wlm4 at cornell.edu (Will
01-04-1995, 09:08 PM
I have been disappointed by the response of the financial industry to the
release of the Home Mortgage Disclosure Act (HMDA) statistics. As a manager
of a community development Credit Union, I see three easy truths:
1) There is a problem. We can not reasonably deny the weight of the
HMDA numbers.
2) Lending discrimination is not about individual racism, that is,
it is not about individuals consciously choosing to discriminate.
Certifying that our staff are not racists is not enough. Our policies may
have a "disparate impact", which means that when a lender applies the same
practice to all applicants, the effect is to discriminate in a way that is
not justified by business necessity.
3) Business are problem solving institutions. The situation can be
made right through value free lending underwriting without accepting
additional risk. Underserved markets represent an opportunity!
There is a problem.
HMDA statistics show that Black and Hispanic-Americans were denied mortgages
at substantially greater rates than were white and Asian- Americans of
similar incomes. The federal report criticized the underwriting standards
of Freddie Mac and Fannie Mae as being "minority unfriendly". These mortgage
markets came under criticism from affordable housing groups for bias against
low-income household income and savings patterns and inadvertent mortgage
loan discrimination.
Upon hearing the HMDA statistics pointing to widespread discrimination in
lending, bankers went into deep denial. But the numbers can't be ignored,
they aren't borderline. The numbers show a two-fold increase in turndowns
if the applicant is a minority. Initial reaction from bankers was that
there has to be something wrong with the numbers: from off the cuff remarks
that income wasn't factored in (it was) to more lengthy arguments of the
minutia of the data. The American Bankers Association emphasized the data
didn't include "information critical to judging the creditworthiness of loan
applicants". ABA lobbyists argued that "the HMDA numbers don't show a whole
lot; they don't mean a whole lot".
Bankers disputed what seems obvious to those of us involved in community
development banking. Banking is not doing a good job reaching minority and
low income borrowers.
A second round of reactions lead to several banks instituting visible (but
small) community lending programs, often in conjunction with fines paid for
past wrongdoing.
Concurrent statistics from the Woodstock Institute of Chicago and the
Federal Reserve Bank of Philadelphia show that loans to low income single
family home borrowers have default and delinquency rates significantly lower
than multifamily housing and a little lower than middle class borrowers.
Put these two statistics together and the problem stands out starkly.
Minority and low income borrowers were approved much less often but repaid
their loans (when granted) more consistently. What causes this negative
approval statistic coupled with this positive repayment statistic?
Let's start by agreeing that this is happening: the mortgage market doesn't
work well for low income and minority borrowers. There shouldn't be any
argument about this anymore.
Disparate impact is not the same as personal racism.
There is a certain myopia in bankers that won't allow us to see value
through difference. Underwriting standards are looking for sameness.
Bankers consistently over estimate the value of middle class borrowers and
under estimate their low income applicants. It's as if bankers expect
minorities to look like pigmented white people and low income people to
have economic profiles like rich people, only smaller. Minorities are
expected to adapt, pull themselves up by their bootstraps, and in general,
"fit in". The counterpoint is that there is value in uniqueness. What we
need is an underwriting rainbow.
The reliance on secondary market standards for underwriting strikes at the
heart of the CDFI mission. We should not willy nilly loosen underwriting
standards and expect no outfall. and CDFIs are no exception. One would
expect a CDFI to:
- Know its community and judge it more accurately than a national standard
allows.
- Look at each loan individually.
- Separate cosmetic flaws from substantial problems.
- Separate technical flaws from substantial problems.
- Balance true flaws with compensating factors.
One Credit Union was criticized by NCUA for a 100% rejection rate for
Asian-American loan applicants. The CU defended its record by saying that
they had only two applicants, both of whom were denied. A more appropriate
response would have detailed the number of Asian- Americans in the field of
potential membership, the efforts made to market to these members, their
representation in staff and volunteers, special products designed to meet
their needs, and the convenience of branches to their populations. Treating
everybody the same is not enough.
Some recent affordable housing programs address one or two secondary market
flaws without giving a thorough analysis of the possibilities open to low
income borrowers. For example, we have a seen a program supposedly targeted
to this market that lowered the borrower down payment requirement to 3%,
BUT included a requirement that a relative give the borrower an additional
2% down payment. This makes the unlikely assumption that a low income
borrower has rich relatives! Programs like the one described provide lip
service only. We can't realistically talk about gifts, co-borrowers or
larger down payments to avoid PMI.
About 60% of the people in the country own homes. We estimate that of the
remaining 40%, ten-twenty percent would qualify for homeownership within
these underwriting guidelines. There are large numbers of people being kept
out of home ownership needlessly.
