[House Hearing, 110 Congress]
[From the U.S. Government Publishing Office]
THE IMPACT OF CREDIT-BASED INSURANCE
SCORING ON THE AVAILABILITY
AND AFFORDABILITY OF INSURANCE
=======================================================================
HEARING
BEFORE THE
SUBCOMMITTEE ON
OVERSIGHT AND INVESTIGATIONS
OF THE
COMMITTEE ON FINANCIAL SERVICES
U.S. HOUSE OF REPRESENTATIVES
ONE HUNDRED TENTH CONGRESS
SECOND SESSION
__________
MAY 21, 2008
__________
Printed for the use of the Committee on Financial Services
Serial No. 110-113
----------
U.S. GOVERNMENT PRINTING OFFICE
43-699 PDF WASHINGTON : 2008
For sale by the Superintendent of Documents, U.S. Government Printing
Office Internet: bookstore.gpo.gov Phone: toll free (866) 512-1800;
DC area (202) 512-1800 Fax: (202) 512-2104 Mail: Stop IDCC,
Washington, DC 20402-0001
HOUSE COMMITTEE ON FINANCIAL SERVICES
BARNEY FRANK, Massachusetts, Chairman
PAUL E. KANJORSKI, Pennsylvania SPENCER BACHUS, Alabama
MAXINE WATERS, California DEBORAH PRYCE, Ohio
CAROLYN B. MALONEY, New York MICHAEL N. CASTLE, Delaware
LUIS V. GUTIERREZ, Illinois PETER T. KING, New York
NYDIA M. VELAZQUEZ, New York EDWARD R. ROYCE, California
MELVIN L. WATT, North Carolina FRANK D. LUCAS, Oklahoma
GARY L. ACKERMAN, New York RON PAUL, Texas
BRAD SHERMAN, California STEVEN C. LaTOURETTE, Ohio
GREGORY W. MEEKS, New York DONALD A. MANZULLO, Illinois
DENNIS MOORE, Kansas WALTER B. JONES, Jr., North
MICHAEL E. CAPUANO, Massachusetts Carolina
RUBEN HINOJOSA, Texas JUDY BIGGERT, Illinois
WM. LACY CLAY, Missouri CHRISTOPHER SHAYS, Connecticut
CAROLYN McCARTHY, New York GARY G. MILLER, California
JOE BACA, California SHELLEY MOORE CAPITO, West
STEPHEN F. LYNCH, Massachusetts Virginia
BRAD MILLER, North Carolina TOM FEENEY, Florida
DAVID SCOTT, Georgia JEB HENSARLING, Texas
AL GREEN, Texas SCOTT GARRETT, New Jersey
EMANUEL CLEAVER, Missouri GINNY BROWN-WAITE, Florida
MELISSA L. BEAN, Illinois J. GRESHAM BARRETT, South Carolina
GWEN MOORE, Wisconsin, JIM GERLACH, Pennsylvania
LINCOLN DAVIS, Tennessee STEVAN PEARCE, New Mexico
PAUL W. HODES, New Hampshire RANDY NEUGEBAUER, Texas
KEITH ELLISON, Minnesota TOM PRICE, Georgia
RON KLEIN, Florida GEOFF DAVIS, Kentucky
TIM MAHONEY, Florida PATRICK T. McHENRY, North Carolina
CHARLES A. WILSON, Ohio JOHN CAMPBELL, California
ED PERLMUTTER, Colorado ADAM PUTNAM, Florida
CHRISTOPHER S. MURPHY, Connecticut MICHELE BACHMANN, Minnesota
JOE DONNELLY, Indiana PETER J. ROSKAM, Illinois
ROBERT WEXLER, Florida KENNY MARCHANT, Texas
JIM MARSHALL, Georgia THADDEUS G. McCOTTER, Michigan
DAN BOREN, Oklahoma KEVIN McCARTHY, California
BILL FOSTER, Illinois DEAN HELLER, Nevada
ANDRE CARSON, Indiana
Jeanne M. Roslanowick, Staff Director and Chief Counsel
Subcommittee on Oversight and Investigations
MELVIN L. WATT, North Carolina, Chairman
LUIS V. GUTIERREZ, Illinois GARY G. MILLER, California
MAXINE WATERS, California PATRICK T. McHENRY, North Carolina
STEPHEN F. LYNCH, Massachusetts EDWARD R. ROYCE, California
NYDIA M. VELAZQUEZ, New York RON PAUL, Texas
MICHAEL E. CAPUANO, Massachusetts STEVEN C. LaTOURETTE, Ohio
CAROLYN McCARTHY, New York J. GRESHAM BARRETT, South Carolina
RON KLEIN, Florida MICHELE BACHMANN, Minnesota
TIM MAHONEY, Florida PETER J. ROSKAM, Illinois
ROBERT WEXLER, Florida KEVIN McCARTHY, California
C O N T E N T S
----------
Page
Hearing held on:
May 21, 2008................................................. 1
Appendix:
May 21, 2008................................................. 57
WITNESSES
Wednesday, May 21, 2008
Hunter, J. Robert, Director of Insurance, Consumer Federation of
America........................................................ 32
Keiser, Hon. George J., Representative, State of North Dakota, on
behalf of the National Conference of Insurance Legislators
(NCOIL)........................................................ 17
McCarty, Hon. Kevin, Insurance Commissioner, State of Florida, on
behalf of the National Association of Insurance Commissioners
(NAIC)......................................................... 15
Neeson, Charles, Senior Executive, Personal Lines Products,
Westfield Group, on behalf of Property Casualty Insurers
Association of America......................................... 38
Parnes, Lydia B., Director, Bureau of Consumer Protection,
Federal Trade Commission....................................... 14
Poe, Eric, Chief Operating Officer, Cure Automobile Insurance.... 36
Powell, Lawrence S., Ph.D., Professor, University of Arkansas at
Little Rock.................................................... 42
Pratt, Stuart K., President, Consumer Data Industry Association.. 40
Rice, Lisa, Vice President, National Fair Housing Alliance....... 34
APPENDIX
Prepared statements:
Carson, Hon. Andre........................................... 58
Hunter, J. Robert............................................ 59
Keiser, Hon. George J........................................ 101
McCarty, Hon. Kevin.......................................... 112
Neeson, Charles.............................................. 168
Parnes, Lydia B.............................................. 173
Poe, Eric.................................................... 181
Powell, Lawrence S........................................... 193
Pratt, Stuart K.............................................. 214
Rice, Lisa................................................... 233
Additional Material Submitted for the Record
Watt, Hon. Melvin:
Information from the Web sites of AllState, State Farm, and
Travelers Insurance Companies on what factors they consider
in determining rates....................................... 245
USA Today article entitled, ``Credit scores' link to
insurance rates tested''................................... 251
Responses to questions submitted to J. Robert Hunter......... 252
Responses to questions submitted to Lydia Parnes............. 254
Responses to questions submitted to Eric Poe................. 257
Responses to questions submitted to Lawrence S. Powell....... 272
Miller, Hon. Gary:
Letter from the American Insurance Association, the Financial
Services Roundtable, the Independent Insurance Agents and
Brokers of America, the National Association of Mutual
Insurance Companies, and the U.S. Chamber of Commerce...... 279
Statement of the National Association of Mutual Insurance
Companies.................................................. 280
Statement of Michael J. Miller and EPIC Consulting........... 289
Statement of the Property Casualty Insurers Association of
America.................................................... 294
``The Use of Occupation and Education Factors in Automobile
Insurance,'' State of New Jersey, Department of Banking and
Insurance, dated April 2008................................ 296
THE IMPACT OF CREDIT-BASED INSURANCE
SCORING ON THE AVAILABILITY
AND AFFORDABILITY OF INSURANCE
----------
Wednesday, May 21, 2008
U.S. House of Representatives,
Subcommittee on Oversight
and Investigations,
Committee on Financial Services,
Washington, D.C.
The subcommittee met, pursuant to notice, at 10:03 a.m., in
room 2128 Rayburn House Office Building, Hon. Melvin L. Watt
[chairman of the subcommittee] presiding.
Members present: Representatives Watt, Gutierrez, Waters,
Green, Klein, Boren; Miller, McHenry, Royce, Barrett, Roskam,
and McCarthy.
Ex officio present: Representative Bachus.
Also present: Representative Lynch.
Chairman Watt. This hearing of the Subcommittee on
Oversight and Investigations of the Financial Services
Committee will come to order.
I will recognize myself for 5 minutes or less for an
opening statement.
This is the second in a series of hearings held by the
Oversight Subcommittee to gain more information about the use
of consumer credit information in the underwriting and rating
of personal lines of insurance, including automobile and
homeowners' insurance.
These hearings are warranted because this practice, known
as ``insurance scoring,'' and its derivative products referred
to as ``credit-based insurance scores,'' or simply ``insurance
scores,'' cries out for careful evaluation to determine whether
it is consistent with good public policy.
We learned at the Oversight and Investigation
Subcommittee's first hearing in October of 2007 that almost all
major insurance companies use credit-based insurance scores in
some way. Consequently, nearly all Americans who drive cars or
own homes must also have good credit if they are to avoid
paying high insurance premiums, regardless of their individual
claims history or driving record.
We also learned through a report by the FTC that while
credit-based insurance scores are predictive of claims risk, or
claims, no one can explain why this is the case. We also
learned from the last hearing that in three out of four lines
of automobile insurance, credit-based insurance scores serve in
some measure as a proxy for race. However, given the data
concerns with the automobile study that witnesses discussed
extensively at the last hearing, the full extent of the proxy
effect still may not be known.
Despite perceived shortcomings of the data, the FTC report
concluded that there was some ``proxy effect'' from the use of
credit-based insurance scores, and as noted by a dissenting
Commissioner, ``Given the incompleteness of the data, it is
unclear whether the actual proxy effect might be greater.''
Even a minor proxy effect for race gives rise to the most
serious public policy concerns. I don't think anyone should
favor a system in which either directly or indirectly, racial
classifications are allowed to hinder a person in their daily
lives, whether in being considered for employment, getting an
education, buying a home, getting credit, or purchasing
financial products like automobile and homeowners' insurance.
Because of these major public policy concerns, two bills
have been introduced. One, H.R. 5633, introduced by
Representative Luis Gutierrez, the chairman of the Domestic and
International Monetary Policy Subcommittee of the Financial
Services Committee, would reign in the use of credit-based
insurance scoring by prohibiting the use of credit-based
insurance scores where the Federal Trade Commission finds
evidence of racial discrimination, or that credit-based
insurance scores serve as a proxy for race.
The second bill, H.R. 6062, introduced by Representative
Maxine Waters, chairwoman of the Housing and Community
Opportunity Subcommittee of the Financial Services Committee,
would prohibit the use of credit-based insurance scores
altogether in underwriting or pricing personal lines of
insurance.
I am going to digress just long enough to say that I am a
cosponsor of both of these bills. I actually think I start with
the assumption that there really shouldn't be a connection
between somebody's credit score and their insurance premium any
more than there should be a connection between somebody's
driving record and whether they get credit.
But I guess I am willing to be convinced that perhaps there
is some utility in the use of these scores, but I am not sure
that I can be convinced that if they are a proxy for race, we
can justify their use as a matter of public policy, even if
there is a correlation between credit scores and insurance
premiums, or underwriting of insurance.
So, I am on both bills. I am trying to keep somewhat of an
open mind on this issue, but we don't legislate in this
committee anyway. We just have hearings and build a record, so
my presence on either one bill or the other probably has no
significance to my role as chairman of the subcommittee. That
is just to put everything on the record.
We hope to shed more light on the pros and cons of each of
these two proposals, H.R. 5633 and H.R. 6062, as well as
consider changes or other options that might be appropriate at
today's hearing.
We look forward to hearing from the witnesses about the
potential impact of H.R. 5633 and H.R. 6062 on consumers and
the insurance industry. And with that, I will recognize my
colleague, the ranking member of the full Financial Services
Committee, Mr. Bachus, for 5 minutes.
Mr. Bachus. Thank you, Chairman Watt, for holding this
second hearing before your subcommittee on the impact of
credit-based insurance scoring on insurance availability and
affordability. Let me say at the onset that I acknowledge your
concerns and those of Mr. Gutierrez. I know that Congressman
Green has what I think are good-faith concerns, as has
Chairwoman Waters. So certainly I enter this hearing with an
open mind.
Credit scores, as we all know, are widely used for a number
of purposes other than lending, including employment
interviews, apartment rental applications, government licenses,
mobile phone services, as well as insurance, which brings us
here today. Credit scoring actually can help individuals who
manage their financial affairs responsibly, I believe, to get a
number of benefits that they might not otherwise receive, based
on traditional underscoring criteria such as age, gender, zip
code, or income.
So actually I believe in certain cases--and studies have
validated this--people have actually benefitted from their
credit scores through cheaper insurance or availability of
insurance. In fact, I think the FTC confirmed this in a recent
study that found the use of credit scores greatly increases
fairness and affordability for consumers of insurance products.
They found that more responsible and thus lower-risk
drivers get cheaper coverage, but they also found that higher-
risk drivers enjoy greater access to insurance because insurers
can more accurately price their risk.
They further found that ``Credit-based insurance scores
appear to have little effect as a proxy for race, although
every predictive factor the FTC analyzed had a slight disparate
impact on certain ethnic groups.'' And I think that, obviously,
is the subject matter of Mr. Gutierrez' legislation.
For example, they found that prior claims history had a
disparate impact on various ethnic groups, with nearly the same
percentage of proxy effect to predictive value as credit-based
insurance scores. So the use of credit scores for various
purposes--not only has the FTC studied it, but it has been
extensively scrutinized by State regulators. I will just
mention two.
The Texas Insurance Department recently analyzed 2 million
insurance policies and found a direct and non-discriminatory
correlation between insurance scores and expected losses. It
found that the average automobile insurance losses for people
with the worst credit scores are double those for people with
the best credit scores, while losses on homeowners' policies
for people with the worst credit scores are triple those of
people with the best scores. The Texas Department further found
that these scores were not unfairly discriminatory or based on
race or income.
A second study, this one by the Arkansas Insurance
Department, yielded similar results, including a finding that 3
times as many consumers received lower insurance rates because
of credit score use than received higher rates. In short, the
evidence from these studies appears pretty clear that credit
scores are one of the most accurate non-discriminatory
predictors of insurance risk available.
However--and I think maybe this would be a good starting
point for us to make some agreement--most States, after lengthy
deliberation, have chosen to adopt a model law developed by the
National Conference of Insurance Legislators, and that model
recognizes the benefits to consumers of using credit-based
insurance scores, but prohibits using credit information as the
sole basis for increasing rates or denying canceling or failing
to renew coverage.
The model act also includes a number of safeguards,
including prohibiting insurers from taking an adverse action
against an insured with no credit history. In other words,
recent immigrants with no credit history would have to be
treated as having a neutral credit score.
In closing, Mr. Chairman, as used in the insurance
underwriting process, credit scores appear to be highly
predictive of, and many times lower the cost of insurance for
consumers. I think they encourage responsible behavior, and
they are closely regulated by the States. And I think any
legislative attempt to limit or prohibit their use in
evaluating risk should be done so very carefully.
I thank you.
Chairman Watt. I thank the gentleman for his opening
statement. I now recognize Representative Gutierrez for 5
minutes.
Mr. Gutierrez. Thank you very much. I ask that my complete
statement be entered into the record.
Chairman Watt. Without objection, it is so ordered.
Mr. Gutierrez. Thank you very much. I thank Chairman Watt
for calling this hearing, and I appreciate the comments of
Ranking Member Bachus, and I thank everybody for joining us
here.
I just think that if you have a good driving record, if you
stop at stop signs, you don't go through red lights, you don't
speed, you don't crash into people's cars, and you don't let
your daughter use the car so she can let her boyfriend crash it
as they go out dating, if you act in all these responsible
manners, you should get a good insurance rate, regardless of
what your credit score might be.
