found
that the essential features of a successful LDP are:

These criteria have been summed up by Campen ( 1985) as the 'well, well, well” criteria;
well - designed , well- implemented, and well- publicized.

Three approaches to ranking residential lenders might be used to evaluate lenders
under a Linked Deposit Program ; ( 1) the " Exhaustive " method implicit in the City's
original RFP, ( 2) the " Specific Area " method focusing on well-defined neighborhoods,
and ( 3) the “ Cohort" method, which we propose as a more attractive alternative. The

*The points made here are drawn from James Campen, A Linked Deposit System for
Massachusetts: Information, Issues and Analysis, a report to the Special Commission on

State Lovestment,State of Massachusetts, June 30, 1977; James Campen, Private Banks
and Public Money : An Analysis of the Design and implementation of the Massachusetts
Linked Deposit Program , John W. McCormack Institute of Public Affairs, University of
Massachusetts at Boston, Winter 1985; and Robert Stumberg , Banking on the States: The
Next Generation ofReinvestment Stondards, Draft Report, Georgetown University Law
Center, May 15, 1990. In particular, see Campen ( 1985) for an in -depth analysis of the
requirements of a successful LDP.
97

304

another on the basis of their loan performance in a variety of different areas within the

City. The difficulty with this approach is that all lending institutions are not directly
comparable due to the vast difference in their scale of lending and the geographic scope of
their lending activities . The importance of maintaining comparability has been recognized
by other jurisdictions in implementing their Linked Deposit Programs.
The Specific Ares approach is useful for targeting well-defined geographic areas. For
the City, these might include areas defined by the Community Redevelopment Agency

( CRA ) and the Community Development Department ( CDD) as being ofparticular interest
or importance. The problem with this approach is its focus on DarTow areas constituting
a small part of the entire City. While useful for particular departments in assessing their

The Cohort approach we propose overcomes the difficulties of scale and scope inherent
in the Exhaustive approach, by grouping lenders based on the absolute scale of their lending
activities. It also overcomes the problem of too narrow a focus associated with the Specific
Area method by looking at areas across the entire City. Rankings are developed within

comparable groups of financial institutions ( cohorts) , on the basis of cohort institutions'
residential loans to targeted areas across the entire City. The Cohort approach is similar to
that used by the State of Massachusetts in its Linked Deposit Program . Another strength
of this ranking approach is that it creates a relatively fair means of comparing institutions
that are in different lender groupings.
The next section reviews the highlights of our " Cohort " ranking procedure. We then

explain this " cohort" ranking procedure in depth , and discuss the general results of an
illustrative " Cobort " ranking procedure. This illustrative ranking uses 1989 HMDA data

to assess the relative performance of 220 residential lenders reporting EIMDA loan activity
in Los Angeles in that calendar year. The final section examines the need to establish a
program of ongoing monitoring. The chapter has four Appendices: the first contains a
98

305

complete set of " Cohort " rankings for 220 residential lenders reporting HIMDA loans made
in Los Angeles during 1989; the next contains a complete set of " Exhaustive rankings for

the period 1981-88, as proposed in the City's original Request For Proposals, and discusses
the weaknesses of this ranking approach; this is followed by an appendix containing a
complete set of " Specific Area rankings for 1989 over Community Redevelopment Agency

and Community Development Department areas within the City. The final appendix
presents the operating manual for the PC program developed by this study to monitor
residential lending performance of financial institutions within the City of Los Angeles.

•

B. HIGHLIGHTS OF THE ANALYSIS
Lender Cohorts in Los Angeles: The 220 financial institutions reporting

306

• Largest, Large and Medium Lender Cobort Behavior:

( 1) Financial institutions in these cohorts, though differing in size, ahibit roughly

C. THE COHORT RANKING METHOD

Three approaches to ranking residential lenders might be used to evaluate lenders'
performance in making residential loans: the " Exhaustive" method implicit in the City's
original RFP; the " Specific Area " method focusing on well-defined neighborhoods; and

the " Cobort" method, which we propose us a more attractive alternative. The Exhaustive
approach ranks all financial institutions reporting HMDA loans against one another on the
basis of their loan performance to a variety of different areas within the City. The difficulty

with this approach is that all lending institutions are not directly comparable due to the
vast difference in their scale of lending and the geographic scope of their lending activities. "

A more detailed description of this method and its results are contained in Appendix 3 - B .
The Specific Area approach is useful for targeting well- defined geographic areas,

as those defined by the Community Redevelopment Agency ( CRA) and the Community De
velopment Department ( CDD ) . Appendix 3 - C presents the results of this straightforward
ranking procedure for each of the Community Redevelopment Agency and Community
Development Department areas within the City.
Home Savings is by far the largest lender with 6,303 loans during 1989 amounting to
$ 1,135,752,000 covering all census tracts in Los Angeles, while many small lenders, such

as First Professional Bank, made only one loan to one census tract.
100

307

The Cohort approach we propose overcomes the difficulties of scale and scope inherent
in the Exhaustive approach , by grouping lenders based on the absolute scale of their lending

activities. Rankings are developed within each of these groupings ( cohorts ) to rank lenders
on the basis of their residential loans to targeted areas within the City. The Cohort
approach is similar to that wed by the State of Massachusetts in its Linked Deposit
Program . A strength of this ranking approach is that it allows comparisons across lender
groupings. One can ask, for instance, ifthe largest institutions have better or worse records

than small institutions in lending to the targeted areas us a percentage of loan activities .

1989. The lender cohorts are defined as :
• Largest Lenders: These are lenders whose ELMDA loans during 1989 in the City

While these four groups ( Largest, Large, Medium and Small) might appear to be ad hoc

groupings, there is considerable justification in the data for such a categorization. There
are clearly marked " breaks ” in the loan volumes between each of the cohorts, suggesting
considerable differences exist in the scale of lending between any two cohorts. This grouping

approach is similar in spirit, though quite different in implementation, from the ranking
method used by the City of Boston in its linked deposit program . The breakdowns in
' The Massachusetts and Boston LDPs are discussed in depth in Chapter 10.
101

308

the Boston program focused on overall asset size rather than

lender's presence in a

particular line of lending activity. As the cohort method focuses on the dollar volume of
residential lending, rather than a fostitution's asset size, we find that some large national

banks, such First Interstate, fall into the second largest category of lenders, rather than
the largest category .

1. Areas over which Rankings are Determined
Each cohort of lenders is ranked separately from other cohorts to avoid comparability
problems arising from differences in lender Kale. While lenders in each cohort are ranked
separately, the areas over which their lending performance is ranked is the same across

cohorts. We have targeted six ( 6) types of areas which appear consistent with the City's
interests in this project's RFP. A lender's performance is ranked against its peers in each

of these six areas on the basis of both numbers and dollars of loans in the area expressed
as a percentage of the lender's total loans over the entire City. The following six areas are
used in ranking lender performance:

( 1) Al Block Grant Eligible Census Tracts: Block grant eligibility encompasses

102

309

The last four areas are the most restrictive, each constituting only 20 % of the total census
tracts in the City, while the first two areas cover at least 40% the census tracts in the

City. The makeup of Community Redevelopment Agency and Community Development
Department areas of interest are primarily low -moderate income and high minority census

tracts and so these areas are contained in the first two ranking areas. Hence lender rankings
over the first two areas are comparable to rankings that would be established by generating
a ranking based on lending to Community Redevelopment Agency and /or Community
Development Department areas within the City. Residential lending to the lowest 20 %
income census tracts in the City is an important indicator of lending for affordable housing.
The majority of the loans to these census tracts are for multi-family rental dwellings, a
particularly crucial need within the City .

It is important to understand that the list of census tracts for each type of area is
independent of the other areas. While it is certainly possible that a census tract may fall
into all six areas, there is no necessary reason why this should be so. The extent to which
census tracts appear in multiple areas simply reflects the demographic make-up of the City.
The close association between lower median incomes and higher minority populations in
Los Angeles , as Chapter 4 demonstrated, makes it likely that many census tracts fall into

both of the first two areas and at least one of the last four areas.

2. Ranking Method
The residential lending ranking method is set forth in the following manner. Within
a particular lender cobort, and for a particular area type, the lenders are divided into

four ( 4) approcimately equal groups based on the percent of their total loans which went
to census tracts in that area type.' There is one score based on the percent of the total
The ELMDA data used in this study do not differentiate between types of multi-family

residential structures. Hence it is not known whether these represent affordable renta]
housing, ordinary rental housing, or even condominiums.

• The number of lenders in each of these quartiles will depend on the overall cohort size.

For the Largest lenders, a quartile will contain 6 ( =24/4) lenders, while for the smallest
103

310

number of loans and another score based on the percent of total dollars of loans going to
the area of interest.
S

As institution's can range from 0 to 3 in each of the six areas . The higher the score, the
" better " the institution's performance, u 2 percent of its total loans, in lending to the
area of interest.

originally proposed exhaustive ranking method .
There are two rank scores for each lender , one for numbers and one for dollars, for

each of the six areas of interest. Given that individual area scores lie between 0 and 3, two
aggregate scores, one for numbers the other for dollars, can be arrived at by sumoming over

scores in each of the six areas. The maximum score possible for a lender over all six areas is
18.0, while the minimum is 0. The higher a lender's score on this scale, the " better " is the
institution's overall performance as compared to other institutions in its lender cohort. In

fact, the aggregate scores can be used to arrive at an overall grade for residential lending
performance. Number scores are converted to letter grades using the following scheme:

cohort, the same quartile will contain roughly 21 ( =83/4) lenders.
104

311

• A = Average score greater than 13.5 and less than or equal to 18.0.

• B = Average score greater than I and less than or equal to 13.5.
• C = Average score greater than 4.5 and less than or equal to 9.0.
• D = Average score greater than 0.0 and less than or equal to 4.5.
• F = Average score equal to 0.0 .

It is quite difficult to fail under this scheme. Only financial institutions which make
fewer loans than 75 % of their cohort, in both numbers and dollars, in all six areas of
interest receive an 'F'.

The aggregate scoring scheme is simple to implement. Each institution is evaluated

on its residential loans to six areas, receiving two scores in each area , one for numbers
of loans and the other for dollars of loans. The scores over the six areas are then added
together and averaged to produce an aggregate score out of 18. This numerical score is
converted to a letter grade of 'A' through 'F' as explained above. This ranking scheme

has the additional attraction of being easily integrated into a more comprehensive scoring
system covering performance in the areas of financial services and lending for economic
development. This comprehensive ranking is derived in Chapter 10.

3. Cohort Ranking Results
To avoid nonsensical results the standards of one cohort should not be imposed on
those of another cohort. To the extent that performance within an entire cohort is deemed
unsatisfactory, different policies or programs may be necessary to address particular defi
ciencies in the cohort's overall lending behavior. For example, a small institution's poor
lending performance to low income, high minority areas may reflect its location in a high

income, low minority area . The appropriate program might be participation in a loon pool
aimed at priority areas, or encouraging it to purchase loans originated by other lenders in
these areas , rather than requiring the lender to originate such loans on its own .
It is also important to remember the backdrop against which the cohort ranking
scheme is being conducted. Chapters 3 and 4 established that regardless of how well
105

312

& lender might rank against its cohorts, lenders in aggregate systematically make fewer
residential loans in low income and high minority areas. It is also clear that racially
disparate lending exists within the City : a substantial portion of these lower loan filows
occur when, for any given income level, a significant proportion of residents are African

American and /or Latino.
Detailed rankings for each of the institutions making residential loans in the City are

in the tables of Appendix 8 - A . Table 5.1, also reproduced in the body of this chapter,
compares best and worst lender lending patterns by cohort to establish qualitative differ

ences in lending performance by size. Tables 5.2 through 5.5 explore differences in lending
patterns within each of the four cohorts. Tables 5.6-5.9 present the actual ranking results
for the 220 residential lenders by the four lender cohorts. These tables contain information

on; ( 1) each lender's rankings within each type of area, ( 2) aggregate scores across all area
types, and ( 3) the " qualitative " ranking received. These tables also include the percentage
of loans, either numbers of dollars, which each lender made to each type of area. These
percentages appear directly below the corresponding rank scores in each of the Tables

3.6-5.9. Table 5.10 is of particular interest; it provides a summary of institution rankings
by alphabetical order for the 220 institutions in the 1989 HMDA data.
Chart 5-1 illustrates that institutions in the Lorgest cohort made 81 %
residential loans in 1989 and 83%

of the 50,000

of the almost $ 11 billion in loans. A further 15%

the market, both in numbers and dollars, is attributable to institutions in the Large
cohort. The remaining 156 institutions in the Medium and Small cohorts have a small,

almost insignificant, market share. The high market share of the Largest and Large
cohorts suggest their record of lending closely follows the aggregate patterns discussed in
Chapters 3 and 4. It is interesting therefore to compare this cobort's overall performance
with those of smaller ' community-based" lenders. Conventional wisdom suggests smaller
financial institutions are invaluable to “ special needs" communities because they provide

services that larger institutions do not. This argument has recently been revived to oppose
106

313

Chart 3-1
Market Share by Lender Cobort

81.16%

0.70%

Largest

3.68%

Large

14.46%

Il Medium Small

Market Share by Lender Cohort
S of 1989 HMDA Loans
83.23 %

0.26% 2.71 %

Largest

60-893 O - 92 - ll
-

Large

13.80 %

II Medium Small

314

TABLE 5.1 :
Best and Worst Lender Profiles

Bloed

Local

Liebest

LARGEST COHORT

Ligbert | Eliebest
20 %

anas

Fichest
20 %
Asian

Bahle

18)

BEST LENDER
13
14

36
38

18
12

| * . % of Total #
$ - $ d Total 3's

0
0

2
1

1
j

* - % of Total to

$ - $ of Total 's

21
26

66
56

AIA

.47
54

#usal Total ho

Ius al Total s'o

0
0

| + n % ofTotal to

75
63

- 12

63
SS

0
0

14
20

6
13

17
22

5
3

sudTotel 3's

3
21

in % of Total mio
I w % of Total $ 's

0
0

% a Totel H'S

23
33

WORST LENDER
2

5
S

LARGE COHORT
ไม้สี

BEST LENDER

17
13

olo

WORST LENDER

MEDIUM COHORT
BEST LENDER
25

38
27

WORST LENDER
Total po
%
sw $ d Total I's

Pultiple Leader )

SMALL COHORT

BEST LENDER
|

WORST LENDER
Multiple Leader )

m % d Total to

315

proposed changes in banking legislation that would encourage mergers and consolidation

within the industry .

The main results of the cohort ranking are somewhat surprising in this light. Chart
8-2 illustrates that relative to the cohorts of larger lenders, lenders in the Small cohort
perform more poorly, on average. Almost 35 % of the Small cohort received a failing

grade of 'F' on their 1989 HMDA lending. Most institutions in this category made very
few loans in the City during 1989. It is therefore difficult to judge if the low scores reflect

a systematic avoidance of low income and high minority areas by these institutions or a
scale so small that they have no substantial lending presence in the areas of interest.

Chart 5-2 illustrates that the grade distribution within each of the Largest, Large,
and Medium cohorts is approximately the same. Within each of these three cohorts,
8%

of the institutions received a failing grade on their 1989 residential lending activities .

Between 20 % and 25% of the institutions in each of the three cohorts received the highest
grade on their lending, while 45% of the institutions received a 'B' or higher in each cohort.
While this suggests that performance is relatively the same across the three cohorts,
even the best of the Largest lenders are never as heavily concentrated in the areas of

interest as are the best lenders in the Large cohort. Similarly, the best lenders in the
Large cohort are not as heavily concentrated in the six areas as are the best lenders in the
Medium cohort. Table 5.1 illustrates this pattern , presenting the lending percentages of

the best and worst lenders within each cohort to each of the six areas used in the ranking.
By contrast, even the worst of the Largest cohort lenders make higher percentages of

their loans to the six areas than do the worst lenders in the Large cobort, who are
correspondingly make higher percentages of their loans in the six areas than do the worst

institutions in the Medium cohort. The dispersion of lender performance within a cohort
increases as the size of institutions within a cohort decreases.
This dispersion is also illustrated by the results in Tables 5.2 through 5.5 in Appendix

5 - A . In the Largest cohort, lenders receiving the best ranking ( =3) over Block Grant
107

316

Court 3-2:
Grades for Residendal Leading

'Large Lender Cobort

Largest Leader Cobort
12.50 %

& 33 %

15.00 %

1.50 %
20.00 %

23.00 %

33.33%

32.50 %

25.00 %

20.83 %

OA NB I c m D IF

DA NB IC ID IF

'Medium' Lender Cobort

'Small’ Lender Cobort

685 %

34.94%

21.92 %

9.64 %

21.92 %
16.87%

26.03%

23.29 %

DAB

C 11 D IF

20.48%

OA SB

18.07 %

D IF

317

Eligible census tracts in the City made 32-51 % of their total loans by number to these
areas. In contrast, to receive the highest ranking over Block Grant Eligible census tracts in
the Large lender cobort, an institution wus required to 34-84 % of its loans to these census
tracts. The requirement for the highest score over these census tracts in the Medium
cohort rises to 35-100 % of total loans. On the other hand, in the Largest cobort, even

lenders receiving the lowest ranking ( = 0) made 2-15 % of their total loans by sumber to
the Block Grant Eligible census tracts , while the worst lenders in the Large lender cohort
made 0-12 % of their loans to these census tracts. This increasing dispersion in lender

behavior exists between any lender cohort and any other smaller lendes cohort.
The results are consistent with increasing lender specialization, both in área and type
of loan , as the size of the institution decreases. Smaller institutions secure market niches
while only the largest institutions have the ability to be all things to all customers in all
places. This specialization on the part of smaller lenders may be good or bad for their

ranking; it is good when they specialize in the targeted areas ( low income, high minority ) ,
but it is bad when they specialize outside the targeted areas ( high income, low minority ) .
Specialization makes it correspondingly more difficult in percentage terms for a lender in

the Small cohort versus a lender in the Largest cohort to receive the best ranking. The
ranking scheme, however, also makes it correspondingly more difficult in percentage terms

for a lender in the Largest cohort versus one in the smallest cobort to avoid receiving
the worst ranking.

4. Other Features of the Ranking Scheme
A lender's score for number of loans should , in general, be fairly close to the corte
sponding score for dollars of loans made to an area . This is true whenever the majority of
residential loans made by the lender are of the conventional single- family purchase type.

Table 5.6 for the Largest cohort in Appendix 3 - A reveals the congruence between the

two scores for the majority of lenders. There are circumstances, however, under which
108

318

the two segregate scores differ substantially. If a leader is primarily involved in home
improvement loans, each loan involves a lower dollar amount than a conventional loan . As

minority areas in the form of home improvement loans.

The two aggregate rankings can also diverge when the lender primarily makes multi
family loans. A multi- family building loan is normally for a higher amount than a conven

tional single-family loan, a given number of multi-family loans entails a far greater dollar
amount than the same number of conventional loans. As a result, lenders specializing in
multi-family loans normally have number rankings that are lower than their corresponding

dollar ranks. In Table 5.6 for the Largest cobort, Éidelity FS & LA is a lender whose loans
to thetargeted areas consist primarily of loans for multi-family buildings . Fidelity's score
of 10 for numbers versus 15 for dollars indicates that each loan it made in these areas was
for a larger than average dollar amount, indicating loans for multi- family dwellings. Hence

the two scores together contain information not only on relative performance within the
cohort, but also information on the composition of each lender's loans. It may be desirable
for the City to use both in evaluating a lender, perhaps rewarding lenders meeting the
Deed for multi-family rental housing, especially in the area of affordable rental housing.

D. ON -GOING MONITORING
An ongoing reporting and monitoring effort is necessary for the ranking procedure to
be effective in encouraging changes in lender behavior. It ensures that incentives exist for

financial institutions to improve their performance in areas desirable to the City. Effective

• The construction loan financing on entire condominium building should be reported as
a multifamily loan under HMDA . Underthese regulations, loans for the purchase of each
condominium unit would appear under the single- family, conventional loan category.
109

319

ongoing monitoring has been identified by Campen ( 1985) as an important component of

a well-implemented program . Marginal lenders have an incentive to improve performance
if they feel that their gains will be recognized . Similarly, " good" lenders have the incen

tive to maintain or improve their performance. An effective ongoing monitoring program
facilitates the publicity aspects of the program , allowing for regular updates on financial

institution behavior. This publicity component is an important part of a successful pro
gram . For instance, Campen ( 1990) notes that one-third of all survey respondents said
they would switch their primary financial institution if they became aware that its prac

tices failed to comply with CRA. Among non -white respondents this proportion increased
to one-half.'

RECOMMENDATION : The City should monitor lender
performance closely, and update and publicize its rankings
regularly.

An integral component of this study is the software package, Lending Assessor, which
allows the analysis of HMDA residential lending for individual lenders in specific areas of

the City by year over the period 1981-88. The program runs on a personal computer and
utilizes Microsoft Windows and Microsoft Excel to provide a user-friendly environment for

analyzing residential lenders. Appendix 3 - D contains the User Manual for the Lending
Assessos program . Users are able to produce a wide variety of customized reports, in
particular, comparisons of the lending performance of an individual financial institution
relative to other institutions or groups of institutions in different areas throughout the
City. The main problem with this software is inherent in the HIMDA data, rather than the

program itself. HMDA data for a given year becomes available with a lag of one year. The
Lending Assessor program contains a facility for updating the HMDA information each
Campen, The Political Economy of Linked Deposit Banking Programs, ( 1990 ) , U. Mas
sachusetts at Boston .
110

320

year .

reinvestment programs.

* CD -ROM stands for Compact Disk -Read Only Memory. It is a technology which stores

information on a compact disk . Unfortunately this service did not become available until
the study had concluded this portion of the analysis.
This data is not without flaws, however. A comparison of the 1989 title transfer data with
the corresponding 1989 HMDA datafor the City indicates substantial. under-reporting
of residential loans in the title transfer data . The title transfer data does not include
information on home improvement loans, por does it distinguish between conventional and
government-backed single family mortgages.
111

321

Chapter 6

Race, residential lending, and income
Individual Transactions data for 1989

A. INTRODUCTION

This chapter explores an additional source of 1989 residential lending data that sup

plements the 1981-89 HMDA data examined in previous chapters. The new source, title
transfer data , permits an analysis of lending flows by income and racial concentration
that complements the HIMDA - based analysis . It provides information on each individual

housing sales in the City of Los Angeles during 1989, as reported to the County Assessor's
Office for the purpose of transfering title to the new owner . In effect, these data represent

information from an independent source, which can be used to check conclusions drawn
from HMDA data. Title transfer data, obtained from a private vendor, provide details of
residential mortgage terms not provided by FIMDA data. Each title transfer transaction

provides information on a property's sale price and assessed value, the amount of the loan
financing the sale, the lender's name, the interest rate on and the length of the mortgage.
We first examine the aggregate numbers and dollars of sales and loans from the 1989
title transfer data by types of areas and financial institutions. The azzlysis is identical to
that conducted in Chapter 3, but is of special interest because it contains information on a
larger number of small mortgage and finance companies than does the HMDA data. The
title transfer data is less comprehensive than the HMDA data, however, as it contains only
home sales, not home improvement loans. It also does not distinguish between loans made

under FHA or VA provisions versus conventional loans . We examine the characteristics of

median home sale transactions,such as interest rate and downpayment, across demographic
areas and financial institutions within the City. We investigate whether variations in loan

terms mirror the patterns of aggregate loan flows across areas. For example, we examine
112

322

whether the median downpayment percent or interest rate on residential loans were higher
in low - income census tracts than in high - income census tracts.
B. TITLE TRANSFER DATA

Title transfer information was obtained on properties changing ownership in Los An
geles County during the year 1989. The data were obtained from DAMAR Real Estate

Information Service, a private vendor collecting title transfer data on Los Angeles County
from a variety of sources . Information on the selling and financing arrangements for a

parcel of real estate is updated whenever the parcel changes owners . This information is
submitted to DAMAR by title companies and financial institutions. We collected data on

sales and financing of single family residential transactions comparable to those covered by
HMDA disclosure during 1989 in the City of Los Angeles . Our main interest was in the
financing portion of any sale, so we collected information only for those transactions that
filled the following criteria : ( 1) were verified by DAMAR, ( 2) contained full information

on the property's sale characteristics, and ( 3) contained the amount of loan and identified
the lending institution financing the sale .