************************************************** **********
William Myers
Alternatives Federal Credit Union
301 West State Street, Ithaca, NY 14850-5431
Voice (607) 273-3582 ext 817 FAX 277-6391
E-Mail Alternatives-Myers@Cornell.edu
************************************************** **********
This post transferred from the cdb-l mailing list
release of the Home Mortgage Disclosure Act (HMDA) statistics. As a manager
of a community development Credit Union, I see three easy truths:
1) There is a problem. We can not reasonably deny the weight of the
HMDA numbers.
2) Lending discrimination is not about individual racism, that is,
it is not about individuals consciously choosing to discriminate.
Certifying that our staff are not racists is not enough. Our policies may
have a "disparate impact", which means that when a lender applies the same
practice to all applicants, the effect is to discriminate in a way that is
not justified by business necessity.
3) Business are problem solving institutions. The situation can be
made right through value free lending underwriting without accepting
additional risk. Underserved markets represent an opportunity!
There is a problem.
HMDA statistics show that Black and Hispanic-Americans were denied mortgages
at substantially greater rates than were white and Asian- Americans of
similar incomes. The federal report criticized the underwriting standards
of Freddie Mac and Fannie Mae as being "minority unfriendly". These mortgage
markets came under criticism from affordable housing groups for bias against
low-income household income and savings patterns and inadvertent mortgage
loan discrimination.
Upon hearing the HMDA statistics pointing to widespread discrimination in
lending, bankers went into deep denial. But the numbers can't be ignored,
they aren't borderline. The numbers show a two-fold increase in turndowns
if the applicant is a minority. Initial reaction from bankers was that
there has to be something wrong with the numbers: from off the cuff remarks
that income wasn't factored in (it was) to more lengthy arguments of the
minutia of the data. The American Bankers Association emphasized the data
didn't include "information critical to judging the creditworthiness of loan
applicants". ABA lobbyists argued that "the HMDA numbers don't show a whole
lot; they don't mean a whole lot".
Bankers disputed what seems obvious to those of us involved in community
development banking. Banking is not doing a good job reaching minority and
low income borrowers.
A second round of reactions lead to several banks instituting visible (but
small) community lending programs, often in conjunction with fines paid for
past wrongdoing.
Concurrent statistics from the Woodstock Institute of Chicago and the
Federal Reserve Bank of Philadelphia show that loans to low income single
family home borrowers have default and delinquency rates significantly lower
than multifamily housing and a little lower than middle class borrowers.
Put these two statistics together and the problem stands out starkly.
Minority and low income borrowers were approved much less often but repaid
their loans (when granted) more consistently. What causes this negative
approval statistic coupled with this positive repayment statistic?
Let's start by agreeing that this is happening: the mortgage market doesn't
work well for low income and minority borrowers. There shouldn't be any
argument about this anymore.
Disparate impact is not the same as personal racism.
There is a certain myopia in bankers that won't allow us to see value
through difference. Underwriting standards are looking for sameness.
Bankers consistently over estimate the value of middle class borrowers and
under estimate their low income applicants. It's as if bankers expect
minorities to look like pigmented white people and low income people to
have economic profiles like rich people, only smaller. Minorities are
expected to adapt, pull themselves up by their bootstraps, and in general,
"fit in". The counterpoint is that there is value in uniqueness. What we
need is an underwriting rainbow.
The reliance on secondary market standards for underwriting strikes at the
heart of the CDFI mission. We should not willy nilly loosen underwriting
standards and expect no outfall. and CDFIs are no exception. One would
expect a CDFI to:
- Know its community and judge it more accurately than a national standard
allows.
- Look at each loan individually.
- Separate cosmetic flaws from substantial problems.
- Separate technical flaws from substantial problems.
- Balance true flaws with compensating factors.
One Credit Union was criticized by NCUA for a 100% rejection rate for
Asian-American loan applicants. The CU defended its record by saying that
they had only two applicants, both of whom were denied. A more appropriate
response would have detailed the number of Asian- Americans in the field of
potential membership, the efforts made to market to these members, their
representation in staff and volunteers, special products designed to meet
their needs, and the convenience of branches to their populations. Treating
everybody the same is not enough.
Some recent affordable housing programs address one or two secondary market
flaws without giving a thorough analysis of the possibilities open to low
income borrowers. For example, we have a seen a program supposedly targeted
to this market that lowered the borrower down payment requirement to 3%,
BUT included a requirement that a relative give the borrower an additional
2% down payment. This makes the unlikely assumption that a low income
borrower has rich relatives! Programs like the one described provide lip
service only. We can't realistically talk about gifts, co-borrowers or
larger down payments to avoid PMI.
About 60% of the people in the country own homes. We estimate that of the
remaining 40%, ten-twenty percent would qualify for homeownership within
these underwriting guidelines. There are large numbers of people being kept
out of home ownership needlessly.
************************************************** **********
William Myers
Alternatives Federal Credit Union
301 West State Street, Ithaca, NY 14850-5431
Voice (607) 273-3582 ext 817 FAX 277-6391
E-Mail Alternatives-Myers@Cornell.edu
************************************************** **********
This post transferred from the cdb-l mailing list