Now I remember when I wasn't a Member of Congress, and I
remember going to get my first--I couldn't get a credit card,
so I had to get a store card--I remember, Montgomery Ward is
now defunct, I think. But that was the only place. And you got
$200 worth of credit there, and then you moved up to J.C.
Penney, and you got another $200 there, and you paid that
faithfully, because that was the only way to get credit. I was
a college graduate, I had a good job, I just couldn't get
credit--couldn't get a mortgage, couldn't buy a house.
But I got those two store cards. Finally, they gave me a
credit card, my first VISA credit card. And I remember that
they suspended it after 2 years just because arbitrarily they
decided one day that I had paid the bills on time, but they
just suspended it. I don't know why. I was pretty angry. I
remember calling the 1-800 number like 100 times, thinking
about how much damage I could cause, inflict some kind of
financial pain on them, because they did it for no reason.
I'm sure they wouldn't have done that if they thought I was
going to be, you know, a subcommittee chairman on the Financial
Services Committee one day. Because I still remember the credit
card company that--they didn't cancel my card, they said,
``Thank you for the $35 annual fee. Keep paying, but you can't
charge anything more on that credit card.''
Now look, we should all understand our personal experiences
and the experiences of consumers in America. I just want to
reiterate: If you drive safely, if you stop at stop signs, if
you don't speed, you don't have accidents, you maintain your
car, and you're a safe driver, that should be primarily how it
is you get scored in terms of how much insurance you pay. And I
think that should be the ultimate goal.
If there are other criterion, maybe we should try to
balance and blend them.
Thank you very much, Mr. Chairman.
Chairman Watt. Thank you for your statement.
I recognize the ranking member of the subcommittee, my good
friend, Mr. Miller, for 5 minutes.
Mr. Miller. Thank you, Chairman Watt, for holding this
hearing today. This is the second hearing we have had on the
impact of credit-based insurance scores on the availability and
affordability of insurance.
As numerous States, Federal agencies, and private experts
have concluded in studies on this topic, credit-based insurance
scores do indeed make insurance more available and affordable
for consumers.
Over 30 years ago, Congress passed the Fair Credit
Reporting Act, permitting insurers to use credit information to
underwrite insurance. Since the law's enactment, several
studies have been conducted on credit-based insurance scores,
showing a strong correlation between credit history and the
likelihood of filing insurance claims.
The credit information enables most consumers to qualify
for lower insurance rates, since most consumers have good
credit. Insurance companies have even reported that credit
scoring may in some cases counter-balance the imperfect driving
record of individuals.
After questioning the legitimacy of using credit scores to
underwrite risk, and expressing concerns that the scoring
method was discriminating against minorities, Congress directed
the Federal Reserve Board and the Federal Trade Commission,
FTC, to study the effects of this practice on credit and
insurance markets, and report their findings to Congress in the
2007 FTC study on the use of credit reports and automobile
insurance, and the Commissioners confirmed that credit scores
are accurate and objective predictors of risk.
That conclusion drawn by the FTC showed that for
financially responsible consumers, credit scores decreased
insurance rates. The FTC also confirmed that credit scores make
insurance more available for many riskier consumers, for which
insurance would not otherwise be able to be determined an
appropriate premium.
The FTC disproved concerns that insurance scores somehow
serves as a proxy for race, finding that ``credit-based
insurance scores appear to have little effect as a proxy for
membership in racial and ethnic groups in decisions related to
insurance.'' Further, the Commission found that insurers do not
use risk models that contain information about race, ethnicity,
or household income.
The Federal Reserve Board reported similar findings in
their study last year on credit. The Board concluded that
credit scoring likely increases the consistency and objectivity
of risk evaluation, thus helping diminish the possibility that
credit decisions would be influenced by personal
characteristics or other factors prohibited by law, such as
race or ethnicity.
The favorable study by the FTC and the Federal Reserve
Board regarding the beneficial use of credit scores have been
echoed by similar findings in the States. For example, the
Texas Department of Insurance conducted extensive research on
credit scores and reported that there is no way to determine
race, ethnicity, gender, age, or economic status by checking a
person's credit information.
The Texas study also found that drivers with good credit
are involved in 40 percent fewer accidents than those with poor
credit. In addition, homeowners' insurance claims for people
with bad credit are triple that of people with a better credit
history. In fact, the vast majority of States have thoroughly
examined the use of credit risk insurance scores and approved
their use for pricing risk.
After years of deliberation and study, the National
Conference for Insurance Legislatures, NCOIL, established a
model allowing the use of credit information in personal
insurance as long as it is not the sole factor used in
underwriting. The NCOIL model has been adopted by 26 of the
States and prohibits insurers from denying, canceling, or non-
renewing coverage due only to credit history.
According to the NCOIL in most of these States, insurers
are unable to deny consumer's insurance based on a thin credit
history or no credit at all. The FTC's conclusion, studies on
auto insurance involved a research team of career Ph.D.'s,
economists, and consultations with communities, civil rights,
consumers, and housing groups, government agencies, and private
companies, examination of records, and assurances of
reliability and independently tested data. Facts and facts and
conclusions are comprehensive and incontrovertible.
Yet after the study was concluded, several of my colleagues
were unsatisfied with the result and challenged the
Commission's data gathering, insisting that the FTC subpoena
further information from insurers. The purpose of this action
is unclear to me, considering the fact the Commission testified
in October that ``The insurance industry was cooperative and
forthright with the FTC throughout the process of gathering
data and analysis.''
They further testified of the extensive cost and drain on
resources to the Commission. In more recent discussions with
the FTC, I have learned that extensive automobile studies have
already cost millions of dollars--that is millions of taxpayer
dollars--and that the compulsory request for data from insurers
for the homeownership study would cost taxpayers as much as
double or triple the amount we have already paid.
I am glad we are going to have the hearing again. I hope
that maybe some new information has been gathered. I don't know
if that is going to be the case; but I just have a concern when
we are using subpoena powers on an industry that the testimony
to-date has said has been cooperative. And I also have a
concern about the privacy of the information we might gather in
the future. The Freedom of Information Act is very broad, and I
am concerned that might apply here.
I yield back. Thank you.
Chairman Watt. I thank the gentleman for his comments. And
while that is not the issue directly today, we had an extensive
discussion about that at the last hearing. That is the FTC's
decision. We have not directed them to do anything; we just
asked them to get us good information. And if they decide that
they need subpoenas, fine; if they decide that they do not need
subpoenas, if they can get us a good report that tells us what
the impact of credit-based scoring is, then that is for the FTC
to decide. But we are not trying to micromanage that; I want to
assure the ranking member of that.
Are there any other members who wish to make opening
statements? I am just trying to get a gauge, so I know how much
time to divide--one, two, three on this side; and one on the
other side. Okay.
I recognize Mr. Green for up to 5 minutes, if he chooses,
and then we will go to the other side.
Mr. Green. Thank you, Mr. Chairman. And if I may, let me
start by thanking you for allowing me to become a part of the
committee. I thank you and the ranking member for accepting me
as a member. This is my first hearing. And to you and the
ranking member, I greatly appreciate your having this hearing,
because it is something that has been of concern to me for some
time.
I especially thank the ranking member for his comments. He
and I have had many conversations, and I have found him to be a
person who is principled and who moves forward based upon what
he sincerely believes to be the case. Notwithstanding the fact
that he and I may have differences, we do have one thing in
common, and that is we enjoy our conversations with each other
about our differences.
My concern with this is as was indicated previously: The
connectivity between one's credit score and one's driving
habits. I am hoping to hear information that can help me to
better understand that relationship between one's credit score
and one's driving behavior.
I especially concern myself with this because we have young
drivers who have no credit scores. It is not unusual for
parents to add their children to their insurance, and these
children, generally speaking, have little or no credit. How is
it that they will inherit the credit of the parent and become a
risk by virtue of having been born into a certain family? Is
that the way it will work? Do they have a different standard
for young drivers who have no credit score, who have not had a
track record of driving at all, but perhaps they have been to a
driving school and they have had all of the safety courses,
such that one might conclude they understand the rules of the
road? They don't have a track record of poor behavior. I don't
see the connectivity between such a person and a credit score.
It seems to me that if we are not careful, we are going to
make it almost impossible to be poor in the richest country in
the world. The richest country in the world; 1 out of every 110
persons is a millionaire. But it costs to be poor in America.
You pay more for your insurance; you ride on roads that will do
more to your vehicle because of where you live if you are poor,
generally speaking. You will probably have to get more wheel
alignments. You probably go a store that has prices that are
higher than in some other neighborhoods.
And I think that at some point, we have to examine the
notion of whether we ought to do things just because we can.
Maybe you do have the right to do it, but the question is: Is
it the right thing to do?
I am looking for the cause of connection between a credit
score and one's driving behavior.
Mr. Chairman, I thank you for the time, and I yield back.
Chairman Watt. I thank the gentleman for his statement. The
gentleman from California is recognized for 3 minutes.
Mr. Royce. Four minutes, Mr. Chairman?
Chairman Watt. Four minutes. Okay.
Mr. Royce. Thank you, Mr. Chairman. I appreciate you
holding this hearing.
I would also just like to express my opposition to the
concept here of banning the use of credit-based insurance
scores, because from the studies I have seen, this is a very
effective predictor of the actual risk. It is a predictor of
the number of claims that the consumers file; it matches the
total cost of those claims. Credit scoring is not based on
race. And I think the FTC's report, which came out in July,
explained this benefit as have other studies. We have seen a
number of studies on this subject.
In the competitive marketplace which exists throughout most
of the auto insurance sector, companies have an incentive to
provide the lowest actuarially sound rates for the customers.
In most instances, a potential customer can get several quotes
on auto coverage in a matter of minutes over the Internet or by
picking up the phone.
Companies have even began to offer their prices along with
the prices of their competitors in the names of attracting
additional business. If there are inefficiencies, if there are
gaps in coverage, I think a logical place to look would be the
current State-based insurance regulatory system. With the
exception of Illinois, every State subjects property and
casualty insurance products to varying degrees of government
price controls. And of course, that discourages companies from
operating effectively and efficiently in those States.
Additionally, the bureaucratic delays weigh heavily on the
rates paid by consumers. The American Consumer Institute
recently found that the cost of excessive regulation at the
State level is $13.7 billion annually, paid for by insurance
buyers through higher premiums. If Congress really wants to
improve the ability of consumers with weaker credit histories
to obtain more economical quotes on insurance coverage, we
should be looking at ways to bring more competition to those
markets.
In the Wall Street Journal, on May 6th, there was an
editorial on the Massachusetts Miracle, and that highlighted
the recent move by Massachusetts to remove its government-set
rates on auto coverage, and as the editorial noted, Progressive
Insurance, the third largest insurer in the country, entered
the market May 5th with rates 18 percent below the old price-
controlled rates. Overall, premiums in the State are going to
fall 8 percent this year as insurers adjust to a world in which
they need to compete to attract customers instead of bargaining
with their regulator for price hikes.
If more States saw the economic implications of price
controls, or if Congress would consider our legislation to
create an optional Federal charter, a greater number of
consumers, including those this legislation was intended to
help, would be on the receiving end of more products, and
certainly with much lower premiums.
So in closing, I would caution my colleagues against
enacting legislation which leads to banning the use of credit
scores by insurance providers as one of the many factors
included when setting premiums. I believe the majority of
consumers would see higher costs for insurance products if that
happened, because their provider would not be able to set
actuarially sound premiums.
And again, Mr. Chairman, I offer this other alternative,
and I would like to thank you for holding this hearing. I look
forward to the testimony, and I appreciate the witnesses coming
out to speak to us today.
Thank you, Mr. Chairman
Chairman Watt. I thank the gentleman for his presence and
for his opening statement. I recognize my colleague from North
Carolina, Mr. McHenry, for 3 minutes.
Mr. McHenry. I appreciate the chairman's recognizing me.
And I do appreciate him holding this hearing as well.
I think we should have a discussion about how to improve
and how to accurately assess credit risk in all financial
service products. But it is interesting here that the
discussion is about whether or not a credit score should be
used, and it is but one of the tools in the tool chest to
assess risk.
From what I have read in some studies, it is one of the
most effective ways of assessing risk. Insurance is not simply
an individual's right, but it should be the ability of the
company to accurately assess risk, so that they can more
accurately seek payment for that. And I think as such, credit
scores are a worthy example of the way an insurance company can
assess risk.
I don't think it should be the be-all end-all, and from
what I understand from the industry, it is not.
But I would just go back to my experience in college with
credit cards. My experience is pretty simple. You know, you go
and rack up the credit card debt, which I did, buying
cheeseburgers, pizza, and many other things in college, but I
had to pay the consequences for that.
And my credit score reflected that, and as such, I was a
greater credit risk because of how much fun I had in college,
and how I paid for it. And I think that is a fair assessment of
how this works.
I think we should go to an additional step--and I would be
happy to work with the chairman on this--I do think the issue
is not about the insurance industry using credit scores; I
think it should be about how these credit scores are derived.
There are a number of different items that are not included
in a credit score that could better assess risk for
individuals. For instance, most of us have to pay a power bill
every month. I think that would be a positive credit indicator.
And I think if the insurance companies could see that they
regularly pay their power bill every month, and have never
missed a payment, maybe that would be a stronger indicator
rather than their overall credit score on whether or not they
will pay for their insurance, and be a greater risk.
I think that's a fair assessment. I think we should look at
credit scoring rather than really I think a symptom of the
underlying disease, which is how these credit scores are
derived. I think that's a positive thing; I think we could have
some bipartisan support, and I look forward to working with the
chairman on those items. Thank you so much for having the
hearing today.
Chairman Watt. I thank the gentleman for his opening
statement, especially his confessions of his college years. I
am glad he cut it off where he did.
[Laughter]
Chairman Watt. I recognize the gentleman from Illinois, Mr.
Roskam, for 3 minutes.
Mr. Roskam. Thank you, Mr. Chairman. And Mr. Chairman,
thank you for holding this hearing today. I attended the last
hearing, and I understand where the Majority is coming from in
holding the first hearing. And that is, it's a pretty
interesting narrative, that if you can thread the pearls to
suggest that there is a racial component to a predominant
American industry, manipulating a marketplace on the backs of
minority groups, that is powerful. That would be outrageous,
and all of us would be outraged, and we would be like-minded
and say, ``That ought not to happen.''
But as I listened to the testimony last time, and
particularly the study from the Federal Trade Commission, what
I heard was essentially that it wasn't happening that way.
There were some consumer groups who were testifying, and the
more I listened--it's kind of like talking on talk radio to the
weird caller that calls in: The more you listen, the more
disjointed it starts to sound. So I kind of discounted that in
terms of testimony.
And then, as I have been thinking about this, I have come
to the conclusion that there are a lot of similarities between
credit scoring and student grades and good student discounts. I
mean, is there a relationship between someone's driving record
and their performance on a history test? Is there a
relationship between someone's driving record and their
performance on their calculus final? Is there a relationship
between someone's driving record and their performance on their
English composition? Well, we can't really articulate what it
is, but it just so happens to be that it always sort of seems
to work out, and that it is a predictor.
So as I was listening to the hearing last time, and I'm
just doing research as to this other hearing that has been
prompted, in Illinois, as it turns out, there is a carrier in
Illinois that is using this, and they are actually increasing
their book of business into South Chicago, which is a
predominantly minority community.
And so I think what we do today--if this sort of goes the
direction that I think it might go--what we do is we risk
taking away tools from carriers to offer more coverage to more
people, regardless of race and ethnicity and the unintended
consequence, I think, becomes a self-fulling prophecy, and it
becomes more difficult for folks to get the type of coverage
they need. They are pushed into more residual markets. They are
forced to go with the substandard insurance carriers with the
great names. What I have learned is the more glorious the name
of the insurance company, generally the worse the coverage is,
and that, I think, is where we ought not to go.