The resulting title transfer data set for 1989 contained 11,256 transactions within
the City. This compared to the 44,841 single family residential loans ( both government
backed and conventional) reported under HIMDA for 1989. While HMDA single family
mortgages include a limited number of refinancings, it is unlikely that refinancings explain
the difference between the two loan figures. The requirement for verified information on a
property's financing arrangements is the most likely reason for the lower number of loans in

the 1989 title transfer data. In the analysis which follows it is important to remember that
the title transfer data is a subset, albeit a sizable one, of all 1989 single family residential
Specifically, we obtained data in five DAMAR categories: single family ( code 163 or SFR) ;

duplex ( code 115 or TFD) ; triplex ( 165) ; quadraplex ( 151) ; or multifamily ( 133 or MFD ) :
These categories encompass residential properties with , respectively, 1, 2, 3 ,4, and 2-4
units.
113

323

loans in the City. Hence results in the sections which follow are inferences based on a large
sample of total residential lending.
The title transfer data contains a wealth of detail about each transaction , including

information on the property's location - including census tract and characteristics of the
sale and financing of the property. To make DAMAR's transactions-level data comparable
to the HMDA data, we matched each transaction to a census tract . Once the title transfer
data was grouped into census tracts , it was combined with the demographic data used in
the analyses of Chapters 3 and 4.

A total of 185 institutions appear in the 1989 title transfer data, compared to 220
institutions in the corresponding 1989 HMDA data. The 185 title transfer institutions are

not completely contained in the 1989 HMDA institutions lenders. Many of the title transfer
lenders are small mortgage and finance companies that do not appear in the HMDA data.

These small mortgage and finance companies are assigned to an " exempt" category, as they

wherever possible.

C. AGGREGATE TITLE TRANSFER LENDING PATTERNS
Table 6.1 in the Appendix to this chapter presents aggregate number and dollar figures
for the title transfer data broken out by types of areas and financial institutions. This

information is comparable to that presented in Tables 3.1 through 3.3 of Chapter 3. Table
6.1 also provides information on average sale price, loan dollar amount, and downpayment
DAMAR transaction records contain a category for census tract. However, this is rarely

filled in; less than20 % of the transactions haveacensus tract number. We compensated
by using the City Planning Department's LUPAMS data set to match the Assessor Parcel
Numbers( APNs) that invariably accompany DAMAR data with census tracts. This pro
cedure allowed us to match the APN's of 11,256 transactions to census tracts. However,

some 162 transactions had no matching census tract. These transactions were deleted from

the analysis after determining that no one financial institution was overly involvedin these
transactions.
114

324

by area and type of institution . Averages are likely to be misleading in this data, however,
because there is a wide dispersion in transaction characteristics. The preferred measures
for inferring transaction characteristics are median figures rather than average. These
median characteristics by area und type of institution are presented in Tables 6.2 and 6.3.
The majority of residential title transfers in our data - 86 % of the total were for

single family residences. The next largest category was duplexes, which involved 8% of
total transactions. The market share pattern across financial institution types is strikingly

different in the 1989 title transfer data from that in the 1981-89 HMDA data of Chapters
3 and 4. The 1981-89 EMDA data showed savings and loans dominating the residential
lending market with more than a 90 % market share of reported loans. Savings and loans
stil have a significant portion of the title transfer loans ( 31 % ) , but their share is lower
than the 41% total market share for HMDA ' exempt" institutions, which are mostly small
mortgage and finance companies. National banks have a significant 25% share of the market
in 1989, higher than their corresponding share in the 1989 HMDA data. HMDA " exempt"
institutions tend to finance lower median dollar sales , leading to a lower median dollar loan
figure than other financial institutions. They also tend to accept lower downpayments than
other types of institutions, although they do not appear to charge a correspondingly higher
rate of interest.

1. Patterns across Income Quintiles
The title transfer data in Table 6.1 exhibit much the same pattern with respect to
income quintiles as the 1981-89 HMDA data in Chapter 3. Larger numbers of loans are
made in larger dollar amounts to higher income census tracts versus lower income trocts .
There were more than twice ( 2.3 times) as many loans made in the highest income quintile
as in the lowest income quintile in 1989. The average dollar loan in highest income census
tracts was roughly three times as large as that in the lowest income tracts.

The use of title transfer data allows us to examine how the median percent downpay
115

325

ment varies over income quintiles. There is a clear trend for higher percentage downpay
ments in higher income quintile census tracts, 20 % in the lowest income quintile versus

39% in the highest income quintile. In dollar terms, median downpayments were $ 23,168 in
the lowest quintile and $ 133,795 in the highest. This pattern suggests that the higher the

income quintile, the more likely it is that a home sale reflects " trading -up,' where existing
homeowners nse equity on previously acquired homes us the basis for their downpayment
On a more expensive residence. First -time buyers seem to be, on average, confined to lower
income quintiles, as indicated by the lower downpayment percentage in these areas.
2. Patterns across Minority Quintiles

The patterns of residential lending across minority areas within the City are similar
to those found in the HMDA data, but not as clear cut. In general, the higher the minority

population of a census tract, the lower the number and dollars of loans going to that tract.
The highest overall minority quintile census tracts received 8% of loans by dollar volume
and 18% of loans by number. In contrast, the lowest minority population quintile census
tracts received 35% of loan dollar volume and 20% of loan numbers. The lowest 40%
minority census tracts within the City received 46% of the loans by number and 61%

dollar volume in the 1989 title transfer data. This difference, while marked, is certainly
not as great as the differential found in the FIMDA lending data using the same overall
minority quintile rankings.
The patterns of fewer loans and small loon dollar volumes are repeated for both African

American and for Latino population quintiles. There is not as dramatic a decline in loan
volume in the highest quintile census tracts as was present in the HIMDA data. This is
due primarily to the inclusion of mortgage and finance companies it the title transfer data.

Small mortgage and finance companies make a higher percentage of their loans to high
minority areas than low minority areas. While racially disparate lending by conventional
lenders remains a problem, it is mitigated to some extent by the activities of these small
116

326

mortgage and finance companies. A larger decline in dollar volume occur from low mi
nority to high minority areas , but this largely represents the lower average sale price and

consequent lower loan amount in the highest minority census tracts. As already noted,
this occurs because high minority census tracts are overwhelmingly associated with low
income areas within the City.

rates , indicating that loans to these areas are profitable at mortgage terms prevailing in

the market as a whole.

D. MEDIAN TRANSACTION CHARACTERISTICS
Title transfer data allows one to calculate the characteristics of median home sale
transactions. This was not possible for the HMDA data since the lending information for

each institution was aggregated by census tract.

* A limitation of the title transfer data is that there is noway to determine whether or not
a loan was made under a federal program like FHA or VA .
The median value of a set of pumbers is that value which half the numbers are below
and half the numbers are above. It provides a useful measure of the " normal" level of a
variable.
117

327

residences, almost 90 % of the 11,256 transactions. There are, however, an appreciable
number of duplexes , triplexes and quadraplexes. Lending terms are comparable across the
various types of buildings, with downpayments of roughly 20 % , although interest rates are
somewhat lower for single family residences.

Loan terms also vary little across financial institutions. A 20 % downpayment is close
to the median for every type of lender. National banks make slightly higher median

loan amounts than institutions with higher interest rates. The assessed value of these
homes , versus their sale price, is lower than for other institutions, indicating that these

homes are in areas experiencing above average rates of price appreciation . Savings and
loan associations offer median loans with the lowest median interest rates of any other
institution , even though the loan size is close to that of oational banks. The HMDA
" exempt” . category, which includes small mortgage and finance companies, makes loans for
the lowest median sale price and loan amounts, with interest rates comparable to savings
and loans. The median assessed value of these homes to their sale price is roughly equal
to that of savings and loans and higher than that of national banks.
Overall, there is a surprising amount of uniformity in median interest rates and down
payments , across institutions and demographic areas. Underwriting standards appear to

be comparable across lenders because of the practice of selling these mortgages in secondary
markets through the FNMA. There is an indication that lending institutions specialize in
particular market niches, both areas and types of buildings. No type of institution , how

ever , seems to specialize in financing affordable single family housing because median loan
amounts for every type of lender consistently exceed FHA and VA loan limits for the Los
Angeles area .
1. Patterns across Income Quintiles
Median sale price and median loan amount increase uniformly with higher income

quintiles . The median sale price in the highest income quintile is $ 380,000, roughly 3.3
118

328

times the $ 114,250 median sale price in the lowest income quintile. Similarly, the median
loan amount in the highest income quintile is $ 285,000, roughly 2.8 times the $ 100,700
median loan in the lowest income quintile. The lower ratio of loan amounts reflects the

rise in the median percent downpayment in higher income quintiles. Despite the higher
downpayment percentage in higher income quintiles, however, the median interest rate on

loans is similar across all income quintiles.
The ratio of median sale price to median assessed value is lower in higher income
census tracts , indicating that these quintiles have experienced higher rates of home price

appreciation than lower income tracts. This tends to exacerbate wealth inequality, suggest

ing that higher income households in Los Angeles have experienced higher rates of return
from home ownership than have low or moderate income households fortunate enough to
be homeowners .

2. Patterns across Minority Quintiles
Increases in minority populations are associated with lower median sale prices and
loan amounts . There is a particularly sharp dropoff in both sale price and loan amount
when one moves to the highest minority population quintile. This drop in sale price is

associated with a corresponding dramatic decline in the median downpayment percentage
and a noticeable increase in the median interest rate ( from 8.87% to 9.25% ) in the highest
minority quintile. This pattern is found across overall minority quintiles, and acros African

American and Latino population quintiles individually.
It is difficult to interpret the associated drop in downpayment percent and the in
creased interest rates found in the highest minority areas. There are valid economic

arguments to support higher interest rates that compensate lenders for the higher risk

associated with a lower downpayment. On the other hand, there are strong indications
of racially disparate lending in the numbers and dollars of residential lending in both the
HMDA and title transfer data. The higher interest rate may simply be another mani
119

329

festation of this racial disparity in lending. Table 6.3 provides median characteristics of

loan terms and amounts across minority quintiles, holding income approcimately constant.
This table is similar to the analysis conducted on the HIMDA data in Chapter 4. After

controlling for the effects of income, the downpayment percent in the highest minority
quintile is roughly the same as that in lower minority quintiles. The median interest rate
in the highest minority quintile is not significantly higher in high minority areas than that
in th : lower minority quintiles. While certainly not conclusive, this evidence suggests that

economic considerations may be responsible for the higher median interest rates when the
downpayment percentage falls.

120

330

PART III

Financial Services, Lending for

Economic Development and Affordable Housing
INTRODUCTION TO CHAPTERS 7-10

Chapters 3 and 4 examined overall patterns of residential credit iows in Los Angeles
using comprehensive data reflecting lender activity. Chapter 5 suggested ways of evaluat
ing banks' performance in channeling residential credit flows to special needs areas — that
is, to census tracts with low median incomes and /or large proportions of minority resi
dents, Chapter 6 then supplemented this evidence, especially by lending insight into the
role of mortgage companies as residential lenders in the Los Angeles housing market. But

residential credit flows, while crucial for housing markets and neighborhood viability , rep
resert only one of the channels through which banking activities affect the welfare of Los
Angeles' residents. The second part of this study, contained in Chapters 7 through 10,
examine other channels.

Chapter 7 looks critically at banking services in Los Angeles: check cashing policies;

the terms and conditions on selected checking accounts; and the location of bank and
thrift branches. Chapter 8 examines banks' role in economic development by considering
evidence on their provision of credit to businesses, especially small businesses, and to indi
vidual households. Chapter 9 then investigates lenders' roles in financing the development

of affordable housing.

The chapters in Part II use data very different from those underlying Chapters 3

121

331

8. Those chapters relied on a vast data base covering all banking institutions. No such
data base is available for the subject matter of these chapters. Instead , discussion relies
on specialized data sources. Many banks and savings and loan associations ( thrifts) in
the city voluntarily made available two sources of information : their 1989 Community

Reinvestment Act ( CRA ) statements , and their responses to a special mail survey sent to
financial institutions. Numerous developers active in the construction of affordable housing
responded to a mail questionnaire we sent them . Small business owners in the City's five
Enterprise Zones attended meetings and returned a modest number of questionnaires on

their interactions with lending institutions.
So in contrast to Part II's emphasis on the quantitative volume of credit flows, Part III

emphasizes qualitative indicators of financial institution performance: ( 1) the quality and
quantity of banking services; ( 2) the scope of credit available to individuals, businesses,
and developers; and ( 3) borrowers ' and lenders' perceptions about problems in particular
credit areas. Chapter 10 concludes this expanded discussion of bank behavior by suggesting
a broader system for ranking financial institutions' performance than wąs developed in

Chapter 5 above. This broader ranking system would use the types of information discussed
in Chapters 7-9 to rank financial institutions on their responsiveness to the service needs
of depositors and non -depositors, and to the credit Deeds of affordable- housing developers
and small businesses. The broader ranking system suggested here is not implemented here;

putting this system into place for all Los Angeles lenders would require more information
about them than is currently available.

122

332

Chapter 7

Banking markets and banking services

A. INTRODUCTION
This chapter explores the availability of banking services in Los Angeles. Banking
services consist of the entire set of institutional means which community residents can use
to manage their cash transfers and payments from their deposit accounts . The provision of
banking services involves both a qualitative and a geographic dimension. The qualitative

character of these services depends on the types of accounts that financial institutions offer
to potential depositors, on institutional policies regarding access to deposit accounts, and
on institutional policies regarding residents of their market areas with special needs. The

geographic accessibility of an institution's banking services depends on the location of its
branch offices. Of particular interest here is the availability of banking services for lower
income residents. These special services include the provision of " lifeline deposit accounts ,
check cashing by non -depositors, and direct deposit of government benefits checks. " Life

line” accounts are deposit accounts that we not costly to open or maintain . As important
dimension of providing these special services is publicizing their availability.
Since this chapter is the first to present results from lenders ' CRA statements, it

also presents some basic information on financial institutions in Los Angeles: ( 1) the

geographic scope of banking markets in Los Angeles; ( 2) whether lenders assess community
credit Deeds; and ( 3) patterns of institutional specialization. This material- contained in
sections D, E, and F - provides a context for the discussion in the remainder of this chapter
and for those that follow .

The emphasis in this chapter then turns to how well community Deeds for banking
services are being met by financial institutions. Section G examines whether banking
services of equal quality ve equally available throughout Los Angeles, regardless of the
123

333

average income or racial composition of the various areas. Section ! then examines the

type of banking services offered; section I looks into the provision of deposit accounts for
economically vulnerable individuals with special banking needs.

B. HIGHLIGHTS OF THE ANALYSIS

1. The Geographic Scope of Banking Markets
• Los Angeles has awide variety of financial institutions with respect to market

2. Do Lenders Assess Community Credit Needs?

• Some 86% of 107 reporting institutions indicate that they have conducted the

3. Patterns of Institutional Specialization
. Most LA institutions provide credit in several activity areas . Of 175 reporting

124

334

able. In particular, only 8 institutions'CRA statementsmention that they pro
vide credit to first- time home buyen . Only 6 lenders indicate that they provide
predevelopment financing for developers, and 6 provide permanent financing for
multi- family projects. Only 17 institutions provide start-up financing for busi

4. The Geographic Location of Bank Branches and ATMs
• Just over athird of Los Angeles County bank branches and ATMs are located

8. The Provision of Banking Services
• Most institutions( 94 % , according to CRA statements, 85 % according to the bank

125

335

• About 71 % of institutions responding to the bank survey provide written infor

6. Special Needs Deposit Accounts
• Most institutions offer special accounts for elderly low - income depositors. About

C. DESCRIPTION OF DATA
This chapter integrates data from three sources. The first is a data bank summarizing

the 1989 Community Reinvestment Act ( CRA) statements of 175 commercial banks and
savings and loan associations. These statements were solicited from the 233 banks and
savings and loan associations filing them in Los Angeles.

126

336

to a standardized format for analysis .

The second data source consists of responses to our survey of banks and thrifts with
offices in Los Angeles . The survey consisted of sixty questions in five topic areas: market
ares and scope of services; financial data; special credit and affordable housing programs;

lending criteria ; and earthquake preparedness. Most questions used a multiple-choice
format. This survey was pretested with the cooperation of four large banks and two
small ones . The survey was then sent to the 175 institutions which had returned CRA
statements; 66 were completed and returned . Included as appendices to this chapter are a

summary of CRA statement data ( Appendix 7A ) , responses to the questions on the bank
survey ( 7B ) , a list of the institutions that sent in their CRA statements ( 7C) , and a list

of the institutions that returned their bank surveys ( 7D ) , and a copy of the bank survey
( TE ) .

institution cooperation with this study. The higher return obtained on the request for

CRA statements can be explained by the fact that CRA statements are public documents
by law , and were prepared by financial institutions prior to the inception of this study."

127

337

institutions into four groups for analytical purposes, based on their volume of residential
lending. This chapter will also sometimes distinguish among institutions. However, it
will divide institutions into two groups, not four. The two larger cohorts of Chapter 6
institutions that have made residential loans totalling at least $ 50 million in the 1981-88
period - are termed " large lenders ' below , and the two smaller cohorts of Chapter 5 are
termed small lenders . "

D. THE GEOGRAPHIC SCOPE
OF BANKING MARKETS :
Residents and businesses in the City of Los Angeles are served by a wide variety of
financial institutions with respect to the size of their market areas. There are numerous
institutions whose market area extends to the entire county, but most institutions have
localized market areas smaller than that of the City as a whole. Of 165 institutions
providing information, 26 % have geographic market areas encompassing all of Los Angeles
County ( if not larger) . Another 5 institutions' market areas incorporate all of Los Angeles
City, but not the entire County. Therefore, 48 institutions cover the entire City of Los
Angeles. At the other end of the scale, a remarkable number of financial institutions with

small market areas operate in Los Angeles . Some 102 institutions ( 62% of those reporting)
indicated their market areas was less than the area of the City.

128

338

RECOMMENDATION : The City should work with lenders to
fund loan pools u appropriate mechanisms for very small or

Darrowly specialized financial institutions to meet credit Deeds in
lower income and minority communities. These pools might tar.

get first- time bome-buyers, affordable multifamily bousing, and
small businesses .

A loan pool" consists of a pool of loadable funds contributed by multiple contributors;
loan decisions to use these funds are made by a fund administrator, in accordance with
the pool's declared social or other purposes . The rationale for singling out first -time

homebuyers, affordable multifamily housing, and small businesses in lower income and
minority areas as targets for pools will become clear in the course of Chapters 8 and 9.

One explanation for the absence of a correlation between the scale of lending flows and
market areas is institutional specialization . Of the 22 smal lenders with City -wide market

areas , only 27% are thrifts; the other 73 % of small lenders are commercial banks, whose
primary lending presumably occurs in other areas. By contrast, of the 13 large lenders with

City -wide market areas, 77% are thrifts . Overall, a much higher percentage of thrifts than
commercial banks have City -wide market areas : for thrifts, 45% of 29 reporting institutions
have City - wide market areas ; for commercial banks, 30 % of 81 reporting institutions do.

* Thrifts ( savings and loan associations) are defined in this study us institutions whose

regulatory agency is the Office of Thrift Supervision ( OTS) .
129

339

E. DO LENDERS ASSESS CREDIT NEEDS ?

Under the Community Reinvestment Act, institutions are responsible for meeting

the credit needs of their entire community ( market area ) , including the lower income areas
therein. CRA statements provide insight into the response of institutions to this duty. Most
reporting institutions - 92 of 107, or 86 %

stated they have conducted a community Deeds

assessment as part of their CRA responsibilities. However, 15 institutions acknowledge not
having conducted a community needs assessment. Anadditional 68 banks did not provide
any information on this question ; presumably fewer than 86 % of this group have done an
assessment.

CRA statements were examined for any mention of involvement with lower income
projects, with special credit programs designed by the reporting institution, or with guar
antee programs or other governmental programs. Some 35 % of reporting institutions
mention none of these activities in their CRA statements. Another 28%

mention one of

these categories; 24.5% mention two ; only 12% mention three or four.

The question of responsiveness also involves direct interaction with residents and
groups within the community. Under the enabling legislation , institutions are required

as a part of their assessment to make contact with community groups. The groups most
likely to be contacted in the course of such community needs assessments are: neighbor
hood groups ( 54 % of 56 responding institutions) ; " mainstream " chambers of commerce
( 30.4 % ) ; and social groups ( 27 % ) . The typical respondent bank contacted two or three

different types of group .'
Responsibility for CRA reporting rests most frequently with a vice president or senior

" There were no statistical differences in the diversity of groups contacted by larger and
smaller lenders. Small lenders contacted, on average, 1.8 different types of groups: most

often , Deighborhood groups, mainstream ( as opposed to minority) chambers of commerce,
and social groups . Large lenders contacted, onaverage, 2.0 different types of groups: most
often , neighborhood groups and nonprofit housing groups or agencies . It should be noted
that " other " was most frequently cited by both large and small lenders.
130

340

vice president ( 79.6 % of the 140 replies to this question) . Assistant vice presidents have
CRA responsibility in 15.7 % of reporting institutions, and compliance officers in just 4.6 % .
Taken together, this evidence indicates that most LA financial institutions have, at
best, a possive approach to assessing lower income housing and banking needs. At worst,
credit needs anessment is completely ignored. This suggests the following recommendation
of this study.

RECOMMENDATION : In conjunction with federal reg

ulators, the City should encourage financial institutions
to more seriously — and more creatively - assess the credit
peeds of low /moderate income and high -minority neigh .
borboods in their market areas .