So I come with an open mind as well. But I also come,
having listened to the testimony of the last hearing and being
completely underwhelmed, and hoping that this bodes better in
terms of the things that we are able to conclude.
I yield back.
Chairman Watt. I thank the gentleman for his opening
statement, and I hope he is not underwhelmed. I thank him for
being at the earlier hearing, as well as today's hearing, and I
think the audience and the witnesses recognize that there is a
range of opinions on this issue, and a willingness and openness
to understand how this system works, so that we can make good
public policy. That is, after all, the reason we have these
hearings, to try to get more information about what is
happening and what the real life impacts are.
So with that, are there any other members who seek to make
an opening statement? We have probably gone a little beyond
what we would ordinarily do in opening statements at a
subcommittee level, but this is an issue that even the
attendance suggests is an issue that people recognize as
important. And so I apologize to all of those in attendance if
they haven't wanted to hear these opinions, but it sets the
basis for our moving forward.
Without objection, all other members and members who have
made opening statements, their full opening statements will be
made a part of the record, if they wish to submit opening
statements.
We will now introduce the members of the first hearing
panel, and without objection, the witnesses' written statements
will be made a part of the record, and each witness will be
recognized for a 5-minute summary of their testimony.
I am going to recognize my good friend from Florida, Mr.
Klein, to do his ``all-politics-is-local'' introduction of his
State insurance commissioner. Mr. Klein?
Mr. Klein. Thank you, Mr. Chairman.
I appreciate that opportunity, having served in the Florida
legislature for 14 years and having the privilege of serving
with one of our panelists today, Kevin McCarty, who is the
commissioner of the Office of Insurance Regulation in Florida.
We have been faced with a number of complicated insurance
issues in Florida, some of which have been taken up by this
committee, and of course today's issue is just another one that
requires some expertise of a broad variety. I think that Mr.
McCarty, with his work in our Department of Labor and
Employment Security, and his work on worker's compensation
issues, will be very helpful.
He has worked in our department for many, many years. He
helped the investigation and response following the devastation
of Hurricane Andrew. He became our first insurance
commissioner, appointed in 2003, and has served in that
capacity ever since, but particularly for today's purposes, he
is very active with the National Association of Insurance
Commissioners, which as we all know, is our 50-State member
organization that gives us the State perspective, and it is
very valuable when we are establishing Federal policy.
So I just want to welcome Commissioner McCarty, and I look
forward to his and our other panelists' comments.
Chairman Watt. I thank Mr. McCarty for being here also. I
will proceed with introducing the other two witnesses on the
first panel, and then I would like to go back and take Ms.
Waters' opening statement, if that is okay with the members.
The first witness on this panel is Ms. Lydia Parnes, the
Director of the Bureau of Consumer Protection at the Federal
Trade Commission. All of the Commissioners were tied up in a
meeting today, and asked us to allow Ms. Parnes to testify on
behalf of the FTC, and we told them that we thought she would
do a better job anyway.
[Laughter]
Chairman Watt. So we thank her for being here.
The third witness on this first panel will be the Honorable
George J. Keiser, State Representative of the State of North
Dakota, who will be testifying on behalf of the National
Conference of Insurance Legislators. We welcome all of the
witnesses.
Without objection, I would like to deviate and go back and
take the opening statement of Ms. Waters, who was just able to
get here. We thank her for being here; she is the lead sponsor
of one of the two bills that we are having the hearing about
today. I recognize the gentlelady for 5 minutes.
Ms. Waters. Thank you very much, Mr. Chairman. I certainly
thank you for convening this second hearing on the impact of
credit-based insurance scores on the availability and
affordability of insurance.
The first hearing you held on this topic last October was
very enlightening, but also troubling. In fact, I was so
disturbed by some of the testimony that I, along with Mr.
Gutierrez, introduced H.R. 6062, the Personal Lines of
Insurance Fairness Act of 2008, to ban the practice of using
credit scores in the underwriting or rating of insurance
premiums.
I am looking forward to hearing our witnesses' testimony on
this topic, but I must say that the findings from the first
hearing deeply concerned me. The hearing covered a report
released in July 2007 by the Federal Trade Commission. The
report found that credit-based insurance scores, which are
developed and used by the insurance industry, serve as a proxy
for race in three out of four lines of automobile insurance.
Specifically, the report found that when credit-based
insurance scores are used to predict claims risk, the predicted
risk of African Americans and Hispanics increases by 10 percent
and 4.2 percent, respectively. Conversely, the predicted risk
for whites decreases by 1.6 percent.
To address the proxy issue, Mr. Gutierrez and Mr. Watt
introduced, of course, as you have already said, legislation
that would prohibit the use of credit scores for insurance
underwriting when a proxy effect is found.
However, I must disagree with this approach. While we must
do something to address the disproportionate racial impact of
this practice, I am also concerned about the overall fairness
of this practice. Specifically, credit scores have little, if
no bearing on how likely a person is to have a car accident, to
break speed limits, or to otherwise engage in risky driving
behavior that could result in an insurance claim.
I know that the industry maintains that there is some
correlation between low credit scores and increased claims
risk; however, a correlation does not imply causation.
I wonder if we would permit other possible correlations, no
matter how unrelated to claims risk, to be used to set
insurance premiums. For example, if research is found that
there was a correlation between zodiac signs and increased
claims risks, would it be appropriate to allow such a
correlation to be used as a metric for setting insurance
premiums?
To make someone pay more for insurance because of a
situation in their financial circumstances that has nothing to
do with their risk as a poor driver or irresponsible homeowner
is simply unfair. It is simply unfair. It is unfair to recent
immigrants, to the elderly, and to low-income Americans, all of
whom have little credit history.
Furthermore, it is unfair to those Americans who have been
hit by the foreclosure crisis, and are now struggling to
rebuild or to re-establish their credit.
I could go and on, talking about whom all it is unfair to,
but recently, friends of mine were hit with an extraordinary
health crisis. They had paid their bills all of their lives and
done well, and because of the burden that were confronted with,
they fell behind in their payments. And of course, their credit
scores went down.
They are good people. Should that credit score have any
impact on their ability to purchase insurance? I don't think
so. Traditional underwriting standards worked with little
problems for several decades before insurance companies began
using them for underwriting purposes.
I am interested to hear our witnesses explain why these
standards were abandoned, and how they continue to justify the
use of credit scores for underwriting, given the concerns I
have raised.
Thank you, Mr. Chairman. I appreciate your accommodating my
coming in a little bit late, and I will yield back the balance
of my time.
Chairman Watt. I thank the gentlelady for being here both
for the first hearing and for this hearing and for her proposed
legislation.
We are now ready to recognize the witnesses, and each one
of you will be recognized for 5 minutes to give a summary of
your written testimony. The green light will come on at the
beginning, at 4 minutes a yellow light will come on, and at 5
minutes a red light will come on. We would ask you, at that
point, to wrap up the thought that you are involved in. We do
have a second panel and a number of members who wish to ask
questions, so we want to try to keep this moving if we can.
With that, I will recognize Ms. Lydia Parnes, Director of
the Bureau of Consumer Protection at the Federal Trade
Commission for a 5-minute opening statement.
STATEMENT OF LYDIA B. PARNES, DIRECTOR, BUREAU OF CONSUMER
PROTECTION, FEDERAL TRADE COMMISSION
Ms. Parnes. Thank you very much. Chairman Watt, Ranking
Member Miller, and members of the subcommittee, I appreciate
the opportunity to--
Chairman Watt. Can you pull that microphone a little bit
closer to you? And if somebody has an empty seat beside them,
would they just kind of raise their hand, so that others who
are standing might be able to take a seat? I think there are
enough seats in here for everybody who is standing, unless you
just want to stand. But if you do, I wish you wouldn't stand,
blocking the door.
Pull the microphone very close to you, because I was having
a little trouble hearing you. And make sure it is on.
Ms. Parnes. Is it working now?
Chairman Watt. Yes.
Ms. Parnes. Better?
Chairman Watt. Thank you.
Ms. Parnes. Okay. Thank you.
I do appreciate the opportunity to appear before you today,
as you consider the impact of credit-based insurance scoring on
the availability and affordability of insurance. As members of
this subcommittee are aware, insurance companies have
increasingly used credit-based insurance scores to decide
whether and at what price to offer automobile and homeowners'
insurance to consumers.
Industry representatives and other proponents contend that
by using these scores, insurance companies charge consumers
premiums that conform more closely to their individual risk of
loss. However, consumer advocates, civil rights groups, and
others believe that the use of these scores results in racial
and ethnic minorities paying higher insurance premiums than
other consumers.
To provide insight on the effect of credit-based insurance
scores, Congress, in FACTA, directed that the Commission study
the effect of these scores on the availability and
affordability of insurance, including the particular impact on
racial and ethnic minorities.
In 2007, the Commission released a report discussing the
results of a study of the use and effect of credit-based
insurance scores on consumers of automobile insurance. The FTC
provided the subcommittee with views about this report during
its testimony last October. Today, I am pleased to provide an
update on the FTC's ongoing study on the use and effect of
credit-based insurance scores on consumers of homeowners'
insurance.
Last week, the Commission approved a resolution authorizing
the use of compulsory process to obtain data for this study.
The FTC intends to issue orders to the nine largest homeowners'
insurance companies, representing roughly 60 percent of the
market of private homeowners' insurance in the United States in
2006.
The FTC has placed on its Web site a draft order setting
forth in detail the information it intends to seek from
homeowners' insurance companies. The Commission is seeking
public comment for 30 days on this draft order consistent with
FACTA's direction that the Agency consult with consumer groups,
civil rights and housing groups, government officials, and the
public at large on the design and methodology of these studies.
After receiving public comments and making appropriate
revisions, the Commission will serve orders on the nine largest
homeowners' insurance firms in the United States. The FTC would
be pleased to keep the subcommittee and its staff informed as
the study progresses.
I know, as you have mentioned, this subcommittee is
considering two bills addressing the use of credit-based
insurance scores. H.R. 5633, the Nondiscriminatory Use of
Consumer Reports and Consumer Information Act of 2008, would
amend the Fair Credit Reporting Act to prohibit the furnishing
or use of a credit-based insurance score if the Commission
determines that the use of scores results in racial or ethnic
discrimination or represents a proxy or proxy effect, per race
or ethnicity. H.R. 6062 would ban the use of credit scores in
insurance underwriting.
The FTC has a longstanding commitment to law enforcement
and education efforts in fair lending, and believes that it is
vitally important to protect consumers from illegal
discrimination based on race or ethnicity.
The Commission, however, has deferred to Congress as to
what legislative measures, if any, are appropriate in this
area. I would note, however, that from a purely drafting
perspective, H.R. 5633 would impose liability based on the
determinations of FTC econometric research studies. As
discussed in greater detail in the Commission's testimony, the
FTC has concerns about using its studies as a trigger for
liability.
Thank you for your attention, and I would be pleased to
answer any questions that you may have.
[The prepared statement of Ms. Parnes can be found on page
173 of the appendix.]
Chairman Watt. Thank you so much for your testimony.
Commissioner McCarty, you are recognized for 5 minutes.
STATEMENT OF THE HONORABLE KEVIN MCCARTY, INSURANCE
COMMISSIONER, STATE OF FLORIDA, ON BEHALF OF THE NATIONAL
ASSOCIATION OF INSURANCE COMMISSIONERS (NAIC)
Mr. McCarty. Chairman Watt, Ranking Member Miller, and
members of the subcommittee, I want to thank you for the
opportunity to testify on the use of credit-based insurance
scores and the provision of personal line insurance products. I
am Kevin McCarty, the Insurance Commissioner in Florida. I am
also here representing the National Association of Insurance
Commissioners.
Proponents have argued over the years that credit scores
are predictive of the future losses based on the insurance
claims experience, and are a necessary and inexpensive
underwriting tool. Critics argue that the use of credit scores
discriminates against protected classes of people.
Technology over the years has allowed insurance companies
access to enormous amounts of new information, including credit
reports. Although some of this information may show actuarial
relationships with insurance claims, this does not
automatically make it an appropriate, fair, and valid criteria
for insurance purposes.
The most notable example of this is the use of race-based
rates. In 2002, the NAIC concluded several multi-State
investigations on companies that historically rated life
insurance differently based on the race of the applicant.
Even today, Caucasians born in the United States have a
longer life expectancy than African Americans. Based purely on
this actuarial science, this would indicate a higher premium
for life insurance. While the outcome of African Americans
paying more is correct from an actuarial perspective, it is
certainly counter to equal protections for Americans and is an
abhorrent public policy.
The use of credit reports represents many potential
problems. Consumer report studies show that 50 percent of the
credit reports contained errors, which can be exacerbated today
by the increased amount of identity theft and the proliferation
of our access to credit.
Thus, even if the methodologies were correct, it is
possible that inaccuracies in the reports may in fact
invalidate their use.
Credit reports also disproportionately and negatively
affect the recently divorced, recently naturalized citizens,
the elderly, those of certain religious beliefs that do not
believe in the use of credit, and younger individuals who have
not established credit histories.
The overwhelming problem with the use of credit scoring is
the relationship between credit scores and race, ethnicity, and
income.
The 2004 Texas Insurance Department study previously
referenced that African Americans have an average credit score
of 10 to 35 percent below that of Caucasians. Hispanic scores
were roughly 5 to 25 percent below.
I do not believe the insurance industry uses credit scoring
to intentionally discriminate or impact minorities. Yet, recent
empirical studies demonstrate a negative impact on these
protected classes.
I am also concerned about other tools that share many of
the same characteristics of using credit scoring. A year ago, I
held a public hearing in Florida on occupational and
educational rating as an underwriting factor for private
passenger autos. Testimony at the hearing and information
gathered as a result of that indicated that insurers would
refuse to study the underwriting practices on minorities and
low-income consumers.
I'm especially troubled by the growing use of occupational
and educational rating, and would encourage the subcommittee to
broaden the scope of its investigation to consider these unfair
regulatory practices.
The 2007 FTC report was very disappointing. The narrative
appeared very one-sided in support of the predictive powers of
credit scoring, while equally downplaying the negative impacts
on protected classes of citizens.
I did agree with one aspect of the FTC report, that the
State insurance regulatory community has identified credit
scoring as a problem and has taken action. As previously
mentioned, 48 States have passed some legislation limiting the
use of credit scoring. Many States have adopted laws that
require regulators to have access to the internal operations of
the credit-scoring models, that the decisions are not based
solely on credit reports, and that consumers be notified of the
use of these reports, and if there is any adverse decisions
based on their credit scores.
It is my sincere desire that the Federal Government assist
the States in its regulatory efforts to address this important
issue and better protect our consumers.
The proposed bill, H.R. 5633, has many favorable
provisions. My colleagues around the country and I welcome a
more comprehensive study by the Federal Trade Commission to
determine if the use of credit reports disparately impacts
minorities and does in fact create a proxy effect.
I am also personally in favor of H.R. 6062, which
implicitly accepts the notion that credit scoring disparately
impacts minorities based on a 2007 study.
Thank you for holding this hearing and for inviting me to
participate. I look forward to your continued leadership on
this very important consumer protection issue.
[The prepared statement of Commissioner McCarty can be
found on page 112 of the appendix.]
Chairman Watt. Thank you so much for your testimony.
Representative Keiser, State Representative, State of North
Dakota, is recognized for 5 minutes.
STATEMENT OF THE HONORABLE GEORGE J. KEISER, REPRESENTATIVE,
STATE OF NORTH DAKOTA, ON BEHALF OF THE NATIONAL CONFERENCE OF
INSURANCE LEGISLATORS (NCOIL)
Mr. Keiser. Chairman Watt, Ranking Member Miller, and--
Chairman Watt. Pull that microphone closer to you.
Mr. Keiser. Thank you very much for inviting NCOIL to
participate in this process.
Using objective methods, which are blind to ethnicity,
gender, income, and other factors, credit scoring may offer a
consistent, accurate, and valid way to underwrite and rate
risk, and may mean lower prices for many consumers, primarily
those with lower risk.
However, NCOIL has taken a position that as State
legislators, we are concerned about any abuses that might occur
relative to the application of the credit scores. We encourage
laws that understand and accommodate and benefit consumers. For
example, our model has looked at the impact on seniors, has
looked at the impact on young people, has addressed the
situation where people have an extreme financial crisis occur
in their life, and we have attempted to adopt that and address
that.
There are 26 States which have currently adopted the NCOIL
model that has been developed and it appears to be working
relatively well in those States. We believe that an appropriate
approach is to allow the States to take the NCOIL model and to
modify it when appropriate for their States.
Well, what is the NCOIL model and what does it do? The
NCOIL model is non-discriminatory. It assists the young, old,
and those who suffer extraordinary events, and requires the
provision of updated credit information. It goes beyond Federal
law by prohibiting insurers from calculating scores based on
income, gender, address, zip code, ethnic group, religion,
marital status, or nationality. It also prohibits denying,
canceling, or non-renewing coverage due solely to a credit
score, or from basing renewal rates solely on credit.
Consumer protection under the NCOIL language, an insurer
must use data taken within 90 days from the time of adverse
action. It must be disclosed to the consumer that when adverse
action is taken, a consumer has the right to appeal or object
to it. The insurance companies are required to review any
objection and to address it.
Relative to consumers, for young people, for old people who
don't have credit--many of the panelists in their opening
comments addressed that--the NCOIL model requires either in the
cases of what would term ``thin credit'' that the credit be
treated either as neutral on the credit score or in a positive
manner.
Inquiries are another big issue relative to credit scores.
The NCOIL model offers common sense restrictions on how
insurers can treat inquiries the credit card companies make
before sending out promotional offers; inquiries based on
consumers wisely shopping around for deals on auto and home
loans; collection accounts related to sickness or other medical
events; and bad credit resulting from extraordinary events like
divorce, illness, or death of a spouse, as mentioned earlier.
The NCOIL model says that insurers can give these
extraordinary victims a credit pass in those situations. The
insurer must re-underwrite and re-rate using new data. If the
consumer has overpaid as a result of a mistake made, then they
are eligible for a credit or refund for that amount.
If the insurer does take an adverse action due to credit,
the insurer must give up to four good reasons why. The insurer
must be clear up-front that credit will be used.
In conclusion, we appreciate the work of the subcommittee
to ensure that credit history is used fairly. The 26 States
regulating credit scoring based on the NCOIL model have
responded effectively to an issue demanding a timely solution.
States as diverse as New York, North Dakota, Texas, and Maine
have successfully used this model to meet their different
demographics.
We ask that you recognize the efforts States have made to
balance consumer protection with the need for healthy insurance
markets and that one-size-fits-all doesn't work.
Federal legislation that would satisfy the laws of these
States is unneeded and may actually bring higher rates for
consumers who are benefitting from their good credit.
Thank you for the opportunity to appear, and NCOIL looks
forward to working with your committee, Mr. Chairman.
[The prepared statement of Mr. Keiser can be found on page
101 of the appendix.]
Chairman Watt. Thank you so much for your testimony, and I
thank all of the witnesses for being here today.
I will now recognize each member of the subcommittee for
questions of this panel, and I will recognize myself for 5
minutes. I may be a little aggressive in enforcing the 5
minutes against us, since we have a second panel to go after
this panel, also.
Representative Keiser, how, if at all, would you
distinguish between this, the use of credit scoring here, and
the public policy position that we have taken with respect to
life insurance, where there is an actuarial, predictive
documented relationship? We have said that as a matter of
public policy, this is unacceptable; and even more recently in
the House at least, and I think maybe even in the Senate, we
have passed a bill that prohibits genetic information from
being used. How do you distinguish this from that, if you are
able to do that?
Mr. Keiser. Well, Mr. Chairman, I'm not sure that I am 100
percent qualified to answer that question. But let me say that
NCOIL, as the policymakers in the State on insurance issues, is
extremely concerned in protecting valid predictors of risk,
whatever they might be--
Chairman Watt. Even if it is race--
Mr. Keiser. Let me just finish. If they can validly predict
risk--and I question, although it's done, that grade point
averages for high school students can be a valid predictor of
risk for insurance companies, that family history can be a
valid predictor for health, when I have an application for life
insurance or health insurance; that age can be a valid
predictor--those are all valid predictors, we are committed to
protecting the industry's opportunity to use valid predictors
and at the same time protecting the consumer to ensure that
invalid application of predictors doesn't occur.
In the NCOIL model, what we attempted to do--
Chairman Watt. I understand that. You are going back, and I
only have 5 minutes, so I am not trying to argue with you on
this; I can't distinguish these things.
Let me also, just as a factual backdrop, get you and Mr.
McCarty, if you would, to distinguish between--or is there a
distinction between being a valid predictor of risk versus a
valid predictor of claims? Which one have you determined that
credit scores are valid predictors of?
Mr. Keiser. Mr. Chairman, I'll answer first, and then
Commissioner McCarty can answer.
Chairman Watt. Go right ahead.
Mr. Keiser. It is my understanding that it is a valid
predictor of claims that--
Chairman Watt. Okay--
Mr. Keiser. That credit score is valid predictor of
claims--
Chairman Watt. Okay. And that is different, is it not, from
being a valid predictor of risk?
Mr. Keiser. Mr. Chairman, from the standpoint of risk of
having an accident or something, yes, I would agree. But from
an insurance company's management standpoint, I would argue
that the claim itself is the exposure to risk that the
insurance company has.
Chairman Watt. So if there were other factors that kept
people, even if they were in automobile accidents from making a
claim, the fact that this predicts their willingness to make a
claim, which is what the insurance policy was written for,
would be acceptable under what you are saying.
Mr. Keiser. If I understand your question--I'm not sure,
Mr. Chairman--but my insurance agent oftentimes tells me,
``Even though you have a claim, you sometimes are better off to
pay that auto damage yourself, because it is a relatively minor
claim, than to apply it to your policy and do the deductible
plus $100 or $200.'' I am well-served from a responsibility
standpoint--
Chairman Watt. You are well-served, but would a poor person
who didn't have the option of paying that claim himself be
well-served? I guess that is the question. Can you just comment
on that, Mr. McCarty? And then I am going to--
Mr. McCarty. Thank you, Mr. Chairman. I agree with my
colleague that the measurement is a measurement of claims.
There's no evidence to suggest that regardless of your credit
score, you have more accidents. Certainly a reasonable analysis
of that data is that credit scoring is really a proxy for your
economic class, or your income; and as a consequence, if you
have lower income, you are not able do to as the Representative
had suggested, to pay out of pocket. And as a consequence,
actually people with more money will forego making a claim,
knowing that their insurance premiums may go up in the future,
and they have the ability to make that economic choice. Lower-
income people do not have that option.
And so what is interesting about the analysis done by the
Federal Trade Commission is addressing the issue of claims,
which I think can be reasonably explained by you having enough
wealth to pay for those claims out of pocket.
Chairman Watt. I recognize the ranking member for 5
minutes.
Mr. Miller. Thank you very much.
Ms. Parnes, you talked about the individuals, the Ph.D.'s
in economics, and you consulted with community groups and civil
rights groups and such in your report. Is it fair to say that
you're confident in the integrity of your initial report, that
you examined the analysis you performed and the findings are
correct in that report?
Ms. Parnes. The Commission definitely was confident in the
reliability of its initial report. And you know that one of our
Commissioners did dissent from that.
Mr. Miller. Is it fair to say that the FTC doesn't support
the legislation, that it would ban credit scores?
Ms. Parnes. The Commission hasn't taken a position on the
legislation.
Mr. Miller. Okay. You heard Chairman Watt say that nobody
has asked you to use subpoena power. When we previously talked
to the FTC, they said that the industry was very cooperative in
providing information necessary to prepare the report. Is that
correct?
Ms. Parnes. The industry has been cooperative--
Mr. Miller. Then why would you use subpoena powers?
Ms. Parnes. FTC studies are important, we think, both in
terms of the actual reliability of the study, and also in the
perception of its reliability--
Mr. Miller. So it's not for the quality of the material;
it's for perceptions reasons that you are doing it?
Ms. Parnes. Well, there was certainly a lot of concern
expressed about our initial report and whether it was adequate,
because we obtained the information voluntarily and for a host
of procedural reasons. We feel that by using our subpoena
authority, we can address those concerns.
And I should add that we use subpoenas often when we are
collecting information in studies, and it certainly isn't meant
to suggest that the industry that we're working with is in any
way uncooperative.
Mr. Miller. Did you review the Federal Reserve Board study?
Are you familiar with it?
Ms. Parnes. I am somewhat familiar with the Board's study.
Mr. Miller. Because it found that some elderly have better
scores, and if this legislation is passed, it would actually
harm the elderly. Is that a fair statement?
Ms. Parnes. I know that it found that the results were
somewhat similar to the results that the Commission--
Mr. Miller. Okay. Mr. McCarty, you said you're representing
the National Association of Insurance Commissioners? Do they
agree with your opinion? They supported that? Have they
supported your opinion today?
Mr. McCarty. The National Association of Insurance
Commissioners supports the testimony today with regard to the
need to continue to study this issue, and is deeply concerned
about--
Mr. Miller. Yes, but that's not what you said. You spoke
against it. And if this legislation is passed, it would
overturn every State law except Hawaii's. Even your own States
would pre-empt it.
Mr. McCarty. That is correct. The NAIC's position is not
supporting--
Mr. Miller. But you said you were representing them, and
that's I don't really think factual, in their opinion. I looked
at data that said after all the States in the jurisdiction
reviewed the use of credit scores extensively, that it's
basically true that only one State out of 56 jurisdictions have
actually banned the use of credit scores, including yours.
Mr. McCarty. I said I personally favor H.R. 6062.
Mr. Miller. Okay. I just want to make sure the record is
very clear, that is not the position of the National
Association of Insurance Commissioners. In fact, as I said, out
of 56 jurisdictions, including States, only one State bans it,
so there is a huge difference between that and--I mean this
would overturn your own State law.
Mr. McCarty. Yes, I understand that. And I want to clear--
Mr. Miller. Well, you're going to have fun going back home,
aren't you, on this one? That is going to be an interesting
process.
Mr. McCarty. I did want to clarify that with regard to H.R.
6062, that was my personal view, not the view of the NAIC.
Mr. Miller. Yes. Okay. I have no problem with your personal
opinion. I mean everybody has a right to one; I just didn't
want a perception to be created or anybody to think that you
represented the opinion of the National Association of
Insurance Commissioners. In fact, it seems to be quite the
opposite; your own State legislators would disagree with your
opinion today, based on what they voted into law.
Mr. McCarty. The National Association supports continued
study of this issue, and is deeply concerned about the
disparate impact on minorities--
Mr. Miller. I have no problem with continued studies--that
is what we are doing today--
Mr. McCarty. And it supports H.R. 5633.
Mr. Miller. Mr. Keiser, you testified that the State
legislators were initially skeptical about credit scores, but
ultimately found that they increased availability and
affordability for consumers, and they were racially blind, and
to help insurers compete. Is that a fair statement?
Mr. Keiser. Representative Miller, that is actually as
accurate a statement as I could make regarding that subject.
Senator Craig Eiland from Texas--
Mr. Miller. My time is up, so just in closing, would you
agree that if this were enacted, it would really harm seniors?
In your opinion?
Mr. Keiser. I could not agree more strongly, and also NCOIL
opposes this legislation.
Mr. Miller. Thank you very much for your testimony.
Chairman Watt. The gentleman from Illinois, Mr. Gutierrez,
is recognized for 5 minutes.
Mr. Gutierrez. Thank you very much, Mr. Chairman. Director
Parnes, it is my understanding that the FTC shared advance
copies of its draft report with the insurance trade
associations, but not with the insurance regulatory community.
Is that the case? And if so, what was the reason behind this
decision?
Ms. Parnes. That is not the case.
Mr. Gutierrez. You absolutely deny that the FTC shared this
with the insurance industry, and not with the regulatory
community? Just answer yes or no.
Ms. Parnes. Well, certainly as far as I know--
Mr. Gutierrez. Thank you. Director Parnes, the July 2007
FTC report found--I read it--that credit-based insurance scores
are a proxy, or a substitute for race or ethnicity in three out
of four lines of automobile insurance: Collision;
comprehensive; and bodily injury. But in your written testimony
for this hearing this morning, you state that the FTC ``found
that credit-based insurance scores appear to have little effect
as a proxy.'' Your written testimony appears to be backing away
from the conclusions of the FTC's report. I hope that is not
the case, but I am going to ask you, for the record, do you
stand by the original FTC report?
Ms. Parnes. The Commission certainly does stand behind the
original report. I think that it may be worth explaining a
little about this proxy effect. Proxy, when used in usual
conversation, it's kind of like an absolute substitute,
something substitutes for another thing. It's an all or
nothing. And when we use proxy effect in the study, we were
talking about the effect, as you understand of course, of a
statistical analysis--
Mr. Gutierrez. I guess I understand that. And we only have
5 minutes. But when I read the FTC report, and I read your
comments and your written statement for this committee today,
they seem to be different. They seem to be backing away. They
seem to be kind of light, kind of like, ``Well, let me
reinterpret, let me re-evaluate what the FTC really meant when
they issued their report.''
They seem different, and I think that most people might--so
I just wanted to ask you if they're different, because I read
the original report, which gave birth to the legislation that
we're proposing. I mean we didn't just base it on thin air; we
read your report. And today it seems like, ``Well, yes, it's a
proxy, but it's no big deal. It's really not that relevant;
it's really not that important.'' That seems to be the way I
interpret what you bring to the committee today, vis-a-vis what
the committee heard when the FTC first reported.
So I just thought I would ask you.
Ms. Parnes. The Agency has no intent to back away from its
earlier report.
Mr. Gutierrez. Mr. McCarty, in your written testimony, you
referred to ``economic advantages'' to insurance companies from
using credit-based insurance scores that have largely been
ignored by empirical studies, including the 2007 FTC study.
What are these economic advantages, and why do they deserve any
scrutiny?
Mr. McCarty. Well, I think the insurance industry and
insurance trade associations would argue that using credit
scoring is an inexpensive underwriting tool, that it would be
more expensive to underwrite if they did not have the ability
to use credit scoring freely as one of many tools in their
underwriting situation.
The concern is notwithstanding that it is predictive, and
that it is inexpensive, if you strike the balances, what impact
does it have if there is a disparate impact on races, and how
much of that is tolerable?
Mr. Gutierrez. I just wanted to share with you, Mr.
McCarty, that when I read your testimony, I fully understood
the difference between your Association and their position and
your personal position here today. So I wouldn't make a big
deal out of it. We are elected officials and we are people who
represent different views.
Let me ask Mr. McCarty, do you have any information that
the FTC may have shared their report with the insurance
industry?
Mr. McCarty. That was our understanding in our Association,
that the report had been shared. I have no evidence to support
that, but that was--it was a common understanding. And the
reason it came to our attention because we would certainly
would have welcomed the opportunity--as the consumer protectors
for the State insurance regulators, would have welcomed the
opportunity to have reviewed the report in advance as well, to
provide some guidance. And hopefully we'll have that
opportunity to work collaboratively with the FTC in the future.
Mr. Gutierrez. All right. Let me just end by saying that I
thank you all for your testimony. I have been here for 16
years, and I assure you that the insurance industry and the
financial services industry has no lack of power, no lack of
influence on the members of this committee, and no difficulty
in getting their way.