F. PATTERNS OF INSTITUTIONAL SPECIALIZATION :

This section focuses on this question : do institutions generally specialize that is , do

they make loans of some types, but not others ? ' There is special attention to the provision
of credit for seismic rehabilitation .
Lenders' CRA statements reveal, for all 175 institutions responding, whether they

make loans for: single-family residences; multi-family residences; businesses; and individu
als ( consumers ) . Some 45% of these institutions make loans in all four.categories. Another

fifth of all institutions lend for single-family residences ( SFRs) and in two other categories:
10.3% make both types of residential loans, plus loans to consumers ; 5% make both cate
gories of residential loans, and loans to businesses; and 5 % make SFR loans, plus loans to

businesses and consumers. Another 13.7% lend only in both residential categories ( SFRs
and multi-famiły residences ) ; 5% make only SFR and business loans. All of this activity

.A closely related question is this: in any given credit category ( such as lending for multi
family residential development) , do lenders offer credit for all the purposes which borrowers
might reasonably require ? This question is addressed in Chapters 8and 9.
131

341

accounts for all but 13 reporting institutions, most of which make loans ( apart from SFR

loans) only to businesses and /or consumers ."

The use of CRA statements to assess the availability of credit offerings raises the question
( discussed in Chapter 6) of whether there are otherfirms which do not report under CRA
but which offer credit in these areas . There is no data on this question. It is reasonable
to think , however , that small businesses and affordable housing developers in particular

do not have significant alternatives to CRA -reporting institutions ( commercial banks, in
particular) assources of credit. There is no equivalent to mortgage companies in the area

of commercialand industrial credit. For individuals, alternatives to borrowing from CRA
reporting institutions ( banks,thrifts, credit unions) do exist. These range from finance
companies to pawnbrokers. Often exorbitant rates are charged. So apart from residential
lending, private sector alternatives to CRA -reporting institutions appear to be important
in providing credit to individuals. Wedid not pursue this further, because our studydid not
concern the provision of credit to individuals. However, see John Caskey, " Pawnbroking in
America: The Economics of a Forgotten Credit Market," Journal of Money, Credit, and
Banking, 23( 1 ) , February 1991, 85-99 .

132

342

make loans in all four categories. The pattern for national banks ( regulated by the OCC)
stands in direct contrast: 22 of 30 national banks ( 73% ) are active in all four lending
utegories; the other 8 make loans in three categories. State-chartered banks, the remain

ing category of CRA - reporting lenders, fall between these two extremes ( though closer to
national banks ) . Only one state bank makes loans in just two categories; 61% of 67 make
loans in all categories.

Seismic Rehabilitation Lending. The bank survey contained several questions
on seismic rehabilitation. Only 6 % of 51 responding institutions perform their own as
sessments of seismic rehabilitation needs in Los Angeles. Almost exactly one third of 43

responding banks indicated making loans for the seismic retrofitting of 1-4 family resi

dences, of multifamily residences, and of commercial buildings. Less than 3% of lenders
indicated they provide below -market rates for seismic retrofitting projects. Only 1 of 19

institutions are involved in a government seismic rehabilitation project.

G. THE GEOGRAPHIC LOCATION
OF BRANCHES AND ATMS :

We have data from two sources on the location of financial institution branches and

ATMs ( automatic teller machines) in Los Angeles County. Our bank survey asked insti
tutions for information on the number of branches operating, and the pumbers recently
opened and closed , inside and outside of the Los Angeles City boundaries. This gives

us some insight into financial institutions' choice between locating in city or in suburbs.
However, this data provides only incomplete insight into the overall pattern of geographic
location: it encompasses just the 69 respondents to the bank survey, and imperfectly de
picts banks' responsiveness to areas of greater and lesser income ( or greater and lesser
concentrations of people of color) .

Within Los Angeles itself are census tracts with a wildly divergent socio- economic
133

343

character. To delve further into the question of branch location within Los Angeles City
itself, we obtained precise geographic information on the location of all commercial bank
and thrift branches us of 1988. This section reviews results from these two sources of
information on financial institutions' branch and ATM locations in turn.
1. City versus County Location Decisions

The bank survey asked institutions to report on the number of branches and ATMs
they operate in Los Angeles City and Los Angeles County. The 69 institutions returning
the survey reported a total of 314 branches and 455 ATMs operating within the City of

Los Angeles, and an additional 513 branches and 767 ATMs operating outside the city
boundaries but within Los Angeles County. So the proportion of Los Angeles County
branches within the City itself is 38% ; the proportion for ATMs is 37% .

The bank survey suggests a location trend from the City into areas of the County
outside the City itself. Institutions were asked how many branches they have both opened
and closed since January 1, 1987. Overall, the 69 responding institutions reported opening
61 branches in the County, of which 28% are in the City. At the same time, some 91
branches have been closed, of which 31% have been in the City. There has been a net
loss of branches since January 1, 1987, in both the City and the County of Los Angeles.

Approximately 7% of all branches have been closed since the end of 1986, about 5% have
opened. The relative loss has been greater in the City than in the surrounding County.
About one- third of responding institutions that have closed branches indicate that

this consolidation is due to corporate restructuring ( for example, mergers and takeovers) .
Given the current trend in banking legislation and in the savings and loan industry, this

trend in branch closures is likely to continue in the near future.

Some 41 institutions operate Los Angeles County branches both in the City boundaries
134

344

and outside them . Only 4 operate branches within the City, without having at least one
Los Angeles County branch outside the City. But 17 institutions have Los Angeles County

branches without having any inside the City itself. Further, the 17 Dewly opened branches
reported within the City were attributable to 12 institutions, while the 44 branches opened
in the County outside the City were attributable to 27 institutions.
2. Patterns of Branch Location within the City of Los Angeles
Thus far, we have emphasized one aspect of bank location decisions -- whether branches

are inside or outside the City boundaries. To supplement our bank survey results, we

analyzed data on the location of all commercial bank and thrift branches in Los Angeles in
calendar year 1988. We mapped branch locations to census tracts, and then related these

in turn to the income and racial quintiles developed for use in Chapters 8-6. Matching
branch locations to quintiles provided several compelling insights.

( CBD ) downtown and in the mid -Wilshire district are excluded, only 33 % of branches are
located in " block -grant eligible " tracts .' Similiarly, only 47%

of all branches are in the

bottom 60 % of Los Angeles census tracts by median income, and only 40%

if CBD and

Mid - Wilshire branches are excluded . Map 7-1 on the following page shows bank branch

locations relative to the Los Angeles' upper and lower income areas.

135

60-893 O - 92 - 12

345

LA
of
Cily
Regions
Bonk es
Branch

LA
of
City

of
City
Angeles
Los
Areas
Income
across
Branches
Bank

free
leor
looo

Aon
Ancona
Upper

346

to 2.86 , while the former drops dramatically to 1.34. This pattern holds up for both income
quintiles. When all branches are considered, there appear to be a paucity of branches for
the first and second quintiles, but more available in the sero ( poorest) quintile.' Dropping

branches in the CBD and Mid - Wilshire from the analysis yields clearer trends: the bottom
three income quintiles have about 1.4 branches per 10,000 residents; the top two income
quintiles average about 3 branches per 10,000 residents.

Apart from residents ' need for banking services, businesses in any locality also need

these services. While data on the aumber of business establishments by census tract
were not available, we did have information on the number of commercial and industrial

( business) buildings in each census tract . We used this information , together with our data
on bank branches , to compute the number of bank branches per 100 business buildings

in each census tract. This measure of the availability of bank branches also indicates
systematic biases. Counting all branches, " block -grant eligible " areas had 1.06 branches

per 100 business buildings, versus 2.62 in other areas. There were 3.76 branches per 100
business building in tracts constituting the highest income quintile , and successively fewer

as income fell; in the lowest income quintile, there were only 0.61 branches per business
building. The same pattern was observed for racial concentration: the lowest minority
* The figures indicate about 2.56 branches per 10,000 residents in the lowest income quintile,
1.38 in the lowest of the middle quintiles, and 3.79 per 10,000 in the highest quintile.
• Data on buildings were drawn from the residential buildings source cited in Chapter 3.
136

347

LA
of
City
Regions
Bronch
Bonk os

LA
of
Cily

Angeles
Los
of
City

Minorily
Lower

de
Areas
Minority
across
Branches
Bank

Minority
Higher

348
quintile had 5.33 branches per 100 business buildings, versus 0.42 for the highest minority
quintile. In all cases, excluding CBD and Mid -Wilshire branches simply makes these
disparities more dramatic.
There are more deposits per residence in higher income areas than in lower income

areas : ignoring CBD and Mid -Wilshire branches, there are $ 8.74 in per capita deposits

( using 1986 population estimates) in block pant eligible " tracts, versus $ 21.40 in deposits
per person elsewhere in Los Angeles. At the same time, there appear to be more loans
made per dollar of deposit in higher income areas than in lower areas. No direct measure
of the volume of loans as of 1988 was available. A simple summation of residential loans
over the 1981-88 period, however, served nicely as a substitute measure of loans in 1988.
This procedure suggested a total of about $ 36 billion in 1981-88 residential loans, versus
$ 50 billion in 1988 deposits. For illustrative purposes, we set the loan /deposit ratio to

1 for Los Angeles as a whole. In 'block grant eligible" areas, bowever, this loan /deposit

ratio was less than onespecifically, 0.81; and elsewhere in Los Angeles, the loan /deposit
ratio was 1.08.
In sum , these data convincingly demonstrate that bank branches are systematically
less often located in lower income and minority areas of Los Angeles than elsewhere. They

also amplify the points made in Chapters 3 and 4: that is, residential loon flows in lower
income and minority areas are lower than would be expected on the basis of deposit volumes
in those areas - even taking into account the paucity of branches in these areas. This
evidence suggests a need for more bank and thrift branches in lower income and minority

areas. Bank officers have observed that these branches are not profitable because of the
low deposit levels such branches will likely attract. Our data do show that deposits per
capita are lower in branches located in low income and high minority areas. However ,
this does not reduce the need for bank branches, especially in light of the special needs
for banking services in lower income areas ( discussed below ) . A creative solution may be
required. This leads directly to one of the study's recommendations.
137

349

RECOMMENDATION : The City should advocate the es
tablishment of multi-bank branch offices serving low and
moderate income areas within the City.

H. THE PROVISION OF BANKING SERVICES

Depository institutions distinguish between depositors and non -depositors when it
comes to cashing checks for residents of Los Angeles with special needs. According to
lenders' CRA statements, only 6.4%

allow non -depositors to cash government checks.10

The bank survey asked a question about this and other practices pertaining to lower income

customers. Banks' responses to our survey also indicate that only 14.7% of 61 institutions
answering this question allow non -depositors to cash government checks in their branches .

The remaining 85% indicated that they do not. Further, some 64 institutions indicated in
the bank survey that they offer direct deposit of government checks for low - income elderly
people. However, only 44 % provide this service for non -elderly welfare recipients ( Aid to
Families with Dependent Children or AFDC, General Relief) .

More detailed information on multilingual services was requested in the bank survey .

Approximately 40 banks indicated they provide written information in languages other
than English on consumer deposits and loan programs ( about 71% in both categories) .
About 93% of respondent institutions maintain personnel who can accomodate non -English
10 This percentage is out of 94 institutions whose CRA statements provide information on
this point.
138

350

speakers, but only 35 % of respondent banks maintain ATMs with non -English instructions.
I. SPECLAL NEEDS DEPOSIT ACCOUNTS :

The bank surveys asked respondent banks about the most economical deposit account
they could make available for senior citisens on low fixed incomes , and for the son elderly
poor. About 61 of 64 institutions provided information concerning various characteristics
of the accounts they would offer to these two clienteles.

form of identification for check cashing.
Some 71%

of respondent institutions' accounts have no monthly service fee; for the

remaining 29 % , the monthly fees they charge usually range from $ 3 to $ 6 per month . An

unlimited number of checks can be written without a per-check fee being imposed at most
institutions; those with such a fee charge an average of $ .25 per check. ATM use is free for
these senior citizen accounts in 39 of 44 reported cases. When a fee is imposed for ATM
use, $ 1 is the most commonly charged per use."

No minimum deposit is required to open an account for 18.6 % of reporting banks.
11 At least one bank restricts holders of its senior citizens' accounts to use of ATMs ( not
tellers at bank branches ) . Since ATMs are evidently much less widely available than bank
branches, this practice, if widespread , would significantly restrict these accounts' cost and
ease of use .
139

351

Another 10% require $ 25 to $ 50 to open an account; but most, 44 % ( 26 institutions)
require $ 100. The remaining banks ( 27 % of the total) require between $ 200 and $ 1000 to
open this account. The majority of these accounts ( 55 % ) require do minimum amount to
avoid fees; but 12 institutions ( 24.5% ) require depositors to hold from $ 500 to $ 2500 to
avoid fees.

non -elderly poor.

140

352

J. BANKING SERVICES AND SPECIAL NEEDS

This chapter has presented some troubling information about the relationship between

Los Angeles banks, on the one hand, and Los Angeles' lower income communities and
residents, on the other . The sheer size of Los Angeles' lower income areas and of its low
income population presents a challenge and an opportunity to the City's lenders.
However, evidence from lenders ' CRA statements indicates that community credit

needs are often assessed haphazardly, and sometimes are not assessed at all. And while
many financial institutions operate in Los Angeles , in some credit areas extremely few
lenders supply loans. Further, lenders' branches are disproportionately located in upper
income areas, and there is some evidence of a shift toward suburban locations.

of the study.

RECOMMENDATION : The City should encourage finan
cial institutions to allow non -depositon to cash their gov

ernment benefits checks, to directly deposit all govern
ment benefits checks for depositors, and to offer and public
cize " lifeline" deposit accounts for all low income residents
of Los Angeles.

141

353

Chapter 8

Lending for economic development

A. INTRODUCTION
Economic developmoent can be defined as all activities that either maintain and in
crease employment opportunities, or preserve and enhance the economic base and physical

infrastructure of the city. Broad -based and incessant economic development is a prereq
uisite for a healthy and vibrant metropolis . Economic development in lower income and
minority neighborhoods of the City is crucial in the present day for Los Angeles. With
out it, Los Angeles threatens to become an ever more polarized city with segmented job

markets, segmented and segregated housing markets, and a bimodal income distribution:
a First World city stacked on top of a Third World city."

residents.

This chapter will present information about economic development and lending based

on information obtained for this study. The discussion investigates some aspects of lenders'
provision of credit to businesses, especially small businesses, and to households. The inclu
sion of credit flows to households may seem inconsistent at first glance. But in a prosperous

Provocative and complementary discussions of the increasing polarization of Los Ange
les are: Mike Davis, City of Quartz: Excavating the Future in Los Angeles,Verso, New
York, 1990; and Paul Ong, Project Director, The Widening Divide: Income Inequality and
Poverty in Los Angeles , Graduate School of Architecture and Urban Planning, UCLA,
June 1989 .
142

354

market economy, all participants require access to credit flows. Indeed , credit for residen
tial construction and purchases is itself a vital component of economic development. For
most households, the equity built up through homeownership is their primary store of
net worth ; and when households start business enterprises, the collateral underlying their
start- up financing is most commonly their equity stakes in a home. So the flow of residen
tial lending for home purchases is intrinsically related to the process of business start- up.
This is an example of a " negative spillover " effect, as introduced in Chapter 1: barriers in
one economic activity affect other seemingly unrelated economic activities . Failure to put
housing in place makes other forms of economic development, including business start- up,

less viable in low and moderate income areas ( and in high -minority areas) ; and failure to
make residential loans on existing housing makes business start-ups and other forms of
economic development less likely.
However, due to an almost complete absence of data on economic development issues,

this chapter cannot explore these issues with the depth they deserve. After section B
reviews chapter highlights, Section C discusses our efforts to obtain information about

economic development and lending in Los Angeles. It particularly emphasizes the problem
of data availability in this substantive area .
Section D then discusses the scope of credit supply to households. Section E turns
to evidence supplied by lenders about their scope of credit supply to businesses, and
section F characterizes their responses concerning their disposition of loan applications,
especially those from small and large businesses . Section G summarizes and reports on
the perceptions of business people in the City's enterprise zones .

B. HIGHLIGHTS OF THE ANALYSIS
• In each lending category, a large number of institutions provide credit: 166 institu

143

355

first- time home-buyers, and just 17 institutions provide start -up financing for

C. DESCRIPTION OF THE DATA
Despite the intimate connections between housing affordability, residential lending,
and economic development, there is a remarkable lack of information about, and program

matic emphasis on , economic development concerns. HMDA reporting requirements do
not extend to commercial and industrial loans. CRA mandates only that lenders list the
types of loans to businesses they make, not the number and volume of such loans. Eco

nomic development concerns also receive li::!e focused programmatic attention. There are
at least 21 separate offices, agencies and units with economic development responsibilities
of some kind , within various departments of the Los Angeles City government. We are
unaware of any economic development credit needs assessment of the City conducted by

or on behalf of any of these units.
The two problems of lack of systematic data and lack of programmatic emphasis are

mutually reinforcing: the lack of data both reflects the lack of priority assigned to this area,
and at the same time precludes the kind of detailed assessment that would encourage more
weight being placed on economic development activities. This suggests a recommendation
144

356

of this study.

RECOMMENDATION : The City should encourage finan
cial institutions to make detalled information on loans in
areas other than residential lending available to the pub

lic on an annual basis. Particularly useful for economic
revitalization wouldbe annual disclosure of data on the

geographic distribution of business loans, and especially
of loans to small businesses.

The material in this chapter was pulled together from three sources. A limited amount of
data was culled from lenders' CRA statements. Lenders' CRA statements do not provide
figures on dollar volume or numbers of loans by category or subcategory. However, as
noted above, lenders filing CRA statements must specify the types of loans they make to

households and to businesses ( for example, business loans encompass start-up financing,
working capital, and expansion financing ) . Knowing the kinds of loans that lenders make
is , admittedly, of limited value. However, each type of loan pertains to a particular phase
in the process of developing a business enterprise: so even this limited information gives

some insight into the stages at which entrepreneurs might face particular credit bottlenecks.
Further, several questions on our bank survey pertained to lending to businesses.
We also attempted to gather information about economic development and lending
directly from entrepreneurs in low and moderate income areas . The City provided us

with lists of small businesses in each of the City's enterprise zones. We then invited a

randomly selected sample of business owners in the City's enterprise zone areas ( 200 in each
zone, for a total of 1000 ) to meetings at which they could discuss their relationship with
lenders. We also developed two multiple -choice format questionnaires - one specifically for
business owners or managers , and another for community residents. Attendance at these
meetings was disappointingly small. Nonetheless, some of our small- business surveys were
completed by meeting participants, and a number of lively conversations ensued at which
we informally gathered some impressions about the subject matter of this chapter. A copy
145

357
of the small business survey is included as Appendix 8.
D. THE SCOPE OF CREDIT SUPPLY

As discussed above, the flow of credit to households is an important component of
economic development. This flow fuels households' demands for goods and services. More
importantly, it allows households to acquire assets , particularly homes, and thus build up

equity which can serve as the basis for initiating entrepreneurial activity.

credit available.

Of the 175 lenders that returned CRA statements, most are active in making both

residential loans and loans to individuals for other purposes . Of the 166 lenders that
are active in financing SFRs, 106 indicated they make home improvement loans; 91 make
construction loans for SFRs; 83 make home equity loans; and 69 make loans for purchasing
SFRs. In each of these subcategories, less than 10% of responding institutions indicated
they did not make these types of loans.'

" Some 166 report making loans for single- family residences, and 132 for multi-family resi
dences; 121 make loans to businesses , and 134 to individuals .

* For categories with fewer affirmative responses, the actual percentage of institutions not
making these loans is presumably lower. It is probable, given CRA reporting requirements,

that institutions that did not state whether or nottheymade loansin a given subcategory,
do not. So whereas a higher percentage of respondent institutions indicate that they make
home purchase loans than make home equity loans ( 92% versus 90 % ) , only 78 responses
were compiled for the former subcategory, as compared to 96 for the second.
146

358

respectively, indicated they are active in these subcategories. Indeed, only 24.2 % of insti
tutions providing definitive information on their CRA statements indicate making purchase

loans for first -time buyers. That is, 25 of 33 lenders that mention first- time SFR buyers
indicate they do not make loans to this group. In the two remaining mubcategories, 46 %
of responding lenders make rehabilitation loans. Refinancing loans are made by 74 % of

responding lenders.

The statistic concerning loans for first time buyers is particularly worrisome if it is
broadly descriptive of lenders' policies. Breaking into the Los Angeles housing market is
a major hurdle for many households, because the down payment for a new home is most

typically and easily obtained from the equity earned by selling one's previous residence.
So first - time buyers are more likely than second- or third -time buyers to be squeezed by
the housing affordability crisis .

For consumer loans, auto loans are most frequently cited — 95 % of 127 responding
institutions providing yes /no responses make these loans. Credit lines are provided by
84.4 % of 94 responding institutions. Credit cards are provided by 65.4 % of 85 responding
lenders.

E. THE SCOPE OF CREDIT SUPPLY
In - Depth Analysis

This section and the next bring together data from lenders ' CRA statements and the
bank survey to shed light on financial institutions' relation to business ( and household)
credit needs. This section sets out the scope of credit offered to businesses by lenders.

Starting and running a successful business enterprise involves financing at several
points: startup financing to augment the owners' equity and provide for the business's
capital assets; working capital and the financing of equipment and inventory to facilitate
the business's day -to - day operations; and expansion financing for business growth.
147

359

Banks' CRA statements reveal that credit in some of these business-need areas is much
more widely available than in others. More than 80% of banks providing yes no responses
indicate making loans for equipment, inventory, and working capital . Expansion loans

are less available only 28 of 40 responding lenders, or 70 % , provide such loans. Start- up
financing is provided least frequently — just 17 of 46 institutions, or 42.5% , indicate making
credit available for this purpose. The relative scarcity of sources of start- up financing
is particularly critical for economic development in special needs areas and for low- and

moderate- income individuals. In both cases, the inability to secure adequate start- up credit
can derail the development process before it has a chance to start.

Respectively, the figures are 65 of 71 lenders, or 91.5% ; 53 of 64 lenders, or 87% ; and 53
of 66 institutions, or 84% .
148

360

responding lenders indicate that they have no minimum loan amount ( or, in 6 cases, no set

loan minimum ) . Another 14 ( 36 % ) do have a minimum C& I loan amount, which is equal
to or less than $ 5,000. Nine ( 23 % ) have minimum loan amounts ranging between $ 5,000,
and $ 50,000 . These responses indicate that most banks do not rule out small loans as a
matter of formal policy. The substance of bank practice may be another matter, as our
informal discussions with small business people indicated.
F. LENDERS ' CREDIT DECISIONS :

Our bank surveys asked several questions about various aspects of lender decision
making concerning loan applications. Respondents ' answers are summarized here.

Loan applications and decisions. Lenders were asked whether they used credit scoring
systems in making loan decisions. These systems are most frequently used in evaluating

Don -residential loans for individuals ( 5 instances of 53 responses, or 9.4% ) . Only one
lender uses credit scoring for either residential or business credit applications. About 60 %
of respondents indicated that businesses must be in operation for a fixed length of time
before becoming eligible for loan flows.