That has been my experience during the last 16 years. So I
am sorry if I am not real sorry for the insurance industry or
for questioning their motivations or their tactics. Thank you
very much.
Chairman Watt. I thank the gentleman. The gentleman from
North Carolina, Mr. McHenry, is recognized for 5 minutes.
Mr. McHenry. I thank the chairman.
Now, Representative Keiser, the Federal Reserve shows
that--one of their studies shows that seniors tend to have
higher credit scores. I don't know if you have seen that fact.
But if this legislation were in place, could it cause higher
insurance rates for those with higher credit scores?
Mr. Keiser. Mr. Chairman, and Representative McHenry, I
think that is the important point to be made today, that if
this legislation were to pass, there would be losers and there
would be winners. Those people who currently are having the
advantage of having good credit are going to pay higher
premiums. Those who have, for whatever reason, not as good a
credit score, are going to pay less. There's no free lunch. The
insurance companies are going to make their money.
Now the question is: Do you reward good behavior in the
form of good credit? Good credit is a fine thing. And again,
NCOIL has been very deliberative on this, and we have attempted
to protect those unique situations that occur. Young people,
old people with no credit line; those people who have
extraordinary circumstances; those people--and I went through
it in my testimony to Mr. Chairman--but the point is there is a
way to address application of credit scores to make it as
reasonable as is possible and as fair as is possible without
throwing credit scores out, to the disadvantage of some groups
who have worked very hard to establish good credit.
Mr. McHenry. Commissioner McCarty, do you have a response
to that?
Mr. McCarty. I'm sorry. Would you repeat the question?
Mr. McHenry. Do you have a response to that?
Mr. McCarty. I don't recall your question, sir. I
apologize.
Mr. McHenry. If you were actually listening to
Representative Keiser, I'm asking if you have a response to
what he just said. I don't know if you were doing what many
behind you were doing, listening to something else. But--
Mr. McCarty. No. What our concern is with regard to credit
scoring is first of all with it in terms of its potential
impact and redistribution with regard to senior citizens; our
evidence and our research has found that many senior citizens
have thin credit files. My grandfather, for instance, didn't
have a credit card; he paid his rent in cash. His credit score
probably would not be good, although he was certainly
financially responsible.
But the Representative is absolutely right. If you
eliminate an underwriting tool for determination of premiums
paid, there are going to be some winners and some losers. And
what the balance is of that is if credit scoring is used and it
has a disparate impact on--racially discriminates against
protected classes of people, where do the public policymakers
strike a balance--
Mr. McHenry. So is your issue with the insurer's use of a
credit score? Or is it your belief that a credit score--or
maybe both--that a credit score has an innate racial component
to it?
Mr. McCarty. My concern is both. Historically, insurance
has been for two purposes: Number one, to provide for financial
security; and number two, loss prevention. And with regard to
loss prevention, I don't see how credit scoring really supports
that insurance principle, since what we want to do is to get
people to drive more responsibly. And I don't see how improving
your credit score serves the purpose of loss reduction.
Mr. McHenry. But do people not also have to pay for
insurance? Therefore, their record of paying or not paying in
other financial service products could be an indicator of
whether or not they will pay for a renewal of their insurance.
Mr. McCarty. Well, that's possible, but the insurance
premium is paid up-front.
Mr. McHenry. It is--
Mr. McCarty. Yes--
Mr. McHenry. But under State mandates, doesn't an insurer
have to cover them for 30 days? Isn't there a gap by which
insurers have to cover?
Mr. McCarty. Failure to pay your policy will result in
cancellation of your policy.
Mr. McHenry. But you have to give them 30 days to do that.
Mr. McCarty. You give them a notification, but they will
notify you if you don't make that payment. They are not behind
in terms of collecting the premium. They will cancel you and
earn the premium that you have paid up to that point.
Mr. McHenry. There should be an expense associated with
that as well, if you're slow to pay or you have to send out
multiple notices. So wouldn't an insurer, wouldn't they be wise
to know that, up-front?
Mr. McCarty. Yes, they would be.
Mr. McHenry. So wouldn't a credit score be useful, then?
Mr. McCarty. Would a credit score be useful? In my opinion,
I think that there are enough built-in costs and expenses, if
you premium finance, that the companies who use premium
financing are able to secure and to pay for those additional
charges.
Mr. McHenry. Interesting. Thank you, Mr. Chairman.
Chairman Watt. Thank you.
The gentlelady from California, Ms. Waters, is recognized
for 5 minutes.
Ms. Waters. Thank you very much.
Ms. Parnes, I want to know how credit scoring is balanced
against the experiences of the driver; for example, in
automobile insurance, I would think that the indicators of
whether or not you have a lot of tickets, you have had
accidents, etc., plays a role. What role does credit scoring
play in the decisionmaking?
Ms. Parnes. I don't think that our study told us exactly
what role it played.
Ms. Waters. Did you ask anybody? Without a study? You are
the Federal Trade Commission. Do you know whether or not they
make these decisions solely on credit scores, or is it a
combination of factors?
Ms. Parnes. It's based on a combination of factors.
Ms. Waters. How do you know?
Ms. Parnes. We know that from talking to people in--
Ms. Waters. What are the other factors?
Ms. Parnes. The other factors like--
Ms. Waters. Who knows this information? What are the other
factors? How do they do this?
Mr. McCarty. They do it on age, your driving history, and
claims made in the past. There is a myriad of factors that
could be used. Most States have passed some laws that say that
you cannot use credit scoring as the sole factor. But that can
be somewhat misleading.
It could be the predominant factor, and in some insurance
companies--not all--but some insurance companies heavily rely
on it because of its predictability.
Ms. Waters. The Honorable George Keiser, do you know the
weight that credit scores have on the decision of the cost of
insurance? How heavily does it weigh?
Mr. Keiser. Mr. Chairman and Representative Waters, our
insurance commissioner could answer that question, because
everything is filed in our insurance--
Ms. Waters. Are there any companies who use it solely? Or
is it 50 percent of the decision? Is it 75 percent of the
decision? How does it work?
Mr. Keiser. In our State, and I believe in all States that
have adopted the NCOIL model--and that would be 26 States at
least--it cannot be used solely. And we have by definition said
``solely'' would be 51 percent cannot be weighted. So it could
be a significant weighting factor, but I cannot answer the
specific combination of factors used or the weights applied.
Ms. Waters. All right. Let's see. Mr. McCarty, would you
agree that if you use credit scoring solely or it's heavily
weighted to make the decision that it reduces your costs of
investigations and of collecting and gathering information, so
that you can make determination about one's ability to pay?
Does it reduce the costs, the personnel costs, the
investigation costs, the vetting costs?
Mr. McCarty. According to industry spokesmen and industry
trade associations, it is substantially cheaper to use a credit
scoring mechanism than it is to do traditional underwriting.
Ms. Waters. Well, you know, I have looked at this, and I
have tried to figure out why there is an argument that somehow
credit scoring is a strong indicator. And it just doesn't make
good sense to me. I cannot make sense out of it. And I don't
even know why the FTC would spend the taxpayer's money, except
I guess you were asked to do it. It just doesn't make good
sense.
So I am trying to figure out why. And as I listen to all of
this, I recognize that the cost of reviewing an application and
determining what kind of experiences these drivers in
automobile insurance have had, whether or not--it costs a lot
of money to do that. And so to just go to the credit score
really reduces the cost of the insurance company. And I'm
beginning to believe that's really what this is all about.
Mr. Keiser, you mentioned that even GPAs are a good
indicator of something. Did you say that?
Mr. Keiser. Representative, absolutely. In our State, good
students get a discount.
Ms. Waters. So you are telling me that a smart student is a
better driver.
Mr. Keiser. Good students can get a discount.
Ms. Waters. Having nothing to do with their driving--
Mr. Keiser. And good students--
Ms. Waters. Just a moment--
Mr. Keiser. That is correct.
Ms. Waters. Had nothing to do with their driving record. A
low GPA indicates that you're not as good a driver. Is that
right?
Mr. Keiser. It is a predictor that is used in some cases.
Ms. Waters. Well, that is absolutely nonsensical. I know
some of the smartest people, I mean geniuses, who are just
stupid. I mean they make good grades, but they can't find their
way to the toilet. And if you're telling me that's an
indicator, I'm more convinced than ever that this is not good.
And so--my time is up. Enough said. I'm moving forward with my
legislation. This doesn't make good sense. Thank you.
Chairman Watt. I thank the gentlelady for her testimony. I
was going to withhold this until the second panel, but since
the gentlelady made inquiries, I wanted to ask unanimous
consent to submit information from the Web sites of three
insurance companies: AllState; Traveler's; and State Farm, on
what factors they consider in determining rates, and I would
invite the gentlelady to take a look at these. She will find
them very unenlightening in trying to figure out what factors
are used.
Mr. Miller. At the same time--I would like to submit--
Chairman Watt. I recognize Mr. Miller for--
Mr. Miller. --a letter from the American Insurance
Association, the Financial Services Roundtable, the U.S.
Chamber of Commerce, the Independent Insurance Agents and
Brokers of America, the National Association of Mutual
Insurance Companies, and the Independent Insurance Agents and
Brokers of America. I would also like to submit: a statement
from the National Association of Mutual Insurance Companies; a
statement from Michael J. Miller and EPIC Consulting; and a
statement from the Property Casualty Insurers Association of
America.
Chairman Watt. I have those, sir, and we will make sure
they get in the record, without objection.
Mr. Miller. Thank you.
Chairman Watt. Mr. Boren from Oklahoma is recognized.
Mr. Boren. Thank you, Mr.--
Chairman Watt. I'm sorry. We have--somebody else came in
that I didn't see. Mr. Barrett from South Carolina is
recognized for 5 minutes.
Mr. Barrett. Thank you, Mr. Chairman. I would like to yield
30 seconds to Mr. Miller from California.
Mr. Miller. Yes. I just wanted to follow up. Mr. McCarthy,
you made an interesting comment. You said that the industry
relies upon credit scores because of its predictability. It's
proven to be right. Nobody knows why, but it's proven to be a
predictable measure of determining risk. Is that not a fair
statement?
Mr. McCarty. Yes. There's a strong correlation between
credit scores and claims.
Mr. Miller. Yes.
Mr. McCarty. Predication of claims.
Mr. Miller. And that's--
Mr. McCarty. Not necessarily accidents, but claims.
Mr. Miller. Okay. But claims. I think that speaks volumes.
Chairman Watt. Would the gentleman yield just for a second?
Mr. Barrett. Absolutely, Mr. Chairman.
Chairman Watt. This is a good time to--Florida did a study
about this claims versus risk issue, in which you found that
doctors, accountants, and lawyers all have higher accident
rates, yet they get lower rates because of their occupation and
education. And your study found that 50 percent of eligible
claims are not reported for fear of a rise in insurance
premiums. Does that play into your response to Mr. Miller's
question?
Mr. McCarty. Yes, sir. And this refers back to a public
hearing we had with regard to occupation and education used as
criteria, where your doctors, lawyers, etc., who may have a
higher claims experience, get more favorable treatment with
regard to the cost of premiums. Those with lower education
levels and with other occupations like mechanics, etc., pay
substantially more, even though the evidence does not support a
higher loss ratio. There may be more frequent claims, but that
again can be explained by the fact that higher-income
individual policyholders have the wherewithal to pay them,
whereas lower-income folks will file a claim.
Mr. Miller. Well, Mr. Chairman, I'm convinced we need to
introduce a bill outlawing attorneys, without a doubt. We need
to stop those people.
Chairman Watt. You mean just attorneys? Not doctors, or
accountants, or--
Mr. Miller. I might need a doctor. I don't need an
attorney.
Chairman Watt. Oh, okay.
I thank the gentleman. I appreciate him, and I will ask
unanimous consent for 2 additional minutes for the gentleman,
so that he is not deprived of his time.
Mr. Barrett. Outstanding. Thank you, Mr. Chairman.
Gentleman and lady, thank you so much for being here today.
I am certainly a free market believer. I think the less
government interference, the better. And I am concerned when we
have government mandates, how that affects the market in the
redistribution. I would like Ms. Parnes and Mr. Keiser to
answer this: Do you think if we ban the use of credit scoring,
that we might have a socialization, meaning the lower-risked
folks subsidizing the cost of the folks who have a higher
credit rating? Either one of you, or both of you, please.
Mr. Keiser. I have already answered that and the answer is
``yes.''
Mr. Barrett. I apologize. Same thing.
Ms. Parnes. As I have indicated earlier, the Commission has
not taken a position on legislation, and really hasn't
considered what the impact of a ban would be.
Mr. Barrett. Okay. For all three of you, if you would
please, let's just say, for example, the credit-based insurance
scoring is banned or curtailed, and we went with something
different. First of all, what might that be? What is a better
way to do that? And would it be less accurate than what we are
using now or more accurate? Please answer that one, if you can.
Mr. McCarty. Prior to the use of credit scoring, insurance
companies had a long, mature history of looking at a variety of
factors, including geographic area, the driving experience,
number of years you have been insured, have you been
continuously insured, your driving history, driving record,
traffic violations, etc. Those have been historically used in
the determination and underwriting of auto policies. Removing
today immediately the use of credit scoring, since it's used
very heavily by the insurance industry, would be disruptive.
That could be ameliorated to some extent by phasing it out over
time.
Mr. Barrett. Mr. Keiser?
Mr. Keiser. I would simply respond that, again, any valid
measures that can be developed for predicting risk exposure
claims should be developed for the benefit of consumers in our
country.
On the other hand, again, we have to make sure that the
abuses that can accompany any of these measures, whether it's
credit scoring or any other measure, not be allowed.
Ms. Parnes. The Commission looked at other models, other
possible models, and was not able to come up with one in its
study.
Mr. Barrett. Okay. I can think of several different
examples where government mandates rather than the free-market
process have led to the consequences that we're not looking
for. In fact, the one that comes closest to mind is ethanol. We
mandate certain things on ethanol, we monkey with the free
market process, and all of a sudden food prices go up.
Can you--in the financial realm, can you give me some
examples of where government intervention in free market
process has led to consequences where the government actually
said that they were going to try to fix something, and the
opposite consequence happened? Can you give me some examples,
any of you?
Mr. Keiser. Well, I think the most obvious one as a State
legislator that comes to my mind, was welfare reform when it
was federalized. The Federal Government was very quick to send
it back to the States once it was entirely mismanaged at the
Federal level. And the States have done a fairly good job with
that. But in the case of the subject at hand, again, to tie the
hands of the industry in terms of valid predictors will create
some offset--and there is no alternative.
The insurance companies, I don't believe, are going to lose
money. Somehow the cost will be shifted if this legislation
were to pass.
Mr. Barrett. I see my time is up. Thank you, Mr. Chairman.
Chairman Watt. I thank the gentleman. Mr. Boren from
Oklahoma is recognized for 5 minutes.
Mr. Boren. Thank you, Mr. Chairman. I want to say thank you
for allowing me to be on this subcommittee, and also, the
Housing Subcommittee, with Ms. Waters. I am honored to join
both of you, and I know we have worked on a lot of contentious
issues over the past couple of years, and hopefully we can work
together in the next few years.
I have a couple of questions, very brief. First for
Director Parnes, you kind of answered this earlier, but again I
would like the answer. Did not the industry provide the data
for auto and homeowners voluntarily? Did they come to the
Commission and say, ``We're going to bring this information
voluntarily?''
Ms. Parnes. The industry did provide the information for
the auto insurance study voluntarily. We went to the industry,
we were directed to do this study, and we talked to the
industry about voluntary submissions. And we were able to reach
agreement on that.
Mr. Boren. And did you respond formally to the industry?
And if you didn't, why did you not?
Ms. Parnes. Respond formally?