About two- thirds of responding institutions indicate that applications for non - residential
loans to individuals or for loans to businesses can be made at branch offices. Another 8%
of respondents take applications at branch offices up to a cap level ( this cap level averaged
$ 132,500 ) . Loan decisions, however, are most typically made in central offices. Less than
20 % of respondents invariably make decisions on individual or business applications at

branches. About a third allow branches to make credit decisions only up to a cap level

( the cap level reported here averaged $ 128,500) , and about three- fifths make all decisions
at their central office.

" These questions pertained not just to business loans per se , but also to loan applications
by and loans to housing developers. Lenders' experience with housing developer loans is
taken up in the next chapter.
149

361

Of applications that are not withdrawn, lenders indicate that about 86 % of small

business applications and 85.4%

of large business applications for credit are approved

at some level. These percentages are somewhat higher than the 80 % and 81% rates,
respectively, for housing developer and home improvement applications. However, approval
rates for home mortgages and personal loans are substantially lower - 60.1 % and 72.4% ,
respectively.

The level and conditions of loon offers. Respondents to the bank survey indicated
that loan applications by individuals, when accepted, are usually fully funded - about 55%
usually fund individuals' full loan requests, and another 38% usually fund between 80 and
100 % of these requests . By contrast, loans to businesses are seldom fully funded . Small
business loans are usually fully funded by just under 40%

of the 49 responding banks.

Another 40% indicate that they typically grant 80-99 % of small business' loan requests.
The remaining 20 % of respondents typically approve less than 80 % . Surprisingly, a smaller
proportion - about 30 % -of respondent banks indicated that they usually fully funded larger
businesses'applications, and a higher proportion of banks - about 27 % -usually fund less

than 80% of these businesses' loan requests.
On average, lenders' loan offers to developers are about 70.7% of the amount requested ;
this compares with 81.6% average loan offers to small businesses , and 73.7% average loan
offers to large businesses.

Another aspect of loan decisions is required collateral. Both large and small businesses
are typically required to have collateral on hand for any loans they request: lenders impose

average collateral requirements on small businesses of 96.6 % ; on large businesses, the
reported requirement is slightly higher, 97.9 % .'

Lenders were also asked to provide some information on the amount that they will
Note that the figures cited here are based on estimates made by those completing the
bank survey .

' As discussed in the next chapter, housing developers must put up collateral averaging
more than the value of the loan received.
150

362

loan on the assets of small businesses. While this information encompasses all types of
small businesses, it is revealing nonetheless. The highest percentage, 74.6% , was recorded

for vehicles; in order thereafter were: buildings / land ( 70.2 % ) ; accounts receivable ( 65.7% ) ;
equipment ( 63.6% ) ; working capital ( 45.1 % ) ; and inventory ( 39.2 % ) .

often ” in all three categories. In all cases , listed Dext among dispositions occurring " most
often ” is “ " denial, with other suggestions." The disposition " withdrawn by applicant”

dispositions " seldom occurs " is selected with approximately equal frequency.
Lenders were given an extensive list of reasons why credit applications might be denied .

For the various types of application, they were asked to indicate reasons that often occur
with a " 1," reasons that sometimes occur with a " 2 ," and infrequently occurring reasons
with a “ 3.” Unimportant reasons were left blank . Weighted averages can then be calculated
for each reason and each loan type, 80 as to estimate the most important overall reasons.
For individual loans, the two leading reasons for denying applications are " debt-to- income
too high " and " credit history ." Following next in order are " insufficient net worth " and

" employment history." For residential loans, by contrast, the same two reasons are most
* The pattern ofdispositions for rejected loans does differ for residential loans. Withdrawal
by applicants is described by lenders as often occurring with somefrequency, and as some
times occurring by most lenders. Further, denials with suggestions to reapply are less
frequent than for these other types of loan application .
151

363
important, but " insufficient net worth " is unimportant; and " lender appraisal of property
comes in ahead of employment history ."

The rankings of reasons for denial come out somewhat differently for business appli
cations. For both small and large business loans, debt-to - income too high " is the leading
reason for denial. A close second, in both cases, is ' inadequate capitalization ." Third
and fourth in both lists is " insufficient collateral and business' credit history ." " Owners'

credit history is as important a factor as 'business' credit history for small businesses,
but it is unimportant for large businesses .

G. PERCEPTIONS FROM SMALL BUSINESS MEETINGS
We now turn to some insights from community residents themselves. Surveys for small
businesses were distributed at meetings convened in South Central, East Los Angeles, and
Wilmington . Attending these meetings were owners and managers of small businesses,
chamber of commerce officials, and community residents.

Too few surveys were completed at these meetings to be statistically reliable, so de
tailed results from these surveys are not reported here. However, we can summarize some
general results from these meetings and from completed surveys of small businesses . Al

most all our survey respondents were minority owners or managers of businesses in Los
Angeles' enterprise zones. These businesses were uniformly small - all grossed less than
$ 1 million annually. The average age of respondents' business was 5 years. Neither the
respondents nor their businesses had ever declared bankruptcy. All but three of these
businesses had been started with personal funds or money from friends or family ; of the
remaining three, two had obtained funds from commercial banks, and one from a finance
company.

businesses had obtained loans, they had pledged collateral of between 80 % and 100 % , and

bad usually received less than their loan request ( sometimes as little as 50 % ) . Lenders'
152

364
most frequent reasons for denying loan requests were " high debt- to - income ratio " and " too

little experience." The most important banking service for these businesses was receiving
cash and making deposits. Only one in six had a bank or thrift branch within a mile of
its business office.

The conversations that occurred during these meetings yielded some further insights.

These small business owners and managers in South Central and East Los Angeles de
scribed a need for working capital loans, for more branches in these areas , and for greater
lender receptivity to their business plans for expansion. They particularly emphasized
the need for more convenient bank branches; they indicated that they and their employees

often had to travel relatively long distances to conduct normal banking business.
Closing the Gap between Small Businesses and Lenders.
The information in this chapter, drawn from both lenders and small businesses ,

demonstrates that a large gap separates lenders from many small businesses that are pa
tential borrowers. The gap is rooted in the entire system of income and credit flows in
Los Angeles. Unequal access to credit and unequal flows of credit to households create
large wealth and net worth discrepancies between individuals and between neighborhoods.
Chapters 3, 4, and 6 have documented that systematically lower numbers and dollar vot

ames of residential loans are being made in low and moderate income neighborhoods and
in high -minority areas. This implies that residents and businesses in low and moderate in
come neighborhoods have systematically lower levels of wealth than do residents elsewhere

( both absolutely and relative to income) ; and residents of high-minority areas have still
lower levels of wealth .

The evidence reported in section F above closes the circle. These results from the

bank survey document that net worth and capitalization are important considerations in
loan decisions, and that any loans that are made must be almost completely backed by
collateral. The result of this circle is that systematically lower levels of loans to businesses
and individuals will be made in low and moderate income areas and in high -minority areas.
153

365

Our interviews with owners and managers of small businesses suggest that, indeed, they
have had little experience as borrowers at financial institutions. Their funds, particularly
for start-up and for expansion , have largely come from elsewhere.

RECOMMENDATION 8: The City should encourage fi
Dancial institutions to sponsor centers in low and mod
erate income neighborhoods offering business counselling
and technical assistance for small businesses operating in
these areas .

154

366

Chapter 9

Credit and affordable housing development:

A. INTRODUCTION
Housing City residents involves two interrelated processes: first, the actual construc
tion or rehabilitation of the housing stock; and second, the exchange of ownership rights
in this housing stock . Credit is essential for both processes. The existing stock of housing
will fully perform its function of housing City residents if healthy credit flowsare available
for exchanging ownership of this stock and for occasional rehabilitation . Chapters 3, 4
and 6 explored patterns in the flow of credit for the City's existing housing supply. It is

equally important that the stock of housing itself be steadily improved and expanded, both
to replace units lost to age or accident, and to respond to the dynamically evolving needs

of City residents for affordable housing. In this process of adding to the housing stock, as
well, the availability of credit is essential.

This chapter examines the adequacy of the flows of credit currently made available
by financial institutions to developers of affordable housing in Los Angeles. This study
has used the term " affordable housing” to refer to the costs of both owning and renting a
residence. When referring to home ownership , the " affordability " question concerns what
percentage of households have the financial resources to purchase a given type of home in a
specific location . Households' financial resources consist of their income and wealth . This
chapter brings the second usage, concerning rental costs, into focus. Here affordability
is based on how much income households must spend to rent residences. The federal

government has established a standard concerning rental cost: a rent is " affordable ” if it
is no more than 30 % of household disposable income. The level of rents which constitutes
" affordable housing" depends on how much income is commanded by potential tenants.
The emphasis here is on affordable housing for low- and moderate- income households. Low
155

367

income households are those with no more than half the income of the median household in

Los Angeles. Moderate income households have between half and four- fifths of the median
household income. Our interest in this chapter thus concerns whether developers can find
the credit to supply housing suitable for these households .

with housing developers are presented here.

The data examined here are unique in that several questions on both the bank and
developer surveys asked about respondents' perceptions of how the markets for housing
construction finance operate. This provides us with the ability to compare borrowers'
and lenders' perceptions concerning, for example, the kind of credit that is or is not
available , and why credit applications are denied . Interestingly, both sides of the market
have complementary, not contradictory, understandings of where the problems are in the
operations and outcomes of the credit markets that support housing construction. This

consensus on several basic points, which is discussed in the last section of this chapter,
suggests steps for improving the performance of credit markets in this area .
* This chapter does not look at another key question concerning poorer households' experi
ence in Los Angeles' housing markets - that is, whether households at these income levels
can find adequate housing at " affordable” rentlevels. Our experience and otherstudies
have

156

368

B. HIGHLIGHTS OF THE ANALYSIS
1. Lenders' Involvement with Affordable Housing
• Roughly 2 out of 3 institutions' CRA statements indicate involvement in projects

2. Lenders' Involvement with Affordable Housing Developers
• Some types of financing for affordable multi- family projects are more widely avail

3. Basic Results from the Developer Surveys

• The 34 respondents to the developer survey are all involved in housing. Half are

157

369

income single- family units for owner -occupants.

• The respondent group is well established. About 80 % ofrespondents' companies

4. Financing Affordable Housing : Developers ' Views

• Developers' primary source of predeveloprirent financing is internal funds, with

5. The Underfunding of Loans : Developers ' Views
• Lenders most often offer from50 % to 80 % of requested loan amounts for devel

158

370

C. DESCRIPTION OF THE DATA
This chapter merges together three data sources. It uses some data from both lender
CRA statements and our survey of financial institutions. These data sources were intro
duced in Chapter 7 and used extensively in Chapters 7 and 8. This chapter also uses results

from a survey of housing developers which, like the bank survey, was conducted specifically
for this study. Participating departments of the City government of Los Angeles provided

lists of developers that have responded to proposals for the construction or rehabilitation
of affordable housing. A questionnaire was then mailed to 300 of the developers appearing

on these lists. Some 34 companies returned their questionnaires. To assure that answers
were unguarded , surveys were returned anonymously. Tables summarizing the results of

the developers' survey appear as Appendix 9A , and the developer survey itself as Appendix
9B .

Definition of Affordable Housing. A block of questions on the developer ques
tionnaire asked respondents to characterize the types of projects they undertake. One

distinction requested was between new construction and rehabilitation . Another was be
tween low-, moderate-, and upper-income projects. As noted in Chapter 7, a wide variety
of definitions of these terms are in use among financial institutions. To avoid this prob
lem in the developer questionnaire, respondents were given explicit guidelines for what
constituted low- and moderate income projects.
The distinction turned on the expected monthly rental cost . A " low - income" house

hold was defined as earning 50 % of median income levels, a 'moderate- income” household
as earning 80 % of median income levels. Households of two and of four were selected . In
turn , each of the resulting four income levels was divided by 12, to get monthly income,

and multiplied by 30% , the current federal standard for how much a household “ should ”
spend for rent. It was assumed, next, that the two person households would properly
be housed in a one-bedroom residence , and the four-person households in a two - bedroom
residence.

159

371

The result of these calculations were four target monthly rent levels for low and
moderate income households. Developers were asked to use these target rent levels ( for
one and two-bedroom residences) to determine whether they were building housing for
" low ," moderate ,' or ' upper " income residents. Developers were considered to be building

" affordable housing if it was designed for " low " or " moderate" income residents ( as defined
here ) . Obviously, this procedure is arbitrary; but it provides a consistent classification of
developer activity.

Financial institutions were treated differently. Ou pretest interviews of bankers es
tablished that lenders typically had their own working definitions of low income” or
" affordable housing. On the survey distributed to financial institutions, we simply asked
what their operational definition was for the " affordable housing " projects with which they

were or are involved. The answers to this question are discussed below .

D. LENDERS, MULTIFAMILY HOUSING
AND AFFORDABLE HOUSING :

We begin this chapter with information supplied by financial institutions that finance

housing. This information, gathered both from CRA statements and from our bank sur
veys, answers several questions about these institutions' behavior in the credit markets
that support housing construction. Do they lend for multi-family residences ? If so , what
types of credit do they provide? Are they financing affordable housing and if so , rental

or owner-occupied units ? This section considers these questions in turn .
The second question , concerning the types of credit provided , is particularly crucial.

Lending flows for multifamily and /or affordable housing construction ( including rehabili
tation) are largely lending flows to developers. The development process involves multiple
stages: predevelopment, which encompasses planning, design, and site acquisition; con
struction ; and permanent financing. The availability of credit facilitates the first stage
160

372

and is crucial in the second two ." The first two stages require medium -term credit obli
gations of varying amounts; the third stage requires a long-term credit commitment for a

given mortgage amount. So while making construction financing available assures a sup
ply of new multi- family housing units, the availability of permanent financing is equally
crucial in assuring that the units get to market.
1. Lending for Multi- Family Residences

About 49% of the 175 institutions returning their CRA statements supply credit for
multi- family residential loans. Lenders' information on the types of multi-family loans they
provide is somewhat sketchy. Some banks indicate affirmatively whether they make loans
of each type. For other banks, detailed information on loan types within a lending category

is missing. Since CRA statements must list all types of residential credit supplied , it can
reasonably be inferred that failure to cite loans of a given type implies that none are made.
The multi -family loans that lenders make most frequently we construction loans;
95.5% of the 86 institutions financing multi-family housing indicate they make such loans.

Within this broad lending category, & more detailed breakdown is available. The most
commonly cited subcategories of lending are refinancing and seismic rehabilitation loans.
About 36 % of these 86 institutions refinance multi- family residential loans ( 12 indicate

they do not) . Some 24 % of these institutions report making loans for seismic repairs and
rehabilitation ( 15 indicate they do not) .*

Only 6 lenders report making loons in either of two sub-categories — the predevelopment
financing and permanent financing of multi- family residential projects. Some 25 lenders
' The developers' survey , discussed below , provides insight into the link between the avail
abilityofcredit and construction of affordable housing. Many respondent developers in
dicatedthat withoutfinancing guarantees in advance, it isimpossibleto proceed with
projects .

* If the institutions that make multi-family loans butprovide no detailed information have
the same rate ofloanactivity, then 74% and 65 % , respectively, of these 86 institutions will
provide credit for refinancing and rehabilitation . But the lower percentages indicated in
the text are more likely to be accurate.
161

373

explicitly indicated they do not provide predevelopment financing, while 6 explicitly pro
vide no permanent financing. It appears that financial institutions do not, in general,

provide developers with project start -up financing or with long-term financing for afford
able multi -family dwellings. Hence, these developers must turn to other avenues for this
financing: to credit made available through the public sector, or — in the case of predevel
opment financing - to internal funds.

2. Lenders ' Involvement with Affordable Housing: CRA Data
Roughly 68 % of lenders' CRA statements indicate involvement in projects assisting

low /moderate income residents. This percentage is higher for larger lenders ( 84 % ) and
lower for smaller lenders . The same pattern emerges with special credit programs. Of
the 37 larger lenders for which we have information, 83% are involved in special credit
programs, whereas 52% of the 55 respondent smaller banks are in volved.

3. Lenders ' Involvement with Affordable Housing:

The bank survey distributed to lenders asked lenders several questions about their

involvement with affordable housing programs. Lenders were asked whether they had se
cently financed , or were currently financing, housing for low- income renters or owners .

Lenders themselves were asked to provide their working definition of “ low -income hous
ing."

Of the 34 institutions indicating involvement with low -income rental housing, 50 %

The terms " larger lender " and " smaller lender ” are defined in Chapter 7.
* It would have been preferable to supply a definition of low income housing to banks,
and then ask whether housing meeting this definition was being financed . Given that this
would have been potentially costly forrespondents, who were completing the questionnaire
162

374

indicated their operational definition of " low income" was not available; another 18%

dicated their institutions did not define this term ; and 32% responded with a definition of
some kind. Of the 34 institutions involved with low -income, owner-occupied housing, 32 %

indicated their definition of " low income was not available, and 9% that the term was not
defined ; 45% provided a definition .
There was remarkable diversity in the definitions lenders did provide. In the case of

low - income rental units, the largest group of lenders ( 5) defined " low- income" as being

less than 80 % of median income. One lender defined it as any below -market rent project.
Another defined it as congruent with HUD-subsidized units. In the case of low -income
owner -occupied housing, only 2 lenders used the 80 % -of- median - income criterion. Some
lenders ( 4) defined low -income owner -occupied housing by the census tract in which it was
located . Three lenders defined low -income as housing below a prospective price: respec
tively, the prices cited were $ 40,000, $ 135,000, and $ 150,000. One lender used the criterion

of $ 20,000 income per year, another the criterion that the residence has been built for the
Community Redevelopment Agency.

Lenders were then asked whether they had been , or were currently involved in , financ
ing low - income housing ( as they defined it) in multi-family rental units. Some 14 of 33
responding lenders had financed units of this kind that were completed in 1989 or 1990: 6

had financed between 1 and 50 units, and 8 had financed 51 or more. Only 6 institutions
indicated they were currently financing units of this kind for which construction had be
gun , and only 5 had committed financing for projects that had not yet begun . One telling
statistic is that only 3 institutions had non -zero activity in all three areas ( finished , 1989
and 1990; underway; and committed ) ; but 16 institutions circled " zero " units in each of
these three categories. Another 9 institutions provided responses other than zero only for
units completed in 1989 and 1990 .

Lenders' loan totals for the projects being undertaken in this area were elicited. Using
on a voluntary basis , this approach was foregone.
163

a rough estimation, these lenders extended approximately $ 116.2 million for projects com

pleted in 1989-90, but only $ 38.5 million for projects under construction, and $ 24.1 million
for committed units. For both new construction and rehabilitation projects, lenders are
most often providing construction financing, and rarely either predevelopment or perma
nent financing.

Lenders' results for low - income, owner-occupied housing were even worse than for
rental units. In this category, 21 institutions ( 65.6 % of the total) had completed no projects
and 9 ( 34.5% ) had financed at least some completed units in 1989 and 1990. Only 4
institutions reported having begun construction on units of this type ( 18 reported none) ,

and only 3 had committed funding for such units ( 17 reported none) . Again, far more bank
credit had been extended for completed 1989-90 projects ( $ 57 million reported) than had

been committed for future projects ( $ 11 million ) or was being used for ongoing projects ( $ 11
million ) . Lenders did , however, indicate a greater likelihood of providing both construction
and permanent financing for low - income, owner-occupied housing.

164

mang

375

376

E. CREDIT FLOWS TO DEVELOPERS
OF AFFORDABLE HOUSING :
In -Depth Analysis of Lenders' Experience

The bank survey we sent to financial institutions asked several questions about banks'
experience with lending to developers. This section summarizes the responses received to
those questions. In so doing, we will sometimes refer parenthetically, by way of contrast,
to lenders' answers concerning their experience with small and large businesses ( already
discussed in Chapter 8) .

The central question is what percentage of the loan applications are funded, of those
not withdrawn. For the 36 financial institutions answering this question, an average of

80% of developer applications are funded . This percentage is somewhat lower than the
85%

and 86 % recorded for large and small businesses, respectively. Specifically, 54 %

of lenders indicate financing 76-100 % of developers' applications; another 19% , 51-75%
of applications; and the remaining 27% of lenders finance 50 % or fewer of developers'
applications for credit.

The level and conditions of loan offers. On average, lenders offer developers
about 71% of the amount they request. This is somewhat less than the 74% and 82%
average loan offers they make to large and small businesses, respectively. Of 29 respondents
to these questions on the bank survey , only 14% usually fully fund developers' applications.
Most ( 55% ) fund 80-99 % of the amount requested; 17% fund 50-79 % and another 13%
average less than 50 % . Lenders overwhelmingly will rely on cosigners if this is necessary
to make a loan application creditworthy ( in lenders' views) .
Overall, higher collateral is required by lenders on loans to affordable housing devel

opers than on loans to small businesses or to large businesses. On average , 100.9 % in
collateral is required against loans to developers. By contrast, the average collateral re

quired on loans to small businesses is 87.4% of the loan amount, and 90.5 % is required
165

377

on average against loans to large businesses. Some 41%
that they require between 51%

of 34 responding banks indicated

and 90 % collateral on loans to developers. Only 4%

respondent lenders indicated that they require 91% to 100 % collateral on these loans. But
19% of lenders require between 101 % and 185 % collateral against loans to developers, and
97 % require more than 125 % .

Unsuccessful loan applications: bankers' perspectives. As discussed in Chap
ter 8, lenders were asked how unsuccessful applications were disposed of, any why they
were typically turned down. Lenders' responses concerning the disposition of unsuccess

ful applications were the same for housing developers as for businesses : flat denial was
most often cited as occurring often or sometimes; denial accompanied by suggestions for

reapplying was cited as occurring sometimes; and withdrawals by applicants occur seldom .
The three most frequent reasons for denying developers' applications, in order of im
portance, are: inadequate capitalization ; insufficient collateral; and debt -to -income ratio
too high. Three additional reasons for denying credit applications were cited with some
frequency, although much less than this first group. These additional reasons for denial
were business' credit history, owner's credit history, and the risk of the line of business.
F. BASIC DATA FROM THE DEVELOPER SURVEY :

The evidence reviewed up to this point in this report tentatively suggests that while
financial institutions have many different lending specialities, and many lenders offer loans
of many different types , some areas of credit need are underserved . The observed disparities

in the levels of residential flows among different areas of the City, documented in Chapters
3 and 4, constitutes indirect evidence of this underserving in the residential credit area .
Chapter 8 and this chapter have used financial institutions' own CRA statements to
unearth areas within loan categories in which credit is infrequently supplied. Further, these
chapters have reviewed lenders' survey responses , which indicate that loan applications by
166

60-893 O - 92 - 13
.

378
either businesses ( Chapter 8) or developers ( this chapter) are frequently either denied or
funded at substantially lower levels than have been requested . This evidence supports the
intuition of underservicing of credit needs, but is only from the supply side of the credit
market. The possibility remains open that these limitations on supply reflect a simple lack
of demand by qualified applicants.
The previous chapter introduced some largely informal evidence to the contrary from

small business owners in the City's Enterprise Zones. This section and the next provide

more detailed information on the demand side of one important set of credit markets
that is, the credit markets that facilitate the construction of affordable housing in Los

Angeles. As discussed in section C above, this information is drawn from a mail survey
of Los Angeles developers. The remainder of this section explores basic information from
this survey about the firms and individuals surveyed. This information will show that
participants in this survey are, by and large, a group with substantial experience and with

1. Characteristics of the Developers Surveyed

Developers of affordable housing are businesses, either for -profit or not- for -profit , in

volved in building multi-family housing for low and moderate income residents. Most of the
34 respondents to the developer survey are developers: 45% are for- profit housing devel
opers and 32% are non - profit housing developers. The remaining 23 % of respondents are
housing professionals. Four out of five respondents ( 81% ) are organized as corporations;
another 9% are partnerships. The respondents display some ethnic and gender diversity :
18% have female control groups, and 12% have mixed -gender control groups. Some 31%
of respondents work in firms controlled by minorities .

An outstanding feature of the respondents to this survey is their record in undertaking

projects. Just under half ( 48.5 % ) of respondents' companies were founded before 1980;
167

379

another 30 % were founded between 1980 and 1985, and 12.1% , between 1986 and 1988 .
Some 87%

of the companies surveyed have undertaken at least one project; 83 % of all

companies initiated their first project in 1988 or earlier, and 40%

before 1980. Most

of these developers ( 72% ) have more than 75 % of their projects in Los Angeles County,
and a majority ( 58% ) locate at least 75 % of their projects in low - income areas of Los
Angeles County. Respondents report 421 completed projects for detached homes, 104
for multifamily owner -occupied units, and 2,178 for multifamily rental units. Another
519 projects for detached housing are either under construction or planned by respondent
developers, as well as 434 multifamily projects for rental units.
Only 7 projects are planned or under construction for multifamily owner -occupied
units. Gross revenues in 1989 exceeded $ 1 million for 35.5 % of respondents; another 26 %

had gross revenues of $ 100,000 or less; and the remaining 39% had gross revenues between

$ 100,000 and $ 1 million. Just under half of all respondents ( 47% ) had more than 10
full-time employees; another 50 % bad between 1 and 3 full- time employees.
Another outstanding characteristic of responding developers is their experience and

their prior record of avoiding insolvency. Just one of 33 respondent companies has ever
declared bankruptcy. Some 22 of the 34 individuals completing the survey have more than
10 years of experience in this field ; only 5 had less than 2 years of experience. And 23 of 34

individuals have 5 or more years' experience as either owner or board member. Meanwhile,
none of 33 responding individuals have ever personally been part of a developer company

that has declared bankruptcy.
2. Developers ' Project Experience

Several questions asked developers to characterize their companies ' building experi
ence and emphasis. They first described the types of ( new ) construction project their

company is currently involved in. The largest number of respondents ( 11) indicated low
income multifamily rental projects; 4 indicated moderate - income projects in this category,
168

380

and 3, high -income projects of this type. The exact definitions of " low ' and 'moderate "
income used in responding to this question are discussed above. The next most frequent
type of project involved single-family owner -occupied units; 6 developers are currently
building such units for moderate - income owners , 5 for upper- income owners, and 4 for

low - income owners . Only 4 developers indicated any current involvement with multifamily
projects for owner -occupants. Rehabilitation projects of any type were mentioned infre
quently. Some 7 developers indicated current projects for low - income, multifamily units;
but only 6 projects were mentioned in all other categories."

The survey set out a list of factors that might figure into the " predevelopment” or
planning stage; respondents were asked to assess whether these were " very," " somewhat,"
or " least " important. The item that was most consistently listed as " very important" was

the availability of permanent financing, followed closely by two other factors: zoning laws
and the availability of construction financing. Generally selected as somewhat important
were lenders' appraisals and others' appraisals, with three other factors just behind: tax

credits , architectural drawings, and the availability of insurance. The least important
factors were engineering estimates and Section 8 guarantees.

' Developers' previous project experience largely mirrored this current experience. The
most frequently cited projects were multifamily rental; of these, 13 had been builtfor
low -income tenants, 9 formoderate-income tenants, and 3 for upper-income tenants. Next
in frequency were singlefamily owner-occupied units, with 7 each in the moderate income
and upper-income categories and 4 in the low - income category .
169

381

large proportion of low -income residents; a concentration of subsidized and public housing.

Most developers responded that a project was more viable if planned for a high -minority
area. Respondents were split vis- a -vis low -income areas: a large number thought it made
a project more viable to be planned for a low -income area, while a large number also
thought this would make a project less viable. Equal numbers of respondents thought that
planning a project for an ares already dense with subsidized and public housing would
make it more or less viable, or would not affect its viability. Two related questions later in
the survey drew a more definitive response on this same general question. These questions
asked whether it was easier or harder to get financing for low /moderate income projects

in high-minority areas , and in low -income areas . The overwhelming reaction was that it
was harder to get financing for low /moderate income projects both in high -minority areas

( 15 respondents chose harder for commercial banks and /or other lenders, 1 chose easier)
and in low -income areas ( 14 chose harder, 1 chose easier ) .

About two-thirds of developers ( 19 of 30 responses) indicate that they commonly con
tact both the Community Redevelopment Agency and the City's Department of Housing
Production and Preservation prior to undertaking projects. The other governmental con

tact mentioned with some frequency was the California state housing agencies ( 13 of 30
responses ) . The most frequently cited purpose of these contacts is to obtain money ( 21

30, or 70 % of responses) . The next most frequently mentioned purpose was technical or
other assistance ( 15 of 30, or 50 % of responses) .

Terms and conditions of loan offers . The developer survey asks about the terms
and conditions on the various types of credit obtained . Predevelopment loans average 25
months in length, and sometimes require a cosigner. Development loans average about 30
months in length , and usually require no cosigner. These loans may carry either a fixed or
floating rate. For both categories of loan , collateral is required in about three of four cases,
usually in the range of 80-100 % of the loan. Some 13 developers indicated specifically what

is used as collateral: receiving 9 mentions each were all company assets and the property
170

382

itself; next, with 8 mentions, were the owner's residence and the owner's other real estate

holdings. Also included us collateral, in fewer cases , are the company's inventory or its
accounts receivable, the owner's financial holdings, and other short- term company assets .

G. DEVELOPERS ' VIEWS ON FINANCING

The survey mailed to developers asked several questions about how they financed
the various phases of their development projects. The primary source of predevelopment
financing is company internal funds ( 14 respondents ) , with outside investors next ( 10 ) .
Commercial banks and savings and loans receive some mention us primary sources in this
phase of a project ( 4) , as do government mortgage funds ( 3) . Overall, a ranking of the
sources of predevelopment financing from high to low finds: company's internal funds,
2.3; outside investors, 1.3; family funds, commercial banks, and savings and loans, 0.6 ;
government mortgage funds, 0.5; friends, 0.4; finance of mortgage companies, 0.3; credit

unions, 0.2 .

The next phase of projects after predevelopment is development ( construction ) per se.
Here, commercial banks ( 15 mentions) and savings and loan associations ( 11) predominated
as the primary financing sources. The only other sources mentioned more than once
as providing primary financing were company internal funds ( 5) and finance or mortage

companies ( 3 ) . Interestingly, all 3 companies citing internal funds as a primary source
of development financing also cited it as such for predevelopment financing . Ranking the
sources of development /construction finance from high to low , as above, yields these scores ::

*A comprehensive ranking of each financing source was developed by assigning a " 3 " to
each mention of it as the primary financing source, a " 2" to each mention of it as a
secondary source, and a " 1" toeach mentionofit as a tertiary source. Underthis method,
e given épancing source would receive a " 3" if all developers uniformly mentioned it as
their primary source, " 2 " if all developers uniformly mentioned it as their secondary source,

and " j" if no developers mentioned it at all ( even as a tertiary source) .
• The converse does not hold : 9 companies cited this source as primary for predevelopment,
but not for development.
171

383

commercial banks, 2.7; savings and loans, 2.0; government mortgage funds, 1.9; outside
investors, 1.7; company's internal funds, 1.6; finance or mortgage companies, 0.7; friends,
0.5; family, 0.4; credit unions, 0.3.10

Cited most frequently as developers' biggest cost item is property acquisition / land

( 46% of 24 responses ) ; also cited were construction materials ( 25 % ) , labor costs ( 21% ) , and
legal/ compliance costs ( 8 % ) . When asked about their second biggest cost item , developers
mentioned a number of items: construction materials ( 29 % of 21 responses ) ; labor costs

and interest costs ( 24% each ) ; and land acquisition ( 19% ) . In half of 22 reported cases ,
developers' fastest growing cost item is legal/ compliance costs; also mentioned prominently
among the fastest growing costs are land acquisition and land costs .

2. Financing Constraints on Development P : ojects
Two questions on the developer survey pertained to whether financing constraints

impede developers from undertaking projects. First, developers were asked whether they
would begin work on a project without a financing guarantee. Some 11 of 29 respon

dents ( 38% ) indicated they would go forward on a development project, hoping to line
up financing later. However, most companies ( 58.5% ) responded that they would hold a
project up until they had secured financing. About half of this group would not start a
10 Most institutions ( 23 of 28 yes / no respondents) indicated that this rank order of credit
sources would not change for a high -minority or low - income area .
172

384

project without construction financing; the other half indicated they needed an assurance
of permanent financing before proceeding." Only 4 of 14 respondents indicated that gov
ernment guarantees are " always " or " sometimes ' required before lenders agree to finance

their projects.
The second question on finance constraints was this : which aspects of low / moderate
income housing projects are commercial banks reluctant to finance ? Of 19 developers
that responded to this question, at least 11 indicated that lenders are reluctant to finance

four cost items: property acquisition / land; construction materials; labor costs; and le
gal/ compliance costs. The only cost item which commercial banks apparently are willing

to finance is interest costs, with just 3 citations. "
3. The Underfunding of Loan Applications:

As discussed above, about 38 % of lenders usually offer between 80 % and 100 % of the
amount of developers' loan applications. But the other 62% usually fund between 0 % and
80 % . For convenience, we describe loan offers in the 0-80 % range of requested loan amounts

as 'underfunded ” loan applications. When asked if they accept underfunded ( 0-80 % ) loan
offers, developers replied that they usually ( 7 of 8 cases) do.

" Some developers characteristically use the same lender for predevelopment and devel
opment financing ( or for development and permanent financing) ; many do not, and these
arrangements are often not made in advance.

19 Interestingly, most developers reported that lenders' appraisals of projects — which are
almost always conducted, according to the survey - are either the same as or lower than
independent appraisals.
173

385

loan applications: debt/income ratio ( 8 mentions, of 14) ; insufficient collateral ( 7) ; risk of
project ( 6) ; too small loan amounts ( 5) ; area of company location ( 2) .

location, 9.13
Lenders also frequently require borrowers to take various actions as a condition of
receiving development financing. Some 15 developers answering this question cited the

following actions: increase equity ( 7 mentions) ; increase collateral ( 6) ; obtain insurance
( 5) ; hire an accountant or consultant ( 2) .
H. SUMMARY : COMPLEMENTARY PERCEPTIONS

These two sets of perceptions from the borrowers' and lenders' sides of the credit
markets that support affordable housing development are very revealing of just how

and

how well, from a social perspective — these markets work. There is a remarkable amount of
consensus among the developers and lenders surveyed about how these markets work, as
evidenced by their nearly identical estimates of what proportion of loan requests are funded,
why loans are denied, etc. This consensus surely reflects the fact that the developers whose
answers are included in the above results are generally experienced and associated with
1 ' One question concerned whether lenders were likelier to deny or set stiffer terms for
rehabilitation projects, as compared to new -construction projects; the few developers re
sponding to this question ( 6 in all) indicated some feeling that rehabilitation projects were,
indeed, disfavored by lenders in these ways.
174

386

experienced and successful organizations.
At the same time, the fact that there is broad consensus among borrowers and lenders

about how these credit markets function does not mean there are no problems with these
markets' workings. The results summarized above do indicate several areas which these
markets miss, and some behavioral biases which may have important consequences for the
end product of these credit markets' operation — that is, important consequences for how
much affordable housing is actually put in the ground in the City of Los Angeles.

• Compared with small and large businesses more broadly, developers of affordable
housing are less likely to have their loan applications approved, and more likely to

have applications which are approved be underfunded . Nonetheless, developers
generally accept underfunded loan offers. Developers are also generally subject
to higher collateral requirements than the other categories of borrowers.
• Borrowers' access to credit markets is generally correlated with their experience.

• It is also generally more difficult to find financing for the early ( planning or

projects.

• Not many lenders ( 6 out of 175 for which we have CRA statements) report mak
175

387

ing loans for predevelopment financing or for the permanent financing of housing
developments; this contrasts with lenders' widespread willingness to make con
struction ( development) loans. Again , there is a possible sticking point here, since
many developers report that they cannot initiate development projects without
pre-arranging both development and permanent financing .

• Developers report that it is harder to get financing for housing projects if these

• Both developers and lenders agree that lenders' principal stated reasons for deny

cate that they often recommend that developers should increase their collateral

or capitalization to become more creditworthy. This, too, poses some particular
problems for newer developer companies, whose lack of " deep pockets " and lack
of reputation may preclude their taking these lenders' advice.
• Developers' biggest cost item , land and property acquisition , is also cited along
with construction materials as the hardest cost item for which it it is most difficult
to finance adequately.

176

388

Chapter 10
Ranking bank services and
A. INTRODUCTION

This chapter suggests how the ranking system for financial institutions, developed for
residential lending in Chapter 5, can be extended to encompass bank services and non
residential lending ( Chapters 7 to 9 ) . The rankings in Chapter 5 compared lenders' relative
performance based solely on their residential lending to low /moderate income areas and

to areas whose residents include many minorities. This ranking should be broadened to
encompass other banking activities, as these too are functions for which financial institu
tions should be held socially accountable . Indeed , the next section ( B) reviews the linked

banking programs developed by Massachusetts and by Boston in some detail. The rank
ing systems employed in these programs encompass much more than simply residential
lending.
A broader ranking scheme can include these items if data pertinent to performance

in these areas of institutional behavior are available. Our experience in conducting this

study suggests that any information needed for this broader ranking can be obtained from
a survey sent to financial institutions. This chapter suggests some ranking criteria for the
three areas of financial institutional behavior covered in the last three chapters:
• Banking Services forConsumers ( Chapter 7) : The location of branches,

177

389

Following the discussion of linked banking programs in other states, sections ( C - E)
develop a set of criteria for ranking financial institutions in each of these areas. Each section
has a common structure. Each briefly explains the standards for social accountability
that are being used implicitly. Then a set of performance criteria are developed that

the City might use in an ongoing linked deposit program . To implement this or some
similiar - expanded ranking system , the City would have to collect information directly
from lending institutions. The final section ( E ) then discusses how a ranking system for
financial institutions incorporating these various dimensions of banking behavior might be
implemented.

B. TWO APPROACHES TO RANKING :

Outside of the Community Reinvestment Act evaluation process conducted by the
Federal regulatory authorities, most lender ranking schemes have been developed in the
context of Linked Deposit Programs ( LDP's ) . Linked deposit programs have been imple
mented by a large variety of municipalities and states. The reasoning behind a linked de

posit program is straightforward: preference is given in allocating public deposits to lenders
whose activities are more heavily concentrated in specific areas and activities deemed to be
publicly valuable. A linked deposit program has two effects: it directly channels funds to
the lenders most likely to use them in desirable ways; and it establishes incentives for other
lenders to shift towards more socially desirable activities, so as to qualify in the future for
benefits under the program .

' In developing suggested criteria for ranking institutions, we have been sensitive to the
need to avoid items with onerous costs for respondent institutions. The bealthy response

to our bank survey for this study is a good indicator that many lenders will participate

when there is some incentive such as good will or revenue - for so doing.
" Perhaps the best known linked deposit programs are those implemented by the state of
Massachusetts and the cities of Boston , New York and Minneapolis-St. Paul.
178

390

behavior is a demanding task . The goals of the linked deposit program must be implicit
in any such ranking scheme. A typical ranking scheme has three components:

• Classification: Lenders are placed in one of a pre-determined number of groups

An institution's ranking is used to determine if the institution is eligible to participate
in the program , and perhaps also determine the extent of benefits it qualifies for under
the program . In case of an LDP, the ranking normally determines whether the institution
receives low cost deposits and the size of the deposit itself.
It is instructive to look at ranking schemes developed by other jurisdictions that have

implemented linked deposit programs. Most of these, including New York City, have
not made the specific ranking criteria they use publicly available. However, the State of
Massachusetts and the City of Boston have. These two programs' ranking schemes are

described in turn

State of Massachusetts Ranking Scheme. The Massachusetts program , which
has been extensively analyzed by Campen ( 1985) , requires banks to submit interest rate
bids on state deposits, along with supplementary information on their in -state lending

activities. Awards under the program are determined based on a " weighted" interest rate
equal to the sum of two components: ( 1) the actual interest to be paid on the deposit, and
( 2) a " social return " based on the bank's lending activities in various targeted areas . In
theory, banks offering the highest weighted return receive deposits under the LDP.
The " social return " offered by a bank is calculated on lending in six targeted areas
plus overall lending within the State plus investment in securities issued by Massachussets'

public entities as a percent of the bank's total in -state deposits. Banks participating in

the program are assigned to one of three leagues. Each league is considered qualitatively
179

391

different, and is required to meet a different standard in terms of minimum offered interest
rate to participate in the LDP. The leagues and criteria are :

Each bank's score is calculated by a formula known in advance by all participants. This
scoring system focuses on the total return to the State, including both the interest rate
bid on the deposit and a social return component, LDS below .
B+C

.: ( *)

Total Return = p + LDS == + .01
where

The C component of LDS targets particular areas and activities of concern to Mas

sachusetts. Campen ( 1985) examines the performance of this program and concludes
that while the scoring system itself was well-designed, the program has been poorly imple
mented. Awards under the program have borne ao systematic relation to scores received,
bank's scores have seldom been publicized, and the ongoing monitoring necessary to induce
lenders to improve their performance has not been carried out effectively.
180

392

City of Boston Ranking Scheme. Mayor Raymond L. Flynn signed an Executive
Order creating the Linked Deposit Banking Program for the City of Boston on September

26, 1989. The Community Banking Commission implementing this Program made its
ranking criteria public in a Report to the Mayor released on June 6, 1991.
Lending institutions are evaluated on their performance in four substantive areas in the

Boston scheme: ( 1) mortgage lending, which is given a 33 % weight in the overall ranking;
( 2) affordable housing and economic development, which also receives a 33% weight; ( 3)
banking services, which is given a 24 % weight; and ( 1) employment, a 10% weight. The

mortgage lending score is based on analyzing each institution's number and dollar volume
of mortgage loans per 1,000 housing units in minority and white neighborhoods of similiar
income levels, using the most recent HMDA data. A variety of factors went into the
mortgage lending ranking, including the number of conventional and VA /FHA loans, and
participation in the Neighborhood Reinvestment Plan . This Plan was a City initiative

designed to involve participating banks in a variety of below -market mortgage products,
new branch commitments, and homebuyer counseling programs.

Affordable housing involvement is based on whether banks had made loans to create
and preserve affordable housing; economic development performance is measured by the
amount of small business lending in Boston's neighborhoods, and especially lending to

minority- and women -owned businesses. The score for banking services is based on the
location of branches and ATMs, with special weight given to branches in lower income and

minority neighborhoods, and to programs targeting the needs of low - income and elderly
customers. Finally, the employment score analyses the number of women , minorities,

and Boston residents in each institution's workforce, with special weights for women and
minorities in managerial ranks, and for programs encouraging the hiring and promotion of

women and minorities.
The City of Boston groups banks into cohorts to insure that rankings are done across
comparable institutions. This " cohort" approach is advocated in this report: the discussion
181

393

in Chapter 5 suggests that institutions be ranked by cohort according to the volume of

their HMDA lending activity. Boston instead groups institutions into cohorts based on
their asset size: " Group

banks have over $ 3 billion in assets; " Group It" banks have

between $ 500 million and $ 3 billion in assets; and “ Group IT consists of all banks with
assets of less than $ 500 million.

C. RANKING BANKING SERVICES

The discussion of banking services in Chapter 7 centers on several basic concerns:
whether branches are located in lower income areas, or in areas with many minority resi

dents; whether the institution makes an effort to respond to the special needs of low -income
people; whether the institution makes an effort to respond to minorities by providing mul
tilingual services .
A Realistic Ranking Scheme.

MEASURES OF BANKING SERVICES
( 1 ) Does the institution offer a " lifeline" deposit account to its elderly low

182

394

areas within its market area ?

( 8) What are the net change in branches ( the difference between branch
openings and

A “ lifeline” deposit account, in light of the survey results from Chapter 7, is defined
as one that can be opened with an initial deposit of $ 100 or less, with no minimum amount
to avoid monthly fees .

These quantitative measures may not perfectly capture the qualitative efforts of a

depository institution to respond to the banking service needs of low and moderate income
residents of their market area . Any institution that is making a good-faith effort, however,
will obtain a positive score for most of these measures .

D. RANKING ECONOMIC DEVELOPMENT LENDING
The discussion of economic development lending in Chapter 8 centered on several basic

concerns: the availability of credit of several different kinds ( start-up, working capital,
expansion) for small businesses, especially those located in low /moderate income areas

and / or in high -minority areas; the availability of " special" credit programs for minority
owned and women -owned businesses; and the level of credit flows to small businesses in

low / moderate income areas and in high -minority areas, us a proportion of total lender
business loan flows.

A Realistic Ranking Scheme.

183

395

MEASURES OF ECONOMIC DEVELOPMENT LENDING

( 1) Does the institution conduct an annual credit needs assessment that

Several points of clarification are in order. A lender's participation in special credit
programs for economic development can encompass lending to non -profit organizations
active in this area . The language concerning minority or women owners who reside in
the City is intended to screen against the use of this provision to benefit absentee owners.
The provision of counseling or other economic development services in lower income areas

should be restricted to facilities or services specifically targeted for this purpose, on the
basis of either programmatic design or geographic location. " Microloans" can be defined

as loans whose dollar amount is in the $ 500- $ 10,000 range. These loans, as discussed in
Chapter 8, are frequently difficult for small businesses to obtain .

184

396

either from CRA statements or from a survey distributed to lenders.
E. RANKING AFFORDABLE HOUSING LENDING
The discussion of affordable housing lending in Chapter 9 investigated every phase

of financial institutions' involvement in development projects for affordable housing in
Los Angeles. Several types of financing are required during the course of these projects; of

these, predevelopment and permanent financing were both found to be less widely available
than construction financing. The bank marvey also reported on lenders' involvement with

affordable housing projects, including their participation in special credit programs.
A Realistic Ranking Scheme.