Mr. Boren. Like in a letter, or some kind of formal
response. If they said, ``We're going to provide for the
homeowners' study,'' for instance.
Ms. Parnes. Well, for the homeowners' study, we began the
process of discussing voluntary submissions. But it was shortly
after we began those discussions, the Commission made the
decision that it would proceed through a subpoena process.
Mr. Boren. Do you think that is premature, to do that, when
you have an industry that is basically begging you, saying,
``Here is the information, so we're going to take the added
step to beat someone over the head when they have actually come
to you.'' That is really not the role of government to hurt
someone when they are actually trying to help you and provide
information. Is that correct?
Ms. Parnes. Well, it certainly isn't the role of
government, and it's not the approach that the Commission has
taken. We do use these subpoenas typically when we're doing
industry-wide studies. And they're not intended to suggest that
the industry that we're studying is not being cooperative.
I think the Commission's concern in the homeowners' study,
as indicated earlier, is that the value of Commission studies
is really based on both their actual reliability and also the
perception of their reliability. And there was a lot of concern
expressed about the auto insurance study as not being reliable
because the data was submitted voluntarily.
The Commission supports the study, the Commission stands
behind it, believes that the study is reliable; but in response
to those concerns, decided to pursue information under subpoena
for purposes of the homeowners' insurance study.
Mr. Boren. Okay. One final one for you: Since there are
questions as to the Commission's legal authority under either
Section 6 of the FTC Act or Section 215 of the FACT Act to
compel information from insurers generally or with regard to
this specific study, is it possible that actually using these
subpoenas would delay the study, or has delayed the study?
Ms. Parnes. Well, we hope that it doesn't delay the study.
Certainly if the subpoenas are challenged, there could be some
delay, but we will--
Mr. Boren. Do you know how long a delay that would be? Or,
is there is a precedent?
Ms. Parnes. I don't, and I don't know if they will be
challenged, either.
Mr. Boren. Okay. That would be interesting to know if they
would be challenged.
I have a question for Commission McCarty. This is
actually--I have an e-mail here from my State insurance
commissioner, Kim Holland, who is a great friend of mine. And
this is what her e-mail basically says: ``Oklahoma's experience
suggests that the vast majority of our policyholders are not
impacted, or actually save money due to credit scoring. We also
prohibit most of the activities found objectionable by other
regulators, such as--rates to be affected by race, gender, or
no credit history.'' Do you have a response to that? And you
know, compare Oklahoma to maybe Florida. What are your
thoughts?
Mr. McCarty. Yes. I actually discussed this issue on the
telephone with Commissioner Holland the other day, and she
feels very strongly as long as there is a predictive value,
that there is an argument to be made that you're actually
providing better value for a majority of consumers, that it is
color-blind with regard to--or from the initial question, since
the asking about the running of a credit report does not ask
the question up-front, so any consequential racial
discrimination is not intentional. And that does represent a
view within our organization. Other commissioners would share
that view. I do not.
Mr. Boren. Well, I would love to ask more questions, but it
looks like my time is up. Thank you all so much for your
testimony.
Chairman Watt. I thank the gentleman for his questions. Can
I just have unanimous consent to ask Mr. Keiser one clarifying
question? You said that the result of not using insurance
scoring for this purpose would result in winners and losers;
some people would be adjusted up, and some people would be
adjusted down. But if that is a greater reflection of actual
risk, are you suggesting that is a bad thing?
Mr. Keiser. Mr. Chairman, I think if it is a valid
predictor, and your committee, through the investigation, can
determine that, if it is a valid predictor and it can be
accomplished inexpensively, it diminishes, number one, the cost
to the insured; and number two, if it's a valid predictor, it
gives those people for whom it validly predicts good credit a
lower rate, those people with poor credit a higher rate.
Chairman Watt. I appreciate that answer, but it is not
responsive to the question I asked, unfortunately.
Mr. Keiser. I apologize. I don't understand--
Chairman Watt. And I understood that you had testified to
exactly what you just said. The question I'm asking is, if
there are winners and losers as a result of using a different
kind of mechanism other than credit-based insurance scoring,
and you get a more accurate reflection of what the risks
actually are, are you suggesting that is a bad thing?
Mr. Keiser. Mr. Chairman, no. I don't believe that
measure--
Chairman Watt. Okay. All right. I didn't think you were. I
just wanted to make sure that we got that on the record. We
obviously know that there are winners and losers. And, as you
say, insurance companies will try to find a way to make a
profit; that is what they are in business for. But there is an
also important factor here in trying to come up with a fair
system that does not shift the burden as a result of unfairness
in credit scoring to poorer people and minorities. And I wasn't
trying to trick you on that; I was just trying to make sure
that I understood what you were saying in your testimony.
Mr. Miller. May I have one minute?
Chairman Watt. Yes.
Mr. Miller. There are a couple of things I didn't bring up.
The State of Oregon decided they wanted to eliminate credit
scores, and they went to the voters, and it was overturned 2 to
1. The people said, ``No,'' they think that's a reasonable way
to predict rates. New Jersey actually reversed itself when they
went the other direction, and came back the other way.
I guess the difference between what a lot of us are
hearing, but even Mr. McCarty, you said that the industry
relies upon credit scores because of its predictability as it
applies to claims, and that is what we are looking at. The
testimony I have heard today said that if we eliminated that,
seniors are likely to be impacted unfairly by changing the
requirements. So until I can find a better way, or somebody
presents a better way of predictability, it seems like the
system today is working, and it is predictable. And if I
thought it was discriminatory, I would absolutely oppose it.
Thank you. I yield back.
Chairman Watt. I thank the gentleman, and I wasn't going to
offer this until you raised the question. But--
Mr. Miller. Oh well, equal time.
Chairman Watt. No. I'm just going to make a unanimous
consent request to offer into the record, since you raised the
question of the Oregon vote, the information about who paid for
lobbying and the amounts that the insurance industry paid for
lobbying in Oregon--just for the purpose of completeness of the
record.
Mr. Miller. Well, those people voted for Obama, too, didn't
they? Something is wrong with that State. I can tell. I yield
back.
Chairman Watt. Thank you. I ask unanimous consent to put an
article from USA Today, ``Credit scores' link to insurance
rates tested,'' by Christine Dugas, into the record. The
article discusses what insurers spent opposing the use of
credit scoring in insurance. Without objection, it is so
ordered.
We thank these witnesses for testifying. I think you have
added immensely to our knowledge base here, and we will excuse
you, and call up the second panel of witnesses.
If I could ask the witnesses on the second panel to be
seated? We seem to be missing one. In the interest of time, I
am going to proceed with the introduction of the witnesses. I
think Mr. Poe is here; he must have stepped out for a moment.
He has returned.
We are delighted to welcome our second panel of witnesses.
They will testify in the following order: Mr. Bob Hunter,
director of insurance of the Consumer Federation of America;
Ms. Lisa Rice, vice president of the National Fair Housing
Alliance; Mr. Eric Poe, chief operating officer of CURE
Insurance, a New Jersey-based insurer; Mr. Charles Neeson,
senior executive, personal lines products, Westfield Group, who
is testifying on behalf of the Property Casualty Insurers'
Association of America; Mr. Stuart Pratt, president of Consumer
Data Industry Association; and Dr. Lawrence S. Powell,
professor, University of Arkansas at Little Rock.
I think all of you were present earlier when I laid out the
rules of the road. Each of you will have your entire statement
submitted in its entirety for the record, and we would ask you
to summarize your testimony in 5 minutes or less. You will get
a green light at 4 minutes, a yellow light for the 4th minute,
and then a red light at the end of 5 minutes, and we would ask
you to wrap up when you see the red light as expeditiously as
possible.
With that, Mr. Bob Hunter, director of insurance, Consumer
Federation of America, you are recognized for 5 minutes.
STATEMENT OF J. ROBERT HUNTER, DIRECTOR OF INSURANCE, CONSUMER
FEDERATION OF AMERICA
Mr. Hunter. Thank you, Mr. Chairman, and members of the
subcommittee. My name is Bob Hunter, and I served as Federal
Insurance Administrator under Presidents Ford and Carter. I
also served as Texas Insurance Commissioner.
Insurance scoring is used to determine whether a customer
will be eligible for coverage, and the premiums that the
customer will pay. In response to a question earlier, I would
say that for many companies, credit scoring can have a greater
impact than claims and other key factors on a final rate of an
individual.
CFA and many organizations, civil rights and others, have
called for a prohibition on insurance scoring because its use
in insurance undermines core functions of the insurance system,
decreasing insurance availability and affordability,
undermining the critical role of insurance and encouraging loss
prevention. It has an adverse disparate impact and
discriminates against low-income and minority consumers. It's
based on reports that often have errors or are incomplete. It's
inherently unfair and penalizes consumers who are the victims
of economic or medical or natural catastrophes. It even
penalizes them for improper lending business decisions that
we've noticed over the last few years. And it violates sound
actuarial principles.
The insurance industry claims a variety of benefits; but
when you boil it all down, it basically says, ``We have a
correlation,'' and therefore we're more predictive with it of
the likelihood of a consumer having a claim. If that were true,
we would expect an increase in delinquencies in bankruptcies
would be matched by an increase in insurance claims. Since
2000, bankruptcies and delinquencies have risen sharply, but
auto claim frequencies have declined sharply. This suggests
there is no correlation.
Why aren't we seeing the correlation at work over time?
Insurers can't tell us what it is about a credit score that is
linked with risk. This is what Ms. Waters and others have
raised. What is it linked to? If you ask them why a person who
suffered a decline in credit because of Hurricane Katrina or
losing a job because of outsourcing is a worse risk, they can't
answer.
Unlike every insurance class before credit scoring was
adopted, credit scoring is not based on a logical rationale
confirmed later by a statistical analysis. A correlation alone
with no thesis being measured means the credit score violates
actuarial principles.
The only thesis insurers have manufactured but cannot prove
with data is that people with bad credit are irresponsible. But
try telling that to people laid off from a job, or after a
major medical problem, or after suffering financial difficulty
from a divorce. These 3 life events account for 87 percent of
family bankruptcies. That is not irresponsibility; that is life
events.
If insurers call this irresponsible, they're even more
heartless than I thought. In fact, there's strong evidence that
insurance scoring is not a predictor of insurance claims, but
rather a proxy for other factors that are related to claims
experience, such as income or geographical location of the car.
But it also is a proxy for race as well as income.
Two independent studies by Texas and Missouri found strong
relationships between race and income. The Missouri department
said race was the strongest, when you look at the correlation
between race and credit scoring. Even the recent flawed FTC
report, as we've already heard, found this correlation.
Hopefully, the home insurance study will be better because of
what we've heard today that they'll actually collect the data
that they didn't have, which was hugely inadequate, the auto
insurance study, for reasons I've listed in my testimony.
Insurers claim that competition would be harmed and
availability reduced if credit scoring is banned. This is
false. I need only to point to California, where credit scoring
is banned for use in auto insurance. In our recent in-depth
study of auto insurance regulation, we found that State has the
best system in the Nation, including the ban on credit scoring.
While the insurers claim competition would be harmed by it, our
data shows that California has the 4th most competitive market
in the Nation, measured by HHI. Plus the assigned risk rate has
dropped to only \1/10\ of 1 percent of the autos insured in the
State.
This proves availability and competition are not harmed by
banning the use of credit scoring. Indeed, Massachusetts, which
was praised earlier, as becoming less State-controlled, still
has prior approval and bans credit scoring. So does Maryland
for home credit scoring.
We applaud the sponsors of H.R. 5633 and 6062. We have
concerns about the first one, H.R. 5633. The legislation's goal
of banning credit scoring if the use of consumer credit
information discriminates on the basis of race is a good one,
but we fear that the legislation will not achieve the results
for a series of reasons I listed in my testimony.
CFA believes that simple banning of credit scoring in
insurance is necessary, and everything we've studied indicates
it's the right thing to do; therefore, we enthusiastically
support Ms. Waters' bill, H.R. 6062.
[The prepared statement of Mr. Hunter can be found on page
59 of the appendix.]
Chairman Watt. I thank the gentleman for his testimony. Ms.
Lisa Rice, vice president, National Fair Housing Alliance, is
recognized for 5 minutes.
STATEMENT OF LISA RICE, VICE PRESIDENT, NATIONAL FAIR HOUSING
ALLIANCE
Ms. Rice. Thank you, Chairman Watt, Ranking Member Miller,
and members of the subcommittee for the opportunity to testify
today on credit-based insurance scoring.
The National Fair Housing Alliance (NFHA) is a consortium
of more than 220 private nonprofit fair housing organizations
and State and local civil rights agencies. Our mission is to
eliminate all forms of housing discrimination and to expand
equal housing opportunities.
It is NFHA's position that Congress should ban the use of
credit scoring in insurance. Studies by the Missouri and Texas
Departments of Insurance have found that insurance scoring
discriminates against minority consumers because of the racial
and economic disparities inherent in the scoring systems.
Even though the Federal Trade Commission used data hand-
picked by the industry for its 2007 study, it found that credit
scoring discriminates against low-income and minority consumers
and that credit-based insurance scores ``appear to have some
proxy effect for three of the four coverages studied.''
Unfortunately, instead of highlighting this discriminatory
connection, the FTC chose to restate the arguments of the
insurance industry that scores are related to responsibility
and risk management.
The industry claims that there are ``intrinsic underlying
individual biological and psychological characteristics of risk
taking in both financial behavior and driving.'' This argument,
however, ignores the fact that racism and discrimination have
always been present in our society, and that discrimination is
inextricably tied to inequality in our lending and financial
markets.
People of color do not have a risk-taking biology. African
Americans and Latinos have lower insurance scores because of
direct and indirect discrimination in the marketplace. America
has a bifurcated lending system that disproportionately
discriminates against borrowers of color.
Countless studies and court cases have demonstrated this.
My own organization conducted a multi-year lending testing
project which uncovered multiple ways in which African
Americans were denied lending opportunities, including
receiving inferior basis information that their white
counterparts were given, being urged, unlike their white
counterparts, to go to a different lender, and being told,
unlike their white counterparts, that they would not qualify
for a loan. This happened even though both the African American
and white consumers were equally qualified.
Parenthetically, NFHA has been involved in conducting
hundreds, probably thousands of tests of insurance companies,
and we have found similar biases there too, based on race.
Our bifurcated lending system has also helped to lead to
the current foreclosure crisis. As we all know, African
Americans and Latinos were disproportionately targeted for
subprime loans and unsustainable mortgages, even when they
qualified for better rates. Thus, borrowers who entered the
mortgage cycle with sound credit are now facing plummeting
credit scores.
It is wholly unfair to further burden borrowers who were
unfairly targeted by unscrupulous lenders with higher insurance
premiums. These borrowers will not suddenly turn into poor
drivers or lax homeowners simply because their credit scores
have decreased. Banning credit-based insurance scores is a
civil rights issue, which is why NFHA supports H.R. 6062.
We also appreciate the efforts regarding H.R. 5633, but are
concerned that the bill lacks an objective standard for
identifying racial discrimination, gives broad discretion to
the FTC, and has no private right of action. Most importantly,
H.R. 5633 could serve to legitimize insurers' use of credit-
based insurance scoring in general.
A recent study demonstrates that if you crash your car, you
can blame the stars. The study found, looking at records of
100,000 drivers that there is a statistically significant
correlation between Zodiac signs and car accidents. Based on
the study's findings, Libras, Aquarians, and Aries are the
worst drivers. Who knew?
The National Fair Housing Administration was involved in
litigation against a major insurance company that utilized a
credit scoring model. An analysis of the model, which we could
do because of discovery under a protective order, found clear
disproportionate impact on African Americans and the price they
paid for insurance, which could not be accounted for by
differences in the risk they posed. In other words, African
Americans paid a higher rate than was commensurate for their
level of risk.