Here, as for the two previous sections, a quantitative list of performance measures is

presented. This list measures performance in financing new or rehabilitated housing units,
not in financing exchanges of existing units. The latter is covered by the ranking scheme
for residential lending in Chapter 5.
MEASURES OF AFFORDABLE HOUSING LENDING

( 1) How many units of affordable housing financed by this institution were

185

397

( 5) Is permanent financing available for affordable housing projects?
( 6) Ls the institution involved in special credit programs aimed at construct

There is a fundamental difference between measuring banking services and measuring
lenders' economic development or affordable housing activities. Questions about banking
services pertain to branch locations or services that, once established , are consistently
available to customers ( until the institution in question makes a policy decision to close

branches or discontinue certain services ) . By contrast, questions about lending for eco
nomic development or affordable housing pertainto activities which a lender may or may
not engage in during a particular time period. As a consequence, greater specificity is
desirable for assessing institutional performance in the areas of economic development
and affordable housing. There is a difference between making predevelopment financing
( measure ( 3) above) available in principle and making several predevelopment loans to
affordable housing developers in a given year. List D - 1 does not contain great detail for
any measure, but additional detail may indeed be appropriate.

186

398

F. IMPLEMENTING A RANKING SYSTEM

Implementing a ranking system for financial institutions involves several steps, which

are linked to several basic decisions about the scope and purposes of this system . In
outlining these steps, it will be convenient to suppose the City's activities are conducted
by a reinvestment office, and to refer to the program for which this ranking is performed
as Los Angeles' linked banking program ( see Recommendations 1 and 2 in Chapter 11) .
Indeed, this report has suggested a variety of activities that might be conducted by the City

in this area; these activities would be most effectively undertaken if they were coordinated
by a single office. This leads directly to another of the study's recommendations.
RECOMMENDATION : The City's reinvestment activi
ties should be coordinated by a Community Reinvestment
Officer or Unit.

1. Selecting the financial institutions to be ranked. The City must define which

institutions operating in Los Angeles should submit information to its reinvestment office .
At a minimum , any financial institution seeking to participate in the City's linked banking

program should provide the data required for a ranking. A more extensive ranking would
result if the City invited all financial institutions to submit the requisite data and be
ranked , even if not participating in its linked banking program . Universal rankings could
be done on the basis of complete information only if the City relied purely on publicly

available information ( that is, on HMDA data and CRA statements) in its measures of
bank activity .

• To encourage full disclosure of the information requested, the City's reinvestment office

should award a 'O' score whenever participating institutions fail to discloseor submit
information.
187

399

in the reinvestment area would pay large dividends. Inducing institutions to participate
will be easier if there are incentives of more than one kind. Deposits represent only one

part of the City's banking business; the City has numerous other financial dealings and
transactions. Making its program a linked banking program , not just a linked deposit
program , would increase the incentives for financial institutions to participate.

RECOMMENDATION : The City of Los Angeles should

consider making this program a " linked banking " program
encompassing all City financial business.

9. Deciding on the scope of the Ranking. The City must also define the set of activities
over which ranked institutions should be evaluated . One approach would be to restrict the

ranking system to residential lending performance, as suggested in Chapter 5. Another
approach would be to incorporate not only residential lending, but other categories of
bank activity, as suggested in sections C through E of this chapter. This broader approach

has been used in the ranking systems developed for both the State of Massachusetts and
City of Boston linked deposit programs. If a broader approach is adopted, the primary

consideration in constructing an overall ranking is how much weight to give each area .
This is ultimately a question of which social goals are most important, which must be
decided in this case by the elected officials and civil servants of the City of Los Angeles.

each participating institution .

HMDA data are available directly from the federal government. However, any HMDA
data provided by the federal government is at least 15 months old . Just before a new
HMDA tape is received, the data for the most recent year are from 26 to 38 months old.
This problem can be addressed only if the City solicits the most recently submitted HMDA
188

400

data directly from each participating lender. This solves the timeliness problem , but at

a potentially onerous administrative cost ( for the City government) . One compromise
solution might be as follows: each participating lender would be told that the City's normal
procedure would be to use the HMDA data contained on the most recent tape released
by the federal government; any lender could substitute, for this, more recent HMDA data

( that is, its data as submitted for the next calendar year, but not yet released by the

federal government) , on the condition that this data be provided in a standardized format
specified by the City.

The next data source is lenders' CRA statements. In principle, these statements can
be used to provide at least some of the data needed for constructing the 'realistic' measures

suggested above. The drawback of CRA statements as a device for evaluating lender per
formance is that they are non -standardized, and contain data of widely varying quality and
detail.' CRA statements would be more useful in analyzing lender performance if they used
a standardized format and contained more standardized information about institutional
activities in areas other than residential lending. A recommendation in Chapter 8 has
already urged Institutions to release detailed information on their non - residential loans,
particularly in the area of economic development. Another innovation in bank reporting
would be for banks to adopt a standardized reporting format for their CRA statements.

This would not increase the amount of information that banks disclose, but simply orga
nize it uniformly for clearer interpretation by the public. This innovation would be best
coordinated by federal regulators. This leads to a recommendation of the study.

The marginal cost of submitting HMDA data directly to the City, instead of to the federal
government alone, would be slight for institutions, since they already prepare and transmit
these data .

' Two examples demonstrate this point. First, many CRA statements do not distinguish
between finance that is supplied for affordable housing development, and thatfor other
( thatis, non -affordable) housing development. Second, recall from chapter 7 that only a
handfulof Los Angeles lenders will makeloans to first-time home-buyers. This statistic
may be true; or it may simply reflect widelyvarying standards in responding to the CRA
statement requirement that all types of credit offered be listed .
189

401

RECOMMENDATION : The City should intercede with

federal regulators to improve the enforcement of federal
laws on financial institutions' performance and report

ing. A standardized format for CRA statements would
enhance public understanding of lender activities.

The third data source is the City's own data request to lenders participating in its
linked banking program . A sample form for soliciting the data needed for this chapter's
proposed ranking scheme ( that is, for the system proposed in lists B - 1, C - 1, and D -1)
is attached to this chapter as an appendix. This sample form contains only the minimal
information required for ranking purposes. The City could also ask questions beyond those
immediately pertinent to constructing performance measures. For example, the City might

ask for lenders' reactions to proposed special credit programs.

5. Grouping institutions by cohort. Once data are in hand, the City will then have
to construct rankings. It might be useful to group institutions by " cohorts" —that is, with
similiar types of institutions. Chapter 5 suggested dividing lenders into cohorts based

on the dollar volume of their residential lending. Two considerations might guide the
construction of cohorts for an overall ranking. One obvious factor, emphasized in Chapter
5's cohorts, is an institution's size. This affects the scale of lending filows in any credit

categor ; the bank supplies. A smaller size also means fewer branches and fewer deposits

across which costs can be spread; and it means less sophisticated computer systems and
more inflexible information systems. Taken together, these factors have two implications.
First, diseconomies of scale may work against small institutions providing services that
are profitable or break -even for larger institutions: for example, special deposit accounts,
advertising, and personnel specializing in particular kinds of credit. Smaller institutions
are simply noton the same competitive footing as larger ones in providing banking services,

because they cannot exploit economies of scale and scope to the same extent. Second, the
costs of complying with any given information request by the City will be greater on a
190

402

dollar-per-asset basis ."
The second consideration relevant for constructing cohorts is specialization by type of

credit. Bank survey results and CRA statements showed that lenders making residential
loans have many different patterns of specialization a large number of thrifts simply do

not make business loans, while numerous banks do not lend for housing construction.
Obviously , an institution which specializes Darrowly will have very low scores on areas

used in the ranking which fall outside its areas of specialization .

their peers makes for fairer comparisons. The key questions are how and whether to com

pensate for differences in size and specialization in ranking performance. These questions
are impossible to sensibly address before a ranking system is actually implemented. The
City's reinvestment office will have to handle these issues as they arise, keeping in mind
the program's goals.
6. Using, publicizing, and updating the rankings. Once the data are in and cohorts
have been grouped , the City reinvestment office must, to retain credibility, use its ranking

as it has promised. Applicant lenders must be sorted into cohorts. The share of deposits or
of the City's financial business received must be linked to institutional performance. In the
case of deposit awards per se, a minimal deposit level must be set that is large enough to

attract participation by lenders. The City's rankings should be publicized. As an integral

part of using any set of rankings, the City should make clear to all financial institutions
when new rankings will be established, and what steps they can take to improve.

* In our interviews with bankers for this study, we found that small banks divided their

creditportfolios into fewercategorical areas than did large

191

403

Chapter 11
Policy Recommendations
A. INTRODUCTION
The expressed goal of the City Council of Los Angeles in approving this study was to

" encourage financial institutions to help meet the credit needs of their comunities , including
the needs of low and moderate income neighborhoods." As a way of implementing this
broad goal, the City asked that we investigate the advisability of a linked deposit program
Specifically, this involved determining whether such a program is needed, and if so , whether
a linked deposit program could be an effective means of achieving that goal.
This chapter encapsulates our policy recommendations. These have been foreshad

owed in the previous ten chapters, which document the need for changes in the pattern
of credit flows and banking services in Los Angeles. We recommend first that the City

government establish a linked deposit program in Los Angeles. Indeed , we recommend
that the City extend its program to encompass a broad range of its banking business, not
just its deposits. As a component of this program , we call for the City to monitur financial
institutions' performance, and to update and publicize new rankings of institutions on a
regular basis.

While we advocate the creation of a linked banking program in Los Angeles, simply
adopting such a program will not alone address the credit and banking needs of Los Ange
les' lower income and minority neighborhoods. A truly effective linked banking program

involves additional policy initiatives. But beyond any specific policies, the success of any

City initiatives in addressing credit and banking needs depends on the City government
of Los Angeles taking a leadership role in advocating reinvestment by financial institutions

in the needy and underserved areas within its borders. This point is so vital that we make
* This quotation is taken from the ordinance authorizing this study, passed by the Los
Angeles City Council on August 24 , 1988 .
192

404

it our overarching recommendation of the report. All later recommendations are steps to
I

fulfill the goal of City leadership .

RECOMMENDATION : The City of Los Angeles should
play a leadership role in encouraging reinvestment and
socially responsible behavior by Los Angeles' financial in
stitutions .

Discussion in Chapter 1 makes clear that municipal and state governments are now
on the front lines in confronting the housing and economic crisis that grips urban America .

Over the past several years, the federal government has surrendered its leadership role in

the wars on poverty and on urban decay. For now , the renaissance of cities like Los Angeles
depends on local governmental initiatives.
A major problem confronts all efforts to turn around urban neighborhoods - negative

spillover effects in neglected areas are large and pervasive, and can overwhelm even well
meant but unfocused investments of time and resources by government or the private
sector. Overcoming negative spillovers requires focused efforts combining public - sector
resources , private sector reinvestment ( particularly by financial institutions) , and neigh
borhood involvement. This lesson — that focused reinvestment is the key to successful
reinvestment - is embodied in the successful ( and profitable) activities of the South Shore

National Bank in Chicago. Among interested parties, the City itself is best positioned to
take the lead in these efforts: the spillover benefits from urban reinvestment and neighbor

hood revitalization will accrue to Los Angeles itself, both in enhanced tax revenues and in
an enhanced quality of life.

encompassing several recommendations.

• First, the City can demonstrate this leadership by adopting and admin

193

405

credit flows and banking services to areas and residents of Los Angeles

that are presently underserved ( RECOMMENDATIONS 5-8 ) . TheCity
can encourage financial institutions in Los Angeles to experiment with

B. A LINKED BANKING PROGRAM
RECOMMENDATION 1: The City ofLos Angeles should
implement a linked deposit program .

Linked deposit programs have been implemented in a variety of forms by a number
of municipalities and states. The reasoning behind a linked deposit program is straight
forward . Preference is given in allocating public deposits to lenders whose activities are

more heavily concentrated in specific areas and activities deemed to be publicly valuable.
most likely to use them in desirable ways; and they establish incentives for other lenders to
shift towards more socially desirable activities, in order to qualify in the future for benefits

under the program .
Normally a linked deposit program involves the commitment of a pool of funds on an
ongoing basis. Financial institutions then bid for the right to receive these deposits. These
institutions' incentive is that they will pay market or below -market interest rates.

194

406

be barred from participating in the linked deposit program , and excellent lenders would
receive higher levels of deposits than simply good lenders.
Elements of A Successful Linked Deposit Program .

A successful linked deposit program must establish explicit, objective links between
lender performance, eligibility to participate in the program , and the amount of deposits
that eligible institutions receive. Studies of other linked deposit programs have found that

the essential features of a successful linked deposit program are :

While these organizational aspects are necessary for a successful program , they are

not suficient in and of themselves to guarantee the program's success . Experiences of
other linked deposit programs across the country suggest that a number of pragmatic con
siderations also should be considered in designing a linked -deposit program . Specifically:

2. Rates offered participating institutions. Strong reinvestment performance
should be rewarded with public deposits that are attractive to participating insti
tutions. Deposits of less than sixmonths have little incentive value to lenders, par
ticularly in light of the collateral requirements that apply to deposits of the City

of LosAngeles. Payment of below market interest rates, especially on shorter
term deposits, will increase incentives. If the deposits offered are long -term ,
below -market interest rates need not be paid .
* The following list is drawn from James T. Campen, " The Political Economy of Linked
Deposit Banking Programs," Mimeo, University of Massachusetts, Boston , January 1991.
195

407

3. A level playing field ” for competition . To achieve a " level playing
field ” for ranking -point competition, financial institutions should be grouped and
ranked according to the size and type of institution .' Credit unions, especially
community based institutions, should be includedamong the qualified competi
tors for deposits. Rankings should not be influenced by past depository or con
tractual relationships with the City .

Criteria for Ranking Institutional Performance.

The heart of any linked -deposit program is its ranking method . We recommend that
any formula be made public, that performance pursuant to that formula be well defined,
and that data be collected in a standardized format. Beyond this, the City of Los Angeles

itself must determine which areas should be included in its ranking scheme - residential
lending, banking services, economic development lending, or lending for affordable housing
construction. Similiarly, the City must decide what measures to use in these areas, and
what relative weights should be assigned to the different areas within the ranking scheme.

Different jurisdictions in the country are taking very different approaches to their
ranking formulas. Some base institutional scores on residential lending performance alone;

others also factor in performance in other lending areas and in banking services. Each
' The City of New York's 1990 ranking of banks competing for its financial business was con
troversial because it grouped banks of disparate size and of diverse portfolio composition.
* Neither shouldrankings be influenced by the financialcondition of applicant lenders. Poor
financial condition could, of course, constitute grounds for denying i financial institution
the opportunity to do business with the City. If it has passed the City's requirements
concerning its financial condition,however, and is eligible to compete for the City's linked
banking business, then an institution's financial condition shouldnot be part of the system
for ranking performance.
196

408
jurisdiction must decide what substantive areas are most important.

Chapters 5 and 10 of this report bave proposed ways of " ranking " financial institutions

in the four areas mentioned above. These chapters suggested a " Yealistic " ranking system
based on these measures:

1. Residential lending performance ( Chapter 5)

4. Affordable housing performance ( Chapter 10 )
• Predevelopment financing for affordable housing
• Construction financing for affordable housing
• Permanent financing for affordable housing
• Involvement in special credit programs
• Involvement in bond financing programs
•

Involvement in government guarantee programs
197

409

The residential lending component of this ranking system can be implemented im
mediately using ELMDA data. For some of the measures in the other three components
of this system , some additional data must first be solicited from applicants. This system

is realistie ” in the sense that it seeks a balance between measures lending insight into
reinvestment activities, on the one hand, and lenders ' costs of providing this information,
on the other.

of redlining and encourage reinvestment. Financial institutions would be encouraged to
lend in neighborhoods, and for credit needs, that have been underserved historically.

RECOMMENDATION 2 : The City of Los Angeles should
consider making this program a " linked banking " program
encompassing all City financial business.

One important aspect of designing an effective linked deposit program is insuring that

there is a healthy representation of lenders when the City opens its competition. If the
City properly advertises the program and announces in advance its intention to publish

rankings, considerations of publicity alone will probably induce some lenders to participate.
Otherwise, lender participation cannot be taken for granted .
The bank survey posed several questions for respondents about doing business with
198

60-893 O - 92 - 14

410

the City. Lenders were asked separately if they were interested in being a depository for
the City's short- term and long -term deposits. With respect to short- term deposits, 54 % ( of
48 respondents) indicated no interest, and 17 % indicated interest only with below -market
rates, while just 29 % were interested without interest- rate concessions . For long -term

deposits, the pattern shifts toward greater receptivity: 43% ( of 57 respondents) indicated
no interest, and only 13% required below -market rates; the remaining 45% were interested
without below -market rates .
Banks were asked to indicate the drawbacks to being a depository of the City of Los

Angeles. Cited most often as the biggest drawbacks were " too much paperwork " and " oner

These responses indicate that the City would mazimize its pool of participating insti
tutions if it made available longer-term deposits at competitive ( or below -market) rates,
or short - term deposits at below -market rates, while also cutting red tape and increasing
administrative efficiency for participants.

Another way to insure healthy competition is to broaden the pool of services opened
for linked -performance bidding to encompass all business relations between the City and

financial institutions. Beyond deposits per se, the financial transactions and services that
might be included are fees on the City's daily transactions and its sweep accounts, fees from

bond issues and rollovers, and dedicated accounts, among others. Including the financial
transactions of independent municipal corporations, as well as those of the City itself,
would add further to the attractiveness of participation.
We recommend that the City of Los Angeles adopt an ordinance that links the de

posits, contracts and other dealings of the City government, the Community Redevelop
ment Agency, and independent municipal corporate entities with financial institutions to

acceptable performance levels as specified by the City.
199

411

Financial Risks from a Linked Banking Program .

The City's financial risks from initiating a linked banking program should be unim
portant if the program is well run . " Financial risk ” in this context can be understood as
involving two distinct items: first, the danger that the City will either lose funds or tem

porarily love access to funds due to financial institutions' failure or reorganization; second,
the danger that the City will fail to maximize its interest income by engaging in a linked
banking program .

The danger of losing funds deposited in financial institutions is effectively zero under
current federal law . There is some danger of temporarily losing access to deposit funds due
to lender insolvency problems. However, unless the City commits a substantial proportion

of its liquid funds to a very small number of institutions, this should have minimal effect
on the City's ability to conduct its financial business. It is certainly sensible for the City
to accept linked -banking business only with institutions that are expected to be solvent
over the horizon of involvement. But once that hurdle is passed, there is no reason to
believe that financial institutions engaged in socially responsible lending are more prone
to bankruptcy than others . For example, the thrifts that failed during the period 1981-88

were no better and no worse in their HMDA lending performance than thrifts that did not
fail over that period.

The second type of " financial risk ” comes into play only if the City offers linked
deposit funds at below -market interest rates. As noted above, this inducement is more
important for short- term than long - term deposits, especially if the City can find ways to

cut the paperwork and collateral requirements it imposes on participating institutions.
In any event, lost interest income from below -market rates on some City deposits, if it
occurs, should be dwarfed in the longer run from the tax -base and employment benefits
generated by boosting Los Angeles lenders' reinvestment activities . In this event, City

decision -makers must determine whether short-run or long- run considerations weigh more
heavily.
200

412

RECOMMENDATION 3: The City should monitor lender
performance closely , and update and publicize its rankings
regularly.

An effective linked banking program must have an ongoing reporting and monitoring
component, so that financial institutions have an incentive to improve their performance.

Marginal lenders have an incentive to upgrade their performance if their improvements are
recognized and publicized, and provide them with entree into the City's linked banking

program . Similarly, lenders who are currently part of the program will need to maintain

or improve their performance to remain eligible in the future. There is obvious mutual
reinforcement between maintaining a healthy pool of program participants, on the one
hand, and ongoing monitoring, on the other.

For instance, a recent survey of people'sfeelings about thesocial responsibility of finan
cialinstitutions in Massachusettsfound that two-thirds ofthe respondentswould change
financial institutions if they became aware that their own institution'spractices were not
socially responsible. This survey is discussed in Campen, " The Political Economy of Linked
Deposit Banking Programs ," Mimeo , University of Massachusetts, Boston , January 1991.
201

413

with a substantial time lag. For example, HMDA data for 1989 became available only in
March of 1991. The program we designed and delivered to the City contains a facility for

opdating the HMDA information each year, but the data will still be more than one year
behind.

At present, there are two ways to fill this gap. The first is to require lenders participat
ing in the linked service program to provide their most recent AMDA -reported information
directly to the City. These data would then be for the most recent calendar year. If these
are requested in standardized form on diskettes - for example, in Lotus 1-2-3 files, using
a given format - the City could readily use this information for analytical purposes. A
second method of obtaining more timely information on residential lending is to acquire

title transfer data from one of the firms that compiles and sells it to users . Until recently,
it has been difficult and costly to use title transfer data services to pull together aggregate
lending information . But this is no longer true. Title transfer data containing information

about real-estate transactions are now available with a lag of several months or less .

lending. Updating these data involves two steps. First, the City could require participat

ing institutions to submit their most recent CRA statements. Like HMDA data solicited
directly from lenders, these would report on the previous year's activity. Second, the City

could send along a brief form as part of the application materials sent to institutional
participants. This form could be made relatively non -intrusive: it would have to include
only the data needed for the ranking scheme itself.

The City might consider several steps to facilitate this portion of its monitoring activ
ity. In particular, it might request that information from CRA statements be submitted

A further advantage of the title transfer data base is that it encompasses extensive infor
mation on real-estate transactions involving commercial and industrial properties. These
data could supplement the economic development component of the ranking system by
providing new insights into the relationship of lenders and businesses.
202

414

in a standardized format. This would facilitate analysis and insure comparability among
institutions. It would also be sensible to suggest that, where possible, these supplemental
data be provided in a standardized format on computer diskette. Finally, the City might
insure the accuracy of this information by either conducting its own random audits of
the information submitted, and /or by checking the accuracy of this information with the
appropriate federal or state regulators of participating financial institutions.

monitoring lending and service activity, and thus of identifying and rewarding the socially
conscious financial institutions that are active in reinvestment activities. The information

collected could find many uses within City departments as they identify and target areas
in need of special attention in terms of banking services and credit flows.
RECOMMENDATION 4: The City's reinvestment activi
ties should be coordinated by a Community Reinvestment
Officer or Unit .

The implementation of a linked banking program requires sustained ongoing effort
and resources . Further, as envisioned in RECOMMENDATION 2, this program would
encompass several distinct activities: the City's investments and deposits; its housing
construction, rehabilitation, and bond programs; and its economic development efforts.
These activities may extend beyond the boundaries of any one of the City's operating
units. Therefore, we suggest the City designate a Community Reinvestment officer or

unit, with several responsibilities. This unit will be responsible for administration of the
linked banking program . As such, it will collect information from financial institutions and

others, as necessary, and use this information to rank lenders, to monitor and publicize
ongoing lender performance, and periodically to issue new rankings.