We urge you to ban the use of consumer credit information
for insurance, and thank you again for the opportunity and the
invitation to speak to you today.
[The prepared statement of Ms. Rice can be found on page
233 of the appendix.]
Chairman Watt. Thank you for your testimony.
Mr. Poe, chief operating officer of CURE Insurance.
STATEMENT OF ERIC POE, CHIEF OPERATING OFFICER, CURE AUTOMOBILE
INSURANCE
Mr. Poe. Thank you very much, Mr. Chairman, and members of
the subcommittee for inviting me today to talk about this
proposed bill. As you said earlier, I am the chief operating
officer of CURE Auto Insurance. We are a regional insurance
carrier based out of Princeton, New Jersey. We are licensed to
write in the State of Pennsylvania, as well as the State of New
Jersey.
I would like to start out by giving some background on our
interest in this particular issue. Prior to 2003, in the State
of New Jersey, insurance carriers were not approved for the use
of credit scores, education, occupation, as well as
homeownership status as factors in underwriting.
However, in 2003, the New Jersey government decided that
they wanted to attract more market players, new national
carriers into the marketplace. And it was at that time they
started permitting credit scores, education, occupation, and
homeownership. As an organization that writes private passenger
automobile insurance, it was at that time that we had to study
and analyze these underwriting methods to determine their
validity.
After significant review, CURE Auto Insurance determined
that while these rating and underwriting variables do correlate
to loss ratios, they merely serve as statistical proxies for
income, which is why CURE Auto Insurance does not employ any of
those factors. However, we believe that we will soon be
compelled to for competitive measures if this is not stopped.
To start, I would like to explain some of the conclusions
that we found when reading the reports that were issued dealing
with credit scores. It appears evident that the auto insurance
industry uses loss ratio models as justification for using
credit scores, education, as well as occupation. Now showing
statistical correlation to these characteristic traits to loss
ratios, the entire industry has been able to validate using
credit scores and all these other factors.
However, I think it is important for everybody to
understand what is a loss ratio. By definition, a loss ratio is
the incurred losses and loss adjustment expenses divided by net
earned premium. In layperson's terms, it's a measure of
profitability which we call rate adequacy.
Surprisingly, what our examination yielded was that the
studies dealing with credit scores, education as well as
occupation, have made an inappropriate conclusion--that simply
because you show a correlation to loss ratios, it means that
the variable that you are testing automatically is a predictor
of risk.
However, it's important to understand that there's an
infinite number of characteristic traits that you can draw
correlations to loss ratios, but they would all be invalid if
you can explain a more valid characteristic trait that's
imbedded in that variable. The best example that was given
before was life insurance and mortality tables with African
Americans in the past, with African Americans having shorter
life spans. Life insurance companies still aren't able to use
it.
My example would be that African Americans have shorter
mortality tables not because they're black; it's because
socioeconomics are involved with them having lower mortality
tables. Because they are in lower-income neighborhoods; it is
more likely there are homicides; it is more likely there is
crime; it is more likely that because they are less educated on
average, they are going to eat worse foods, they are going to
have diabetes, and they are going to have high blood pressure.
Those are the reasons that are the driving factor of why
mortality tables for life insurance may be shorter for African
Americans than whites; it's not because of the color of their
skin.
Similar to loss ratios that we have here with credit
scores. In an example that I just gave, I would say that with
credit scores, what we found is that they are very good valid
predictors of income, but not necessarily good valid predictors
of risk in terms of anything else.
Now our fellow members of our industry would like to
disguise the public policymakers as regulators, as well that
these rating variables of credit scores possess some
unexplainable commonality of why they correlate to risk, and
therefore are valid predictors of risk. But this is in light of
the fact that all of these variables that we're talking about
here have an obvious correlation to income, and it is income
that is correlated to loss ratio or profitability for our
industry.
Now speaking about credit scores specifically: To clarify
first credit-based insurance scores, when we examined them, the
differences between an insurance score and a credit score is at
least we found to be negligible. Asking how many oil accounts
and gas cards somebody has really didn't make a big difference
with what you would yield in terms of a credit score versus an
insurance score.
But the reason why we concluded that credit scores are
correlated to income is because if you look at the FICO credit
scoring model and you look at the models in which the companies
that offer credit scores to us, you would notice that although
35 percent of your credit score is determined based upon your
prior history of making on-time payments, number two in that
list is 30 percent going to credit utilization.
Now due to the fact that credit lines by lenders are
directly calculated, based upon a borrower's income, this is
why we believe the credit scoring model more significantly
based our belief on the conclusion that this was a strong
predictor of somebody's income.
The best example that I can give is really, looking at
somebody who does not make a high income, has a $1,000 credit
line granted; somebody who makes a lot of money has a $20,000
credit line granted, if they charge $800 in groceries, there's
an 80 percent utilization factor for those people who have low
incomes because they are using 80 percent of their credit line.
But I think the reason why the industry really wants to
adopt this practice of credit scores is for three reasons.
Number one, the industry itself wants to attract high-income
drivers for three reasons: Number one, we produce higher
revenue streams for people; number two, the insurance industry
data-mines for higher-income individuals that yield a lot of
money; and number three, richer people can absorb more of their
claims, which was spoken to earlier.
Finally, in summary, I just think that the subcommittee
needs to be more aware of the fact that the bigger problem in
our industry is the use of education and occupation as
underwriting variables in our industry. What companies are
doing, specifically GEICO, is they are basing whether or not
you have a 4-year college degree and whether or not you work in
a traditionally high-income earning job, as the basis of
putting you in the most expensive insurance company that the
affiliate has, and they don't tell them.
So, thank you very much.
[The prepared statement of Mr. Poe can be found on page 181
of the appendix.]
Chairman Watt. Thank you for your testimony.
Mr. Neeson, senior executive, personal lines products,
Westfield Group, you are recognized for 5 minutes.
STATEMENT OF CHARLES NEESON, SENIOR EXECUTIVE, PERSONAL LINES
PRODUCTS, WESTFIELD GROUP, ON BEHALF OF PROPERTY CASUALTY
INSURERS ASSOCIATION OF AMERICA
Mr. Neeson. Chairman Watt and members of the subcommittee,
thank you for the opportunity to comment on H.R. 5633 and H.R.
6062, legislation that seeks to prohibit the use of information
on credit reports for issuing and setting premiums for motor
vehicle and property insurance. My name is Charles Neeson, and
I appear before you today as a senior executive with Westfield
Insurance, and as a representative for the Property Casualty
Insurers' Association of America, a national trade insurance
association, of which Westfield is a member.
I am also a member of the American Academy of Actuaries and
an associate of the Casualty Actuarial Society. An insurance
company's ability to more accurately predict losses is a
critical component of property underwriting risks. Our industry
is united in our concern over the negative impacts that
restricting the use of credit-based insurance scores will have
on American consumers.
When insurers are able to properly underwrite risks,
consumers benefit with lower rates and more choices. Because
credit-based insurance scoring is an objective and accurate
method for assessing the likelihood of insurance loss, we
strongly oppose the passage of H.R. 5633 and H.R. 6062.
Insurance is an incredibly competitive business, and one
way for an insurance company such as Westfield to distinguish
itself from its competitors is to develop better ways of
gauging risk to more accurately price an insurance policy.
Westfield Insurance began using insurance scores in 2000 to
improve the pricing of our automobile and homeowners' insurance
products.
In analyzing the relationship between credit information
and our loss data, we found a strong correlation. Used in
conjunction with more traditional rating factors such as
vehicle age, performance, gender, territory, and driver age,
credit-based insurance scoring allowed Westfield to more
accurately price our products and improve our competitive
position.
Mr. Chairman, today approximately 75 percent of our auto
and home package customers pay less because of insurance
scores, while only 8 percent pay more. Outside of Westfield's
own experience with credit-based insurance scoring, an annual
survey published by the Arkansas insurance department shows
that insurance scores either benefit or have no effect on the
vast majority of consumers in Arkansas.
The latest survey shows that 90.2 percent of automobile
insurance policyholders, and 90.8 percent of homeowners'
insurance policyholders either received a discount or were
otherwise unaffected by the use of credit-based insurance
scores.
In July 2007, the Federal Trade Commission issued a report
to Congress on insurers' use of credit-based insurance scores.
In that report, the FTC concluded that insurance companies
which use credit-based insurance scores are more likely to
price automobile insurance more closely to the risk of loss
that the consumer poses.
This results, on average, in higher-risk customers paying
higher premiums and lower-risk customers paying lower premiums.
The use of credit-based insurance scoring is subject to
extensive regulation by the States. The National Conference of
Insurance Legislators (NCOIL) promulgated model legislation
regarding its use. And most States have either enacted that
model or have adopted restrictions similar to those contained
in the model.
Insurers that consider credit information in their
underwriting and pricing do so for only one reason: Insurance
scoring allows them to rate and price business with a greater
degree of accuracy and certainty. Sound underwriting and
rating, in turn, allows insurers to write more business, which
is a direct benefit to consumers. Without the ability to
consider credit, many insurers would be less aggressive in
their marketing, and far more cautious in accepting new
business.
Every serious and reputable actuarial study on the issue
has reached the same conclusion: There is a very high
correlation between insurance scores and the likelihood of
filing insurance claims. And while it is a common criticism of
insurance scoring that the exact reason for that correlation is
unknown, there are also numerous other rating factors, of which
causality is also unknown.
For example, even though there is no definitive explanation
as to why married individuals represent less risk than single
individuals, marital status is a widely accepted and widely
utilized rating variable. Credit-based insurance scoring is an
effective tool for insurers, and a fair one to consumers. To
protect competition and consumer choice, it is imperative that
insurers be permitted to fully price risks, using non-
discriminatory and statistically valid tools, such as credit-
based insurance scores.
Thank you very much for allowing me to come and testify
before you today, and I would be happy to address any questions
that you may have on this subject.
[The prepared statement of Mr. Neeson can be found on page
168 of the appendix.]
Chairman Watt. Thank you, Mr. Neeson, for your testimony.
Mr. Pratt, president, Consumer Data Industry Association,
you are recognized for 5 minutes.
STATEMENT OF STUART PRATT, PRESIDENT, CONSUMER DATA INDUSTRY
ASSOCIATION
Mr. Pratt. Chairman Watt, Ranking Member Miller, and
members of the subcommittee, thank you for the opportunity to
appear before you today. We commend you for holding this
hearing, and my comments will focus on just a few key points
drawn from our written testimony.
First, the States have fulfilled their mandate to protect
consumers through careful deliberations and extensive oversight
of the use of credit histories and scores for insurance
underwriting.
Second, our members management of the quality of data in
their databases is a proven success story.
And third, the market is addressing the question of
consumers with a thin credit report or no credit report at all.
In 1945, Congress enacted the McCarran-Ferguson Act, and in
doing so left the regulation of the business of insurance to
the States. And perhaps the question before us today is how
have they done with regard to the use of credit histories and
credit history-based insurance scores as a factor in
underwriting for personal lines of insurance? I think the
answer is clear and positive for all of us as consumers.
Virtually all States permit and regulate the use of credit
histories and scores. These decisions have been made with an
eye towards fairness. Studies by regulators have found that the
use of credit histories is fair and predictive.
In 2003, in testimony offered on behalf of the NAIC before
the full House Financial Services Committee, with regard to a
report from the American Academy of Actuaries, they stated the
following: ``The Academy members have reviewed studies and
believe that credit histories can be used effectively to
differentiate between groups of policyholders. Therefore, they
believe credit scoring is an effective tool in underwriting and
rating personal lines of insurance.''
There is no dearth of quality oversight regarding the use
of credit histories and scores. However, some suggest that
credit reports are not accurate and thus shouldn't be used for
underwriting. We could not disagree more strongly. Never before
have we had so much definitive data with regard to the accuracy
of credit reports.
In 2004, the Federal Reserve studied 300,000 credit reports
and they found the following to be true: ``Available evidence
indicates that the information that credit reporting agencies
maintain on credit-related experiences of consumers and credit
history scoring models derived from these experiences have
substantially improved the overall quality of credit decisions
while reducing costs of such decision-making.''
Consumer experiences in reviewing their credit report
disclosures validate the conclusions of the Federal Reserve.
Out of 52 million free credit report disclosures provided, only
1.98 percent of these reviews resulted in a dispute where data
was deleted. Often-cited studies with regard to accuracy have
been rejected by the Government Accountability Office, and in
their 2003 testimony, they state, ``We cannot determine the
frequency of errors in credit reports based on the Consumer
Federation of America, U.S. PIRG, and Consumer's Union
studies.'' Two of the studies did not use a statistically
representative methodology because they examined only the
credit files of their employees, who verified the accuracy of
the information, and it was not clear if the sampling
methodology in the third study was statistically projectable.
Our members data management is a success story, and we can
all have full confidence in the data upon which decisions are
based.
Some suggest that credit history should not be used because
some consumers do not have a credit report which can be scored
or don't have a credit report at all. State laws address this
by prohibiting insurers from denying, canceling, or non-
renewing a policy based solely on credit information. And we
agree with this position.
The good news, and the real good news for consumers is that
CDIA's members are leading the effort to expand the types of
payment data, which can be used for underwriting. Several of
our members have already brought to market public record data
products, which allow a user to consider assets where there's
an absence of credit payment history.
Some CDIA members already include utility and
telecommunications payment data in traditional credit-reporting
databases. Other members of the CDIA are aggregating checking
account consumer payment data, where such data is reported
directly by the consumer's bank to the database, and some CDIA
members provide services where they validate payment data
provided by the consumer.
The Political and Economic Research Council's empirical
study of 8 million credit reports found the following to be
true, including alternative data such as those that I've
discussed, are especially beneficial for members of ethnic
communities and other borrower groups. Hispanics saw a 22
percent increase in acceptance rates. The rate of increase was
21 percent for blacks, 14 percent for those age 25 or younger,
and 21 percent for those who earned $20,000 or less annually.
In conclusion, we believe the right balance has been struck
with regard to the Federal and State laws and that no new law
is necessary. The States have fulfilled the role expected of
them.
Our members' data contributes to fair treatment. A May 18,
2008, Washington Post story reported that a study of an entire
year's FHA applications turned up the additional fact that FHA
lower-income borrowers typically had higher scores than those
with larger incomes. This is powerful new data that should give
us confidence in the core value of credit histories. Our
members' data is blind to race and ethnicity. Our members' data
helps consumers. Consumers want to be recognized for their
years of care and responsible actions, regardless of their race
or ethnicity.
I thank you, Mr. Chairman, for your time, and I look
forward to answering your questions.
[The prepared statement of Mr. Pratt can be found on page
214 of the appendix.]
Chairman Watt. Thank you, Mr. Pratt, for your testimony.
Dr. Powell, professor, University of Arkansas at Little
Rock, you are recognized for 5 minutes.
STATEMENT OF LAWRENCE S. POWELL, PH.D., PROFESSOR, UNIVERSITY
OF ARKANSAS AT LITTLE ROCK
Mr. Powell. Thank you, Chairman Watt, Ranking Member
Miller, and members of the subcommittee. I'm honored to be
invited to share information with you about insurance scoring.
I appear on behalf of the Independent Institute as a research
fellow. I have a Ph.D. in insurance, and I hold the Whitbeck-
Beyer Chair of Insurance and Financial Services at the
University of Arkansas at Little Rock.
This is an important topic, given the financial stability
the insurance industry provides consumers. And accurate pricing
is a cornerstone of the insurance mechanism.
My perspective is that of an educator and a researcher. And
I think it's instructive to begin with a big-picture view of
insurance pricing. Insurance companies face an unusual
challenge; they must set prices for the products they sell
before they know all of the costs. To meet this challenge, they
employ complex pricing methods developed by actuaries, using
economic and statistical techniques.
It should then come as no surprise that some aspects of
actuarial science and insurance pricing are puzzling to people
who have not developed substantial expertise in this field.