203

415

for both financial institutions and community groups. It might work with lenders to de
velop innovative ideas for meeting credit and banking needs in low /moderate income and
high -minority communities. It might also provide technical support and information to
non -profit advocacy groups interested in achieving reinvestment and affordable housing
goals. It might also serve as a consultant to, or even representative for, the City govern
ment in negotiations with financial institutions or public regulators on substantive policy
questions.

C. INNOVATIVE IDEAS FOR REINVESTMENT

This second set of recommendations proposes innovative ideas for reducing inequities
in the distribution of banking services and credit flows. These recommendations go beyond
the implementation of a linked banking program . They are included here both because they

emerged in our discussions with City officials, lenders, and residents, and because they have
promise as tools for reinvestment. The City of Los Angeles could aptly demonstrate its
leadership role by encouraging financial institutions to sponsor and experiment with these
or other innovative reinvestment policies.

RECOMMENDATION 5: The City should work with
lenders to fund loan pools as appropriate mechanisms for
some institutions to meet credit needs in low /moderate in
come and high -minority communities. These pools might
target first -time home buyers, affordable multifamily hous
ing, and small businesses.

The selective use of loan pools as a means of achieving reinvestment goals is suggested

by several findings of our study. One of our findings is that many small lenders operate

within the City of Los Angeles. These small lenders typically have extremely small pro
fessional staffs and restricted market areas spanning only a portion of the City. In some

cases , small institutions' market areas fall almost completely outside the geographic areas
204

416

of Los Angeles with lower median income and high proportions of minority residents .

A loan pool with participation by certain lenders could assist such small institutions

in meeting their CRA responsibilities for reinvestment in two ways. First, the pool would
be operated by loan officers and other professionals with expertise in the specific activ
ity areas of the pool, including start -up credit for small businesses in low and moderate
income areas , financing for affordable housing construction , and so on . This centralized
expertise substitutes for a lack of experience by small institutions with risk - taking in these
areas of credit need. Second, loan pools would provide indirect mechanisms for socially
responsible lending by smaller institutions with market areas that do not substantially

overlap with lower income or high-minority areas of the City .' While their market areas
may not encompass underserved areas, these institutions have social responsibilities un
der CRA and other contemporary legislation . Making well-advised reinvestment loans in
underserved areas without knowledgeable staff could be hazardous; loan pools could effi
ciently intermediate between such institutions' resources and credit needs in underserved
areas.

Loan pools should not be a substitute for good -faith performance in credit and banking

services by institutions that do or should have the capacity to fairly assess the risks of,
and make loans to, residents and businesses in underserved areas of the City. In general, a

healthy credit market — like any other — is one in which numerous suppliers and numerous
demanders confront one another in the marketplace. Funneling all credit supplied to low
and moderate income areas ( or to all high -minority areas) through one or two suppliers

will not create a healthy, thriving credit market in these areas: it maintains the idea that
these areas that are inappropriate risks for regular credit flows.

* The CRA itself specifies that financial institutions' regulators must determine whether
their market areaboundaries have been drawn in a way that unfairly excludes geographic
areas that are low /moderate income or that have a high proportion of minorityresidents.
* Indeed, the City may wish to credit loan poolparticipation only if the lender in question
meets certain criteria , such as being small, not supplying loans in that area , and having

an appropriately delineated market area that overlaps minimally with lower income areas .
205

417

RECOMMENDATION 6: The City should advocate the

establishment of multi-bank branch offices serving low and
moderate income areas within the City .

A central finding of our investigation of banking services ( Chapter 7) is that bank
branches are disproportionately located in apper income areas and in highly-developed
business districts within the City. As a result, the nearest bank branch for many low

income people may be literally miles ( and a bus ride) from their residences. Officials
of financial institutions have advanced the argument that locating more branches in low
and moderate income areas is too costly in terms of the dollar volume of deposits those
branches would service. Without investigating the merit of this claim , one solution is
the establishment of bank branches in low and moderate income areas by a consortium

of banks. The notion of sharing facilities is already well-established in the various ATM
networks to which most financial institutions belong. In this case, the facility shared would
be a branch office.

Shared-bank branches might perform limited functions -- for example, transactions
services plus referrals to the nearest loan application facilities of participating institu
tions. However, the availability of these functions could make a major difference for

residents -- and especially economically vulnerable residents in these areas. In addition,
these branches could address one of the problems cited in our contacts with operators of
small businesses the lack of nearby facilities for conducting bank- related business.

RECOMMENDATION 7: The City should encourage fi
nancial institutions to allow non -depositors to cash their
government benefits checks, to directly deposit all gov

ernment benefits checks for depositors, and to offer and
publicize " lifeline " deposit accounts for all low income res
idents of Los Angeles.

Having available a bundle of services that are relevant to the special needs of low
income residents is just as important to those residents as conveniently located bank
206

418

branches . The three services listed in this recommendation are potable largely for their

scarcity or absence in Los Angeles. Virtually all banks in our survey allow elderly deposi
tors to directly deposit their social security or SSI checks; but less than half provide this
service for non -elderly depositors receiving government benefits checks. Further, almost
no institutions allow non - depositors to cash these checks, even though their risk of default
is zero .

The provision of these services would substantially enhance the personal security of
some of the most economically vulnerable citizens of the City. Clearly, this recommenda
tion's benefits increase if financial institutions also locate branches in underserved, lower
income areas , as suggested above. Provision of these services and opening branches in

lower income and minority neighborhoods will have a particularly powerful impact if fi
Dancial institutions take steps to publicize their availability. The City should encourage
the use of publicity by incorporating it directly into its ranking system .
Recommendations 6 and 7 are both instances in which the City can act as an advocate

to encourage particular policies by financial institutions. It can put some teeth into its
advocacy by tying adoption of these policies to points in its ranking system . For example,
the ranking scheme suggested in this report rewards institutions that allow non -depositors
to cash government-benefits checks and that locate branches in low -moderate income areas .
RECOMMENDATION 8: The City should encourage fi
Dancial institutions to sponsor centers in low and mod

erate income neighborhoods offering business counselling
and technical assistance for small businesses operating in
these areas .

In conducting the study, we obtained insights into the problems involved in lending to small
One of the two institutions offering this service is reconsidering its policy. This is per

haps another area in which negative spillover effects may condition bank policy. That is,
financial institutions' reluctance to offer government benefit checks to be cashed by non
depositors may reflect an effort to protect their " image" by controlling for the clientele
that makes use of their branch offices. However, if a substantial number of bank branches

offered this service as a matter of course, this " image" aspect of this policy would be moot.
207

419

businesses both from officers of lending institutions and from operators of small businesses

in lower income and minority areas . The picture that emerges is of a self-reinforcing
vicious cycle which binds potential borrowers and potential lenders and inhibits economic
development in Deedy areas. For example, lenders are generally reluctant to make loans to
businesses that are not thriving. At the same time, businesses in low and moderate income
areas often find it difficult or impossible to borrow . Operating without adequate credit ,

they will be unable to implement business decisions at appropriate times. They will then
fail to thrive, apparently justifying creditors' reluctance to provide them with credit.10

10 Another area inhibited by poor coordination between potential borrowers and lenders is
business expansion . Lenders report that they only make loans to businesses, on average,
if they are fully collateralized. This means that any business that wants to expand using

bank creditmusteither be well capitalized or have excess assets. Ofcourse,such businesses
would not necessarily need bank credit to expand. On the other hand, businesses which
are neither adequately capitalized nor asset -rich, as in the case of most small businesses in
lower income and minority areas, will find it difficult to borrow for expansion.
208

420

D. RECOMMENDATIONS INVOLVING CHANGES

No matter how successful, a linked banking system and other innovative programs in
Los Angeles can be only partially successful in achieving the goals of reinvestment and

equitable banking services and credit flows. The regulatory and statutory environment
within which the City and Los Angeles financial institutions operate is a fundamental
determinant of patterns of banking services and credit flows. Reducing racial redlining

and making credit and service flows more equitable depends on changes in this broader
environment. Efforts by the City to alter aspects of this environment constitute another

means of achieving its broad goal in launching this study, and of succeedingin its future
reinvestment initiatives. Therefore, our last four recommendations call for the City of Los
Angeles to demonstrate its leadership by acting as an advocate for changes at the level of
bank reporting, federal enforcement and state legislation.

RECOMMENDATION 9: The City should encourage fi
nancial institutions to more seriously and more creatively
assess the credit needs of lower income and high -minority

neighborhoods in their market areas .

A better grasp of banking and credit needs is essential if more socially responsible

banking practices are to be more than blanket commitments with ill- understood conse
quences. But needs assessments by institutions have often not provided this grasp. Some
lenders fail to assess community credit Deeds, as required under the Community Rein

vestment Act. Further, it appears that many or most of the assessments now performed
add little or no new information for the institutions performing them . The idea of as

sessing credit needs is widely perceived as an inherently empty exercise ( everyone needs
more money, and especially residents of lower income areas — so what else is new ? ) . But
perhaps what has been lacking in the needs assessment area is a failure of both effort and
209

421

imagination
The City should work to convince lenders to try creative new approaches to map

ping out credit and banking needs: for example, surveys, community meetings, and /or
interviews with community opinion and business leaders. The City might also take a high
profile role: for example, sponsoring an annual community needs assessment in conjunction

with lending institutions ( perhaps, with those participating in its linked banking program ) .
Whatever leadership role is played, the end - product- more coherent understanding of
credit and banking needs — will be useful in several ways. City officials can use this in
formation to urge specific institutions to meet specific needs that have been established .

This information can also be used to refine the measures used to evaluate lenders, and to
encourage their efforts in the areas of greatest need.
RECOMMENDATION 10: The City should encourage
financial institutions to make detailed information related
to their CRA responsibilities available to the public on

an annual basis , in areas other than residential lending.
Particularly useful for economic revitalization would be
annual disclosure of data on the geographic distribution of
business loans, and especially of loans to small businesses.

The provision of affordable housing is linked not just to residential lending flows, but

to economic development processes in the broadest sense, especially in lower income and
minority communities. We have highlighted the importance of banking activities other
than residential lending by including three other components - banking services , economic
development lending, and lending for affordable housing development – into the ranking

system suggested above.

The City's efforts to assess financial institution performance on a fuller basis than
residential lending would be eased if better information about lender activities in these
other areas becomes available. One useful step would be the adoption of uniform CRA re

porting. A standardized reporting form for CRA statements would both facilitate analysis
210

422

of bank activities and insure that a basic level of information was consistently provided.

A particular problem at present, us discussed in Chapters 8 and 9, is the paucity
of available information on financial institutions' activities in lending to businesses and in
lending for affordable housing development. Lenders might be asked to disclose information
annually on :
• The number and dollar volume of loans made annually to businesses located in

• The number and dollar volume of loans made annually to support affordable

• Information about lender participation in City or State programs promoting eco

• Information about special assistance offered by the lender - in the areas of coun

Information of this type is suggested for use in the City's ranking system . But uniform
disclosure by all institutions, not just those participating in a linked banking program ,

would have a real benefit: it would provide the City with an accurate picture of all ( or at
least most) credit flows in these areas. Better planning and smoother coordination of City
projects with ongoing economic and housing development would result.

In evaluating whether to request additional lending information from financial insti
tutions , it should be kept in mind that CRA disclosure guidelines have been tightened

in 1989 legislation ; so lenders are already upgrading their reporting capability to comply
with changes in federal law . Further, the City might consider coordinating its requests to
lenders for more information with federal regulators of lending institutions. This leads to
the next recommendation.
211

423

RECOMMENDATION 11 : The City should intercede with
federal regulators to improve the enforcement of federal

laws on financial institutions' performance and report
ing. A standardized format for CRA statements would
enhance public understanding of lender activities.

Federal agencies that oversee financial institutions under CRA and HMDA have not
required extensive reporting under their regulatory authority. They also have not used the
full force of their regulatory authority to enforce the substantive provisions of federal law .
More forceful oversight would facilitate the City's own reinvestment efforts.

affordable housing development. The City might persuade federal regulators to require

more detailed annual information from lenders in these credit areas. An optimum would
be HMDA -like annual geographic reporting of loans to small businesses and housing de
velopers.

More sustained attention by federal regulators to lenders' economic development per
formance would also be useful. The City might also persuade federal regulators to insist

that credit needs assessments be better done. Finally, the City might consider involvement
in cases of branch closure and lender merger or acquisition when these might materially

affect the provision of credit and banking services in underserved areas of Los Angeles.
212

424

THE COLOR OF MONEY

Home mortgage lending practices

cdiscriminate against ( blacks

The Atlanta Lournal
STHE ATLANTA CONSTITUTION

425

“ The Color of Money," a four-part series, was published May 1-4, 1988,
in The Atlanta Journal and The Atlanta Constitution, 72 -Marietta St. N.W. ,

P.O. Box 4689, Atlanta, Ga. 30302. Telephone 404/526-5151. Publisher, Jay
Smith. Editor, Bill Kovach. Managing Editor, Glenn McCutchen. Circulation:
morning Constitution, 265,262; afternoon Journal, 186,896; combined Satur

day Journal-Constitution, 501,823; combined Sunday Journal-Constitution,
646,904 ( ABC Sept 30, 1987) . The Journal and Constitution are members of

the Cox Enterprises group. Follow-up articles included here were pub
lished as of May 13, 1988. For additional copies of this reprint, please con

tact the Journal-Constitution Marketing Department at 404/526-5690.
Part 1

Atlanta blacks losing in home loan scramble
Part 2

Southside treated like banks' stepchild ?

.... 11

Part 3

A test that few banks fail – in federal eyes

.... 21

Part 4

Bank protesters in Atlanta make ready to flex muscle

.... 27

Part 5

Follow-up and reaction

.... 37

426

Journal
AtlantCOa NS
The LA
TITUTION
THE AT
NTA

Part 1

Sunday, May 1 , 1988

Atlanta Hectus beingin hare en sort FMDuets

old 777

Atlanta blacks losing
in home loans scramble
Banks , S & Ls favor

city's white areas
by margin of 5-1

As part of a five-month examination

'We're talking about disinvestment, capital flight from the

executive officer of Bank South. “ You
A federal law, the Community Reincan prove by the numbers that the At- vestment Act of 1977, says deposit- gathlanta bankers are discriminating ering institutions have an affirmative
against thecentral city. It's not a willful obligation" to solicit borrowers and dething. The banks really are considered positors in all segments of their
the pillars of the community. If some

communities.

of compliance with the Community Re
investment Act, the Journal-Constitu

cents of each dollar deposited by whites
in home loans to white neighborhoods.
• The offices where Atlanta's largest
banking institutions take home loan ap
plications are almost all located in pre
dominantly white areas . Most savings

427

Scramble
From Page 1
white areas
• Lenders are not required to dis
close information on loan applicants by
race. However, two of the largest lend
ers volunteered that information, which

showed they rejected black applicants
about four times as often as whites.

'It's institutional racism '

" Since the 1950s forward , we've had

a substantial and identifiable black
middle class.

family structures in white neighbor

428

le nul que's.
Fucking Millinds. "
tioning the accuracy of the lending rig.
ures, offered a variety of explanations

ROC

KDA

Some bankers cited the aging of

for the differences.

429

lines bullieen banks and savings and banking is so intimately connected

Scramble

loans. Banks across the country are do

with the public interest that the Con

Ranking lenders on black vs. white loans
Comparing lending to middle -income neighborhoods
Black :White
Rank

Institution

How the rankings were determined :

430

and the lowest of any bank its size in

black nung!!.. huds. ülitut in lee Lower-income also affected

the country in 1986, according to the
Federal Financial Institutions Exami-

loans

In middle - income black neighbor

Institution

Rank

Score
0-100

Citizens Trust Bank *

2.
3.

Mutual Federal Savings and Loan *
Liberty Federal Savings and Loan
Anchor Savings Bank
DeKalb Federal Savings and Loan
California Federal Savings and Loan
First Federal Savings and Loan

4.
5.
6.
7.
8.
9.

First Union Bank

Decatur Federal Savings and Loan

10.
11 .

Home Federal Savings and Loan

12 .
13 .

Georgia Federal Bank

14 .

Fulton Federal Savings and Loan
Trust Company Bank

15 .
16 .

17.

C& S Bank

Bank South

First American Bank

First Atlanta Bank

How the rankings were determined :

431

Scramble
From Page 5
have turned to the banks. And Atlanta

has some of the most profitable banks
in the country
Last year First Union Corp. of
North Carolina and SunTrust Banks of

Atlanta, parents of First Union and
Trust Company respectively, led all U.S.
banks in net income. First Atlanta, C& S

and Bank South have consistently been
in the top half of their peer groups na

tionally in profits, according to the fed
eral examination council.

432

them $ 5.773 69 at 18 percent interest
plus $ 3.180 in " discount points " and
other add -ons, raising the effective in
terest rate to 27.1 percent, according to
the loan papers.

.
Fulton's Michael Lomax: ' If I can't

get a loan , what black person can? '

that turned him down, but he did say

banks last year to get a loan to add a
Fletcher went in 1984 to Citizens one of them was where he had banked
guesthouse in Adams Park, an upper and Southern Bank ( C& S) , his bank for for 25 years.
middle -class black neighborhood in 10 years.

433

Strici

! .. Lib.vitit

Lomax

H
HII

E niin

18
/

'The first reaction from the bank was, " Why do you
want to invest that much money in your neighbor

hood ? ' ”

MICHAEL SCHWARZ/Staff
recalled Michael Lomax. But that's the

neighborhood my house is in . '

434

ANDY SHARP /Staff

In Gresham Park ( left) , Elijah Blount, 7,pauses Blount, washes the car. In the McLendon area
during a bike ride while his father, Clifford ( right) , chiropractor John Dull rakes his yard.

Two neighborhoods: One black, one white
By Bill Dedman
Staff Writer

'They always find the smallest

Advisory Council
borhoods. The residents of Gresham

Journal -Constitution.

it," said Jan Reese, whose " for sale "

435

the il

Contrast

Tie irriters

" ...

*There have been people who

)

A look at Gresham Park, McLendon
Gresham Park residents in south DeKalb eam more money, are better educated and live in newer homes than
residents of the McLendon area in north -cental DeKalb- but they receive one eighth as many of their home-pur

chase loans from banks and savings and loans. One possible reason: Most residents of Gresham Park are black.
ul

Gresham McLendon

d
HillsAd .

.
t
ot lvd
B

l
McLendon
ia

Ponce de Leon Ave)

m
Me

Gresham

.
Rdr
dle
Can .
S

285

Note : Loan figures are for 1986-87 . Amounts for home
purchase loans by banks and savings and loans do not

include unregulated mortgage companies owned by them .

Sho

Fla

als
20

Sources: Loan data from Real Estate Data Inc. , which
compiles information on real estate sales from county
records. Demographic data from the U.S. Bureau of
the Census , 1980.

436

THE VITA CONSUTI TION

Part 2

Monday, May 2, 1988

Nuhtim

ined by

banks'

ptur

U.S. planning

Southside treated

to hike forpign
weapons wile

like banks ' stepchild ?

ROC

KDA

By Bill Dedman

437

Souiiside
From Page 11

438

:: :: ‫ نه؛ نه د مزمن‬:
torniood

not

that ü

from the Northside. He called me Blacks rejected more often

don't refer homebuyers to banks and

CLOSED
NOTICE

LOSED
The East ants Brands
has been comb
with our Lakewood Bran

011

We invite you bransadabisa

T
FRANK NEIMIER /Staff

Trust Company closed its East Atlanta branch, although it had more deposits than some other branches.
13

439

Southside

Where biacl : s bank

Based on 1.461 telephone interviews

440

.: 10

rritu

! u ! committee of the Imerican Insti:

lute of Real Etale Ippraisers He
might attribute it to something more
wearing out of the public in
bland
something that isn't
frastructure
such a dangerous area . "
l'nderappraisals – called lowballs

Willie Clyde lost a chance to sell this
53 Bisbee Avenue home, which he
remodeled, because it was appraised
at $ 10,000 less than the contract
price.

M
MICHAEL SCHWARZ/Staft
15

441

Southside
From Page 15
1

percent in black Cascade Heights. 84
percent in white Sherwood Forest, but
26 percent in black Adams Park: 71 per
cent in white Peachtree Hills, 6 percent
in black Collier Heights.
Low appraisals also deter further

investment in a neighborhood by cur-,
rent owners. The amount of a home-im
provement loan is usually limited by the
owner's equity - property value in ex
cess of debi. If the value of a home
drops, there may be no equity left for a

Trust Company

home-improvement loan. ,

Lower- income also affected

T24
16 - O
Yana

60-893 O - 92 - 15
-

442

don
SET

3. FETTUCE GREEN

200Z

HEAD

ANDY BI

KENNETH WALKER /Stall

Irvin Betts inspects a box of lettuce at his Forest million or $ 15 million in business if I had the
Park produce company. 'I could have maybe $ 10 working capital,' he says .

He built $ 3 million lettuce firm

but can't get a business loan
By Budi Dedenan
svafWriter

to admit it's still a problema ."

A federal law , the Community Rein

443

Belts

Srall -business lending

71
66
64

444

Poor may be left behind by bank deregulation
By Bill Dedman

mir

Traren

445

Pickets target HUD secretary's residence
By Bill Dedman
Staff Writer

RICK MCKAY/ Cox Washington Bureau

Gale Cincotta of Chicago bangs on the door of his high-rise building in Washington. The protest
Housing and Urban Development Secretary Samu.

ers, who were denouncing reductions in federal

el Pierce after she and other protesters entered housing funds, got no answer.
20

446

The Atlanta Journal

Part 3

Tuesday, May 3, 1988

Fast boy used astrology be
an

A test that few banks

fail - in federal eyes
By Bill Dedman
StaffWriter

The regulators have been haphazard . There's a pattern of the

447

Test

Activists say the law doesn't al

448

70
THEY'RE a pain in the neck,' saysFirst Union's Edward Crutchfield of challenges from community

' IT'SNOTourjob to allocate the credit geographically,'
says Martha Seger, a governor of the Federal Reserve

groups .

Board.

of the CRA at the same time. Several

449

Despite its 'good citizen' image,
Trust Company finds itself in battle
By Bill Dedman
StuffWinter

450

RICH ADDICKS/Staft

'I think we get a bad rap,'said Jim Mynatt. " We don't have a map up there with a redlined district.'
challenged a larger merger in 1986 that ing, less transient areas. It offers only million dollars my way since the negoti
brought Tennessee's Third National adjustable-rate mortgages, which ap- ations began," said the land trust's
Bank Corp. into SunTrust, but it backed peal more to residents who will sell in founder, the Rev. Craig Taylor.
only a few years.
" They've co-opted me to absolve them
off as negotiations improved.
" We took them at their word last

" Thus, the fact that fewer mortgage selves of their larger responsibility."

" I just can't believe Craig said that,"
year on the basis of their promise to lis- loans were made by Trust Company
ten to ourconcerns in good faith ,” Gold- Bank in black neighborhoods than in said Trust Company's Mynatt. " Trust
Bank has entertained any
Company
the
of
the
result
is
stein said. “ That might have been a white neighborhoods
mistake . "

451

a ponit . praising Trust i umpany

Trust Co.

Trust Company Bank has en

Banks need ' color-blind' policies

By Bill Dedman

from black areas.

452

CO
THE ATLANTA NSTITUTION

Part 4
Bank proti Atleta

Wednesday, May 4, 1988

seks me wint

Bank protesters in Atlanta

ter City, schod
har bond
i voled down

it!

make ready to flex muscle
By Bill Dedman
StaffWriler

453

Protesters

60-893 O - 92 - 16

tind bell !: ! I primi utt . ( 'oli 11

454

C
(

MER

NICK ARROYO /Staff

Demonstrators, red ribbon in hand, protest in front ofGeorgia Federal last fall over alleged ' redlining.'
Showing off the neighborhood chase loans at lower interest rates in from the banks; ACORN claims ACRA is
inner-city areas, financial counseling just a bunch of white liberals with no

One of the most effective tech- for potential homebuyers at inner-city community base.
niques, community organizers said, isto branches,and low -cost money orders at
Still, they agree that minority and
rent a bus.

455

‘ Anybody can sit out there and
mouth off:He gets things done'
By Bill Dedman

IT
TIO

policies.
Residents also began to push the
" If you can do housing in this neighJonesboro Road has three peepholes borhood , you can do housing any city to enforce laws against trash piles
and a slot – so former tenants could where," Taylor said.
The back door of Taylor's office on

HOW

T:'TUT

456

RENEE HANNANS/Staff

Church volunteers Dan Rodrigue ( left) and Con- being restored under the auspices of the South
nie Trainor work on a house on Bisbee Avenue

Atlanta Land Trust.

" I ask them a question and look di- some inner-city neighborhoods.
rectly in their eyes and see if they're
telling the truth. If they're honest, I'll

qualify for a loan because of an unpaid

Taylor suggests that lenders look at student loan.
it from a different perspective :

457

City Hallclout could sweeten home-loan pot
By Bill Dedman

for Underground Atlanta and loans to

first." Jackson said. " You're talking

Mayor Andrew Young expressed help finance the Democratic National about conservative institutions. That's
more ambivalence.

Convention.

458

result
the

ER.B.
MICHAEL BARRETT/Special

Former Mayor Maynard Jackson and Mary Webb Association of Communities Organized for Re
listen to Annette Wilcoxson ( left) , president of the form Now, question bank lending practices.
businesses. Chicago, Iowa, Michigan,
Missouri, Ohio and Wyoming have tried
such measures.

'I like to try to negotiate things first [ before removing deposits
Washington , from banks) . You're talking about conservative institutions. That's
stick: Conn..
Chicago,
Colorado,Illinois,
D.C.,The
Hartford,
MassachusettsandWest Virginia place not going to change.Thereare probably banks that are going to

government funds only inbanks that be moreresponsive than others; we have to findthose banksand

meet specific standardsfor lending.

son told a recent meeting of the Associ

banks. Beginning January 1987, the city

459

ivionths of work, but lending pool stillbone -dry
By Bill Dedman
Staff Writer

'The worst thing we could do is

mortgage," Crum said. " People can only
" I can't tell you how fast this will afford a home valued at a little more
- As proposed in 1986 by the alliance, the pool would have included $ 10 happen,” said Johnson, a Trust Compa- than two times their annual incorre, so
million a year each from 10 banks for ny spokesman. “ We're not dragging our this program would be good for anyone
single -family homes, a total of $ 100 mil.

feet over here."

460

‫الحدين‬

Nie

RENEE HANNANS/Staff

Activists like Karen Pleas, of104 Bisbee Ave., hope to create new housing in Atlanta with non-profit corporations.

Self-help the aim of non -profit housing corporations
By Bill Dedman
StaffWriter

Of 3,000 community develop

a loan pool for home loans, although

461

How study of Atlanta home loans was conducted
The Atlanta Journal - Constitution

study oflending patterns trackedhome. The newspapers used the Freedom ofInformation Act to obtain
made by everybank, savings and loan lenderreports compiled by the Federal Financial Institutions Ex
association, and large credit union in
metro Atlanta from 1981 through 1986.
The newspapers used the federal

amination Council and matched the data with demographic data

ey on real estate than savings and loans,

462

Part 5

Follow -up
and reaction

Sunday, May 4 - Opinion Page - Atlanta Journal -Constitution

Redlining : An economic war
waged on black communities

tlanta's black neighborhoods are under attack, and
under attack from some unlikely sources: the city's
most reputable banks and savings and loans. Their

write fewer loans to blacks, but it doesn't: When blacks and
whites of similar income are paired, blacks still fare far
worse in getting loans.

Wednesday, May 4 + Editorial Page * The Atlanta Constitution

If it's not redlining, prove it
The chairman of a local bank has already said it
best : " Those numbers are damning, " said Frank
Burke of Bank South. “ Those numbers are mind -boggling ."

an exhaustive study by The Atlanta Journal-Constitu
tion of home-purchase and home-improvement prac
tices of local banks and savings-and-loan institutions.

463

American families accumulate wealth . The practices

Redlining

Monday, May 9

of the city's major financial institutions result in a

Editorial Page The Atlanta Constitution

Banks can fix racial disparities
In the wake of Atlanta Journal and Constitution
reports showing that Atlanta's banks are making, proportionally, far fewer home and business loans in

black parts of the city than in comparable white
ones, black state legislators have called for a meeting

with Georgia's banking commissioner. In addition,
City Council President Marvin Arrington is forming a

to instruct loan officers to pay close heed to their
showing in black neighborhoods – and expect to
have their prospects for raises and promotions
judged in part by whether they succeed in closing
the racial gap.

464

brunclie's only where they call make money with

for comparison next year.

them .

ATLANTA
BANKS

LANO

MARLEL FOMUS
HEANACA CONSTRON
11

YOU NEED TO SPEAK TO ONE OF OUR LOAN OFFICERS ! "

Saturday, May 7 A Editorial Page The Atlanta Journal-Constitution

因
NA

AMERICA'S FIRST BLACK ARCHBISHOP

MARLEITE 2017
THE TWTANSTITUTO

JUST DON'T EXPECT TO GET A MORTGAGE LOAN ! "
39

465

May 5

Panel appointed to probe banks'lending policies
By Bill Dedman

" I don't take any quarrel with the
they do not intentionally discriminate.

practices.

466

tay 7

State banking chief urges lending law
By Bill Dedman

Fulton approves fair-housing law

laws track federal regulations.
At the Capitol , the comments by

State Rep. Billy McKinney ( D -At-

Last week, at a specially called com

Commissioner Dunn came after several lanta) and Atlanta City Council member mission meeting, the ordinance was
black legislators requested to meetwith HoseaWilliams appeared before the held becauseof internal politics on the
him. Dunn said the meeting would be
arranged next week

commission to urge adoption of the or-

commission . Joyner has been accused of

467

May 10

Atlanta NAACP calls for boycott, probe of banks
By Bill Dedman

The Return of " Redlining
T

468

Biay 13

Banks to lend $ 65 million at low interest
By Bill Dedman

ACO

469

Lay 16 Editorial Page The Atlanta Journal

Banks act swiftly to close
lending gap in black areas

old the tough talk. NAACP. Cancel your

rate they reinvest in white communities. Our series

boycott, Concerned Black Clergy. Atlanta

showed that in black communities the rate of rein

ATL
AN
BAN TA
KSI
S

BASSTATISTI JOWWAL "

" YOU CAN IMAGINE HOW SHOCKED AND DISMAYED WE WERE TO READ OF LOAN DIFFICULTIES FOROUR FINE BUCK CITIZENS"

470

CHE
1
Bill Dedman researched and wrote series on home loans.

The people who worked on this series
“ The Color of Money” was researched and written over a period of five
months by Journal-Constitution staff writer Bill Dedman. A native of Chat
tanooga, Tenn., Dedman, 27, joined the Journal -Constitution in 1987.

The project was supervised by Hyde Post, assistant city editor for spe
cial projects, and copyedited by Sharon Bailey . Dwight Morris, assistant

managing editor for special projects, supervised the analysis of lending
The statistical analysis used methodology largely developed by Calvin

data.

Bradford and Charles Finn of the Hubert H. Humphrey Institute of Public
Affairs at the University of Minnesota. These researchers together have
nearly 20 years of experience in the field of community economic develop
ment Research assistance was provided by Stan Fitterman , a graduate stu
dent in city planning at Georgia Tech.

The Journal-Constitution also adapted some methods and ranking sys
tems used by researchers at Johns Hopkins University and Temple
University.

471

For release on delivery
12:15 pm EDT
May 11, 1992

The Future of Banking :

Choosing the Right Model

Address by

Lawrence B. Lindsey
to

The California Bankers Association
Long Beach , California

May 11 , 1992

472

THE FUTURE OF BANKING : CHOOSING THE RIGHT MODEL
It is a pleasure to be here today . Observers of American
culture have often noted that one can see the future of America
by looking at the California of today .

Events of the past few

weeks have shown us that the future will not inevitably be
It will require boldness , vision , and

brighter than the present .

innovation to assure that the California of the future - and the
America of the future · are places we can be proud to leave to
our children .
Your industry is one key to that future .

You are in the

position to be the financial heart of a more vibrant economy .
With an entrepreneurial spirit that is willing to take on new
am

challenges , your future and that of America can be bright .
an optimist .

But we should be clear that a bright ,

entrepreneurial future is not inevitable .

It is one which can

only be developed with hard work and a commitment to a broader

vision of the mission of your industry .
Today the banking industry faces two very different
competing models of the future .
regulated utility model .

The first I shall call the

You are all familiar with the trend

toward this model in your daily lives as bankers , particularly in

the last few years .

It is a model of increasing government

intervention in the way you run your banks .

You may be expected

to be banker , policeman , and social worker all at once .

addition , your profitability may become increasingly determined
by fiat as increasing costs are passed along to you based on the

political perception of your industry's capacity to absorb those

473

costs .

The second model of the future of banking is the competitive
market model . A truly competitive market would involve a much

more intense and rigorous form of competition than the industry
has experienced to date .

While government would continue to

stand as ultimate guarantor of the integrity of the system , the
taxpayer would be protected by adequate capital rather than

intense regulation .

Under this model , the industry's success or

failure will depend upon its ability to provide a vital service

to the economy : efficient evaluation of credit worthiness .
Granted , other financial institutions will be attempting to

perform the same function .

The bank form of credit evaluation

has the unique advantage of a physical presence in the community
This necessarily also entails higher costs .

which it serves .

Your success in the intensely competitive financial market place

of the future will require making the most of your community
presence .

I do not know which path the industry will ultimately take :
that of a regulated utility or that of a community based
competitive provider of financial services .

I prefer .

I can tell you what

This is one regulator who fervently hopes that his

job's responsibilities do not become any more intrusive than they

already are .

Turning banks into regulated utilities is not good

for the banking industry and it is not good for the country .
Unfortunately , neither the public nor their elected

representatives seem disposed to easing off on bank regulation .

474

The battle ahead will be a long one .

will be fought are changing.

During the late 1980s and early

1990s , the basis for bank regulatory action has been the issue of

safety and soundness .

In this vein , last year's banking bill

established highly prescriptive regulations regarding activities
which , in the judgment of Congress , were risky to the banking
industry .
The regulatory issues in the 1990s will not be limited to

safety and soundness , but will increasingly emphasize fairness :
whether or not banks are fulfilling the needs of their
communities .

Today we all know this as C.R.A.

Reinvestment Act

the Community

but it is potentially much broader .

The

existing CRA rules are deliberately non - prescriptive and I

support them in their current form .

Congress wisely chose to

avoid explicit allocation of credit in enacting CRA , letting the
local bank define its mission to the community and the means it
would use to meet that mission .

it would never come to this , one could envision credit rules
allocating the types , volume and location of loans that banks
could make .

One could also imagine racial , ethnic , income and

geographic guidelines regarding the recipients of those loans .
shift in the regulatory framework in this direction is not

475

inevitable , nor , however , is it inconceivable .
Let us consider why .

Racial discrimination is not only

illegal , it is morally repugnant .

Racial discrimination tears at

the very fabric of our national ideal .

A society in which each

individual is evaluated on his or her own worth is what we seek
- not one in which individuals are treated based on the group of
which he or she may be a part .

enterprise system .

This is the basis of a free

Allowing discrimination to continue in either

spirit or substance is not only antithetical to our political
institutions , it is destructive of our economic liberty as well .
Because so much is at stake , our political leaders will take
whatever actions they deem necessary to combat it .

Those actions

may not , by their very nature , necessarily advance our
Constitutional ideal of individualism .

Those actions may not

advance economic liberty or promote economic growth .

But they

will be taken in spite of their costs because combating racism

represents a moral imperative .
In the case of banking, an increasingly prescriptive set of

rules regarding lending practices could be the response to the
perception that existing lending practices are unfair . That is
how the regulated utilities model of banking is likely to emerge.
We already have legislative calls to examine the level of small
business lending done by banks .

Serious legislative concerns

have already been raised about the Housing Mortgage Disclosure
Act data ( HMDA data )

regarding differential rejection rates in

mortgage lending for different racial and ethnic groups .

476

Let me add that although I had settled on this as the topic

of my speech many weeks ago , the tragic events which started not
far from here 13 days ago can only reinforce the message I am
delivering this morning .
.

I can assure you that questions will

emerge in the weeks ahead , if they have not already , about the
roles your institutions have played in the past , and can play in
the future in funding economic development of inner city areas .

What you will be defending are not the policies of your
banks .

Discrimination need not be overt or even conscious .

it was , your task would be obvious .

largely subjective or a matter of perception .
it any less real .

What you must address is

This does not make

But it does make it harder to both identify

the problem and devise the solution .

It will require exceptional

diligence , commitment , and creativity on your part .

that is already against the stated policy of your organization
You are right . But , the mere

and perhaps not even real .

perception of unfairness , not to mention its reality , may drive
policymakers to take action .

The likely outcome may not be good

for you , good for the country , or particularly fair .

I therefore

mention four ideas for combatting the perception of unfairness ,

not to dictate to you how to run your institutions , but to
suggest ways that may help you solve the problems you face .
Let me begin with the subject of perceptions . You are
service business .

In other service businesses , such as hotels ,

477

restaurants , and retail stores , management often hires employees
to shop and report back on the level of customer service .

Indeed , this practice is not unknown to banking .

How else do you

tell whether your tellers give reliable information to your
customers in a pleasant manner ?
Specifically , you might have minority individuals try

shopping for credit and other services at a bank branch where
The anecdotal reports we

they are not known, then report back .

have are that the different treatment that does take place is not
overt , but subtle .

It involves loan officers not extending the

same courtesy , or keeping minority applicants waiting longer than
non - minority customers .

It may involve loan officers putting

forth less effort to provide qualifying tips to minority
applicants or mentioning to those applicants fewer loan products
and options . This is important information for you to find out as
a service business .

Fed and other bank regulators to perform this shopping function

as a part of our enforcement activity .
unfortunate if it came to that .

I think that it would be

The Federal Reserve has

enumerated a number of reasons for its reluctance .
these is efficacy .

Key among

As a regulatory body , our evidentiary

techniques could not rely on anecdotal evidence , but would most
likely entail the gathering of a statistically significant
sample .

This is not only expensive for us as a regulator , it

would also involve a substantial burden on you .

478

In addition , we are dealing with information which is

inherently subjective and not easily quantified .

As an

enforcement agency we cannot easily measure whether subtle
variations in conduct occurred a statistically significantly

number of times , for different applicants.

On the other hand ,

the impression that your shopper gets is valuable to you as a
That impression is by its very nature of great proprietary
value and relatively little regulatory value .

One concern that has been expressed about shoppers is the
potential liability banks might suffer in having the reports
subject to discovery in class action suits .

I think that

Congress might be well advised to consider some sort of safe

harbor protection in this regard .

But , even absent such a safe

harbor , shopping can still be done profitably by your institution

to report back non - quantifiable impressions .

These impressions

could prove very valuable to your organization in a proprietary
sense .

I therefore commend shopping by your own employees as a way

of gathering important information about the way you treat your
customers from different backgrounds .

Such activity could go far

in removing the perception of unfairness by attacking it at its
roots .

It is likely that your employees may not even be aware

that they act in it manner which is considered discriminatory or

offensive .

Simply providing information to these employees may

prove to be a very important consciousness raising step .
Shopping may prove an important adjunct to your other training

479

and educational techniques .

Let me also say that it could prove

to be a very profitable step in improving customer relations and
creating new lending opportunities.

A second idea I would like you to consider gets to the heart
of the subject of mortgage discrimination .

I believe this is

particularly important to deal with because of the vital role

homeownership plays in our society .

In my travels throughout the

country to neighborhood reinvestment projects , I am repeatedly

struck at the difference home ownership can make to the
individuals in a community .

Advancing homeownership is something

that is good both for your customers and for the communities in
which you do business .

mortgage review boards .

These organizations are completely
voluntary and can be totally private sector in nature . Two

models of mortgage review boards are now working .

In Boston and

Detroit rejected mortgage applicants may forward their
applications to the board to appeal that outcome .

Members of the

review board are banking and thrift institutions who are active
in the local mortgage lending market as well as representatives
from local community organizations .

Rejected applicants who meet

acceptable criteria are provided loans by board members on a

rotating basis .

This approach not only provides homeownership

opportunities for those who might not otherwise have them , it

also enhances the perception of fairness in the eyes of the
public .

An opportunity for appeal is created and that appeal is

480

based on a willingness to reassess applications as agreed upon by
a large number of reviewers .
A second approach to the mortgage review board process is
practiced in Philadelphia .

There , a group of mortgage lenders

including local banks and thrifts targets key neighborhoods where

more flexible lending standards could have a substantial impact .
Applicants for mortgages in these neighborhoods whose

applications are likely to be denied , have their application
automatically reviewed by this committee of lenders .
application is forwarded directly by the lending institution and

requires no effort on the applicant's part .

More flexible

criteria regarding employment stability and credit history are
used .

This process avoids the stigma of an initial rejection and

represents a more direct , proactive approach to mortgage lending .
The key merit of both approaches is their voluntary nature .

No quotas are being filled .

No institution is required to take a

mortgage which objective criteria indicate is not likely to be
repaid .

This seems like an idea worth looking at .

Whether it

might work for you in California is for you to decide .

Again ,

the emphasis of my comments today is on finding creative ,

voluntary , and private sector -oriented actions you might take to
end unfairness , both real and perceived .

A third idea I'd like to mention is consumer education .

am increasingly aware from consumer education surveys that many
Americans
especially young Americans
have little
understanding of the basics of consumer credit :

what a debt - to

481

income ratio is or what the consequences of defaulting on a
department store bill a few years ago may have when they decide

to purchase a home .

Many Americans also have little

understanding of the implications of repayment with interest .

Too many Americans feel that they have met their financial
responsibilities if they are able to cover the minimum payment on
their credit cards each month .

Little thought is given to the

concept of long term saving to finance consumer purchases .
In addition , innovations in consumer education could be used

to diminish some of the perceptions about mortgage lending and
other banking activities and ultimately to help more consumers be
successful home purchasers and customers .

Consumers who are

familiar with the lending process are more likely to be at ease
during a loan interview and less apt to misunderstand an
explanation of their eligibility for a particular mortgage , or
for credit with the lender at all .

I firmly believe that our country's consumers must be

educated for success , by learning at an early age the most

critical areas of personal finance that will determine whether
credit doors are opened or closed in their future .

I hope that

you will work actively with local community groups and
particularly the schools in your communities to promote basic
financial education and understanding .

That educational effort ,

along with other efforts , such as I've outlined today , will go a
long way to helping future consumers be creditworthy .

Helping

these customers own their own homes and fully participate in the

482

economic life of the country is good civics .
business .

It is also good

From an industry perspective , consumer education will

go a long way to increasing the number of customers who are
active users of banking services .
A fourth and final idea I would like you to consider is a

much more active role in lending to small business , particularly
minority small businesses .

Let me be candid .

The nature of the

banking franchise has changed and will continue to change in the

years ahead .

Direct access to capital markets will become an

increasing fact of life as securitization of business loans by
the financial market and the process of deepening of the
commercial paper market continues.

These developments will

inevitably shrink the size of your traditional lending markets .

For the banking industry to survive and prosper , it must do
better than other financial intermediaries at credit evaluation
and allocation .

Some argue that banking is at a disadvantage

because it is a high cost financial intermediary .

Those costs

are reflected in the physical networking into communities that is

not done by say , a discount broker or underwriter of commercial
paper .

The higher costs are also the result of the maintenance

of a large credit evaluation staff .

You will survive in a

competitive environment only if the physical network of branches
and the large staffs at your disposal provide advantages which
offset their costs .
Let's face it .

Institutions with high operating costs

simply cannot profitably compete in low margin businesses such as

483

the commercial paper market or in the government securities
However , you can be profitable in retail banking . This

market .

involves not just consumer lending , but lending to small
businesses who do not have access to the lower cost markets .

The

physical presence of branches and loan officers within the
community give you the advantage in determining the credit

worthiness of the individuals in that community .

The lack of

non - bank competition in this market assures that the margins
exist for you to compete profitably in that community .

That is

where opportunities for growth lie .

In the 1990s America will continue to rely on new businesses

and small businesses as the engines of economic progress .

We are

probably unique in the world in having the vast majority of our

net new jobs created by such businesses .

Entrepreneurship is not

only our engine of economic growth , it is also the means our
economy has provided for climbing the economic ladder .
The Wall Street Journal called the 1980s the decade of
minority capitalism .

Between 1983 and 1987 there was an 83

percent increase in the number of Hispanic owned businesses .
There was a 50 percent increase in the number of businesses owned
by African Americans .

More black owned businesses were created

from 1982 to 1987 than in any other comparable five year period
in our history -- and by a wide margin .

More Asian Americans and

more women went into business than at any other time .

your natural customer base .

They are also your natural allies in

484

combatting the impression that the system is unfair .

Each

minority owned business which gets its capital from your bank
stands as a living refutation of the notion that the system is
unfair .

These businesses are also the vehicle by which economic

opportunity is brought both to individuals and to the communities
in which they live .

Reaching out to those new entrepreneurs is not going to be
easy .

It may require a change in the way you do business .

will require you to go out of your traditional market and find
new markets .

But , that is the essence of free enterprise .

You

must find and develop new markets for the survival of your
institutions .

That is your competitive challenge for the 1990s .

These new customers are your opportunity if you have the
competitive capacity to rise to the challenge .

becoming a regulated utility .

In some ways the political

challenge to meeting the perception that banking is unfair to
minority groups augments the economic challenge you face to reach
out and develop new markets and new customers .

As a regulator

who doesn't want more regulations I want you to succeed .
Furthermore , as an American who believes in this country and the
concepts of individualism and free enterprise that America

represents , I hope that you do succeed .

The alternative is a

future that none of us want .

60–893 ( 492 )

Full text of Report on the Status of the Community Reinvestment Act : Views and Recommendations : Report on the Status of the Community Reinvestment Act : Views and Recommendations, Volume II of II

FRASER