Insurance scoring is an example of a beneficial innovation in
insurance pricing that causes some people concern.
It's my opinion that thorough consideration of insurance
scoring should lead one to conclude it is not only appropriate
for insurers, but that using it creates value and promotes
fairness in society. There are many compelling arguments in
favor of these conclusions.
Given the time limit in this forum, I would like to share
with you a fundamental reason why insurance scoring is a good
practice, and a fundamental observation suggesting that any
potential misuse of insurance scoring cannot persist in the
market.
The first fundamental point is that insurance scoring is an
extremely powerful and accurate predictor of insured losses.
Evidence of this is conclusive. Studies by the Texas Department
of Insurance, the Federal Trade Commission, and several others
showed that the subset of drivers with low insurance credit
scores submit more claims and cause more total loss payments
than those with high credit scores.
In fact, it has been shown that drivers with two or more
prior losses, but good credit, are less likely to have a loss
in the current year than drivers with clean driving records and
bad credit. There are many benefits to using accurate
predictors of loss in insurance pricing models.
For use of innovative, accurate predictors of loss, such as
insurance scores, availability of insurance has improved,
competition in insurance markets has increased, and costs have
decreased for many insurance consumers.
Many experts believe the coinciding advent of insurance
scoring and the decrease in residual market populations for
automobile insurance are directly linked. By introducing new
information to the insurance pricing models, insurers were able
to find acceptable risk they were previously unable to
identify.
Accurate loss models also benefit society by producing fair
outcomes in which insurance premiums are commensurate with risk
of loss. When insurers cannot use accurate predictors of loss,
low-risk drivers must pay higher premiums to subsidize high-
risk drivers. In addition to a general sense of fairness,
accurate loss predictors also create incentives for high-risk
drivers to take more care in driving.
Effective competition is a fundamental characteristic
observed in U.S. insurance markets. Competition prevents
insurers from charging excessive or unfair premiums. In 2005,
the NAIC data show an average of 157 insurance companies
underwriting the private passenger automobile cover in each
State. It's therefore reasonable to believe that an insurer
cannot systemically overcharge a group of drivers, because then
one of the other 156 existing companies, or perhaps a new
company, has an opportunity to cover that group of drivers at
an equilibrium price.
But we're not here because everyone likes insurance
scoring. I've heard critics describe potential or anecdotal
unfair outcomes associated with insurance scoring. And I do not
dispute the fact that some consumers have encountered
individual rating scenarios that seem to lack intuition.
For example, I know of a consumer in Arkansas who received
an increase in his premium because his wife canceled a credit
card they were not using. However, he called a few competing
insurance companies and found one that offered him the same
coverage at a significant discount from what he was paying
before the change in his credit. And this is an example of
competitive markets reaching an optimal outcome.
While competitive markets are very effective at making
goods and services consumers want available to them, critics
have voiced concerns that when a drop in credit is unrelated to
insurance risk, some individuals could be mistreated by
insurance scoring. In response to such concerns, almost every
State has regulations in place to recognize the benefits of
scoring, while limiting its use in these certain scenarios.
I think it's worth noting that many insurers offered the
same protections as these regulations require before the laws
were enacted. And this is another example of competitive
markets creating an optimal outcome.
Thank you again for this opportunity to share with you
today. I look forward to addressing your questions.
[The prepared statement of Dr. Powell can be found on page
193 of the appendix.]
Chairman Watt. Thank you, Dr. Powell, and I thank all of
the witnesses for their testimony.
I will recognize myself for 5 minutes for questions. Not
for the purpose of discounting your testimony, but for the
purpose of making sure that we understand that there is some
vested interest that the members of your organization have in
this, the members of your organization provide the credit
scoring that insurance companies rely on? Reports?
Mr. Pratt. To clarify, Mr. Chairman, our members do two
things. They provide the underlying credit data, the credit
history that is the basis for the score, and in some cases they
may be the score provider; in some cases there may be a third-
party company that is providing the score which is used by the
insurer.
Chairman Watt. Okay. Would you have access to information
about what part of your members' business is related to
providing insurance credit scoring as opposed to other
information?
Mr. Pratt. I can't answer the question here at the table.
Chairman Watt. I understand that. But would you have access
to the information if we ask you to obtain that?
Mr. Pratt. I don't know, because it might be market-based,
and publicly traded companies sometimes make different
decisions about what they want to make public and what they
don't, Mr. Chairman.
Chairman Watt. Okay. But I suppose it would vary from
company to company?
Mr. Pratt. I have no doubt that different companies would
claim different market shares.
Chairman Watt. Mr. Neeson, your company did not use credit-
based insurance scores until 2000, is that correct?
Mr. Neeson. That's correct, sir.
Chairman Watt. And what were the factors that you were
using prior to your use of credit-based insurance scoring?
Mr. Neeson. The industry is very competitive, and--
Chairman Watt. I'm talking about your company.
Mr. Neeson. Well, I'll just say that the sort of things
that we used in the past would be age of driver, marital
status, sex of driver, use of car, location of the risk, the
limit of liability, the value of the car, the age of the car.
Things like that. Prior accidents--I don't know if--
Chairman Watt. Those kinds of things that people would
normally associate with having some connection to risk when
you're driving?
Mr. Neeson. I would say it would also include things that
we talked about earlier, like good student driver discounts.
Chairman Watt. Okay. What percentage, how much weight does
your company give to credit scoring versus those other more
traditional underwriting factors?
Mr. Neeson. We don't deal in weights in the pricing of a
vehicle, but I do know about the kinds of pricing factors--
Chairman Watt. Well, if I walked into your office--
Mr. Neeson. Okay, the weight--
Chairman Watt. You're saying you wouldn't--I mean you have
to have a weight.
Mr. Neeson. Yes. For example, I don't know if you've had a
teenage driver before, but a teenage driver added to the policy
would increase the rates substantially, 3 times, and so forth,
whereas the value of having a--
Chairman Watt. Well, would the weight of a credit report be
less for a teenage driver?
Mr. Neeson. Far less than that. A 16-year-old driver versus
someone with--
Chairman Watt. Mr. Neeson, please listen to my question.
Mr. Neeson. Sure.
Chairman Watt. What percentage weight do you give to credit
scoring in determining rates? I appreciate the information
about being a teenage driver, but this hearing is about credit
scoring, and so it's that particular factor that I'm trying to
find out what weight you give to it.
Mr. Neeson. I don't have the exact figures, but I would say
that in automobile insurance, that would range from about a 15
to 20 percent discount to a surcharge of 50 percent, something
in that neighborhood.
Chairman Watt. I am not talking about the discount; I am
talking about the underwriting decision.
Mr. Neeson. Westfield does not use the insurance score as
any part of its weighting or whatever, for underwriting. We
only use it for pricing, not for underwriting.
Chairman Watt. Oh, okay. Yes.
Dr. Powell, you have talked about the predictive value of
insurance-based scores for risk. Let me be clear about whether
you are talking about risk or claims. Which one are you talking
about when you say risk?
Mr. Powell. The risk of claims. I don't see where there's a
difference if we're talking about an insurance mechanism.
Chairman Watt. Well, there is a difference if somebody has
an accident and elects not to file a claim.
Mr. Powell. Not in the amount of money that is paid out by
the insurance company.
Chairman Watt. Okay. So you are talking about the actual
amount of claims that people, pay is what you are talking
about.
Mr. Powell. Yes.
Chairman Watt. Okay. That's fine.
My time has expired. I will recognize the ranking member
for 5 minutes.
Mr. Miller. Thank you. We've talked about a lot of things.
We've talked about underwriting standards, pricing, loss
ratios, profitability, risk have all been used in
conversations, premiums based on higher risk, lower risk.
Insurance companies are not nonprofit organizations; they are a
for-profit industry.
And in order to set premiums, you have to consider risk and
the probability of a loss or how many losses, and the factors
associated with it.
And I know, Mr. Poe, you testified that CURE does not use
credit scores because of your belief that they are proxies for
incomes. And that's really completely different than most of
the witnesses we have had today; their opinion has really not
said that; they have not raised that issue. And yet FHA has now
determined they're going to use a new policy; they're going to
use credit scores for risk-based pricing. And in fact, the
report they just completed, they found that lower-income FHA
borrowers have average FICO scores that are higher than for
borrowers with larger incomes.
I know that's kind of shocking to people, but that's how
FHA is going to do it in the future. Do you think that's a
reasonable and appropriate thing for FHA to do?
Mr. Poe. You know, I haven't seen the study, but I don't
know, so I can't really comment on that.
Mr. Miller. Okay. Now the study that was generated earlier
by the FTC and the Federal Reserve basically said that for
whatever reason, credit scores is a predictor of risk loss. I
mean, that is what their report came out and said.
Mr. Poe. Correct.
Mr. Miller. Okay. Thank you. Your argument was different,
then. I just wanted to get an opinion.
Okay. Mr. Neeson, you were testifying on behalf of PCI, who
adamantly opposes H.R. 5633 and H.R. 6062 because these bills
would increase prices and reduce the availability for most
consumers. What are the facts? And are most of the consumers
helped or hurt by credit score usage, in your opinion?
Mr. Neeson. From our own company experience, the vast
majority of our customers are benefitting from the use of
insurance scores. In my written testimony about our auto home
package policyholders, three-quarters of those received
discounts and another 15 percent or so are neutral.
In the Arkansas surveys that are run annually, comparable
numbers of people benefiting, and neutral persist. In my
testimony earlier it was 90.3 and 8 percent. Amazing
percentage.
Mr. Miller. Can you take a credit score on an individual,
is there any way you can glean from that gender or race from
that credit score?
Mr. Neeson. No.
Mr. Miller. Okay. So it's pretty much a neutral score. You
wouldn't have an idea if it was male, female, black, white--
Mr. Neeson. In the reverse, all insurance companies really
have no idea of the race of their customers. Westfield would
have no idea of that. And what we do know is that the vast
majority of our customers do benefit from insurance scores. By
eliminating that I also know that there would be a vast number
of people then that would be severely harmed, and those would
include groups such as senior citizens on a fixed income, you
know, lower-income people who are working hard to pay their
bills, to pay their gas bills, to pay their electric, and food
costs. In these economic times, it would be very difficult for
so many people to have higher payments such as that.
Mr. Miller. And if insurers are unable to price for risk,
for example, because credit score usage is banned, does this
increase overall costs for consumers, in your opinion, as the
Federal Reserve Board found, because you have to charge higher
premiums for risk uncertainty?
Mr. Neeson. Yes. In my opinion, that would be the case. It
would be no different. I'm sure that all of you are aware of
the difference between Treasury bonds and junk bonds; the risk
of each causes a higher rate to be charged for the risks with
higher risk of loss and insurance.
Mr. Miller. Mr. Pratt, you found that the elderly have
better scores on average? Based on reports we have seen, the
Federal Reserve said that seniors have better credit scores on
average. And Mr. McCarty was on the panel before, and he
testified to the opposite. But who, in your opinion, would you
believe to be correct in that?
Mr. Pratt. The preponderance of the evidence supports the
conclusion that seniors more often have higher credit scores.
By the way, the reason for that is in part because they have
been in the marketplace longer, so as they have built a history
over time and demonstrated--you can have a consumer with a 1-
year credit history and a consumer with a 50-year credit
history. And even if their credit reports looked exactly the
same, there would be some difference in the score, because one
consumer is demonstrating good hard work and good behavior for
a year, and the other consumer is demonstrating it for 50
years.
Mr. Miller. Okay. So banning it could particularly harm
seniors?
Mr. Pratt. That's possible. That's a possible outcome.
Mr. Miller. I'm in the situation you're in; I'm out of
time, Mr. Chairman, so I yield back. Thank you.
Chairman Watt. Could I ask unanimous consent for 30
additional seconds and ask you to yield just on this last
point, because I'm trying to square Mr. Pratt's testimony and
Mr. Neeson's testimony.
Mr. Miller. Oh, sure.
Chairman Watt. Seniors have higher, better credit scores,
yet Mr. Neeson said they have--one of the traditional factors
that they were taking into account was age. That suggests to me
that seniors may have higher incidents of accidents. Is that
correct? Or am I wrong about that? Mr. Neeson?
Mr. Neeson. Seniors do have better insurance scores from
what at least I've heard.
Chairman Watt. I got that from Mr. Pratt. I'm talking about
their driving record.
Mr. Neeson. I've seen where people 60 years old and so
forth have better driving, and as they get to be 80 or 90 years
old get worse. And that's what I've seen.
Chairman Watt. So you would factor that in--
Mr. Neeson. However, the improved insurance score, these
work independently. It's like your value of your car, and the
location that you--
Chairman Watt. I appreciate it. I understand they work
independently, but they work counterproductively, it seems to
me. If you are taking credit scores into account, and seniors
have better credit scores, then you must be saying they have
less accidents, or they at least, according to Dr. Powell,
submit less claims for this to make sense. Otherwise--but I
yield back to the--
Mr. Miller. Can I have 30--
Chairman Watt. Sure.
Mr. Miller. But I heard the testimony earlier that
underwriting standards are different, and that would be the
loss ratio, accidents and stuff that they tend to have. And
then this would be used after that. Is that not correct?
Mr. Neeson. That's correct. Different insurance companies
do--
Chairman Watt. In his company.
Mr. Miller. Yes.
Mr. Pratt. Could I just add one additional comment, Mr.
Chairman?
Chairman Watt. Yes, sir.
Mr. Pratt. It's a personal experience. When I was 24 years
old in Texas--
Chairman Watt. Does this have something to do with aging?
Mr. Pratt. It does. It has something to do with aging.
Chairman Watt. All right.
Mr. Pratt. And I needed insurance and I needed a new car,
so I went out and bought my new car, and I got my insurance
policy. And because I was 24 years old, my insurance premium
per month was higher than my car payment per month.
Chairman Watt. I'm sure that has something to do with what
we were just talking about.
Mr. Pratt. When I was 25, my insurance premium--
Chairman Watt. You're going to have to make this point a
little bit quicker, because my time--
Mr. Pratt. Well, the bottom line is I think the age issue
works on both ends. In other words, had a credit score been
used, it might have been a counterbalance and actually caused
the insurance company to be able to rate me differently and to
allow me to pay a lesser price, and not to have used age as the
preponderant factor in determining my premium. I think it's
just worth the consideration that it works on both ends of the
scale, Mr. Chairman.
Chairman Watt. I hope somebody understands the value of
that. Because I don't. I'm sorry.
Mr. Hunter. Mr. Chairman, could I add something?
Chairman Watt. No. All of our time expired 5 minutes ago. I
am sorry, but I don't want to penalize the other members of the
committee.
I recognize the gentlelady from California for 5 minutes.
Ms. Waters. Ms. Rice, with the National Fair Housing
Alliance, will you please explain the disparate impact on
racial minorities from the use of credit-based insurance
scores?
Ms. Rice. Well, as is identified in our testimony,
discrimination in the marketplace cannot be excised from the
credit repository data. And there are so many instances of
discrimination in our marketplace where African Americans and
Latinos are disparately impacted or disproportionately
negatively impacted, such as the current foreclosure crisis
that we are experiencing. If you compare the rate of
foreclosures across various demographic designations, you'll
see that African Americans and Latinos are harder hit by that.
Now they're harder hit not because they posed a greater
risk, but they're harder hit because they were
disproportionately marketed loan products that were non-
performing and that were unsustainable. It had nothing to do
with their individual level of risk; it had everything to do
with discrimination in the marketplace. And we feel that using
credit information, particularly at this juncture, is going to
do more harm than good, and we're going to see even greater
disparities.
As you've heard other people say before, African Americans
and Latinos score anywhere between 10 points and 35 points
lower than their white counterparts. And again, we argue that
is not because they are more intrinsically or inherently risky,
but due to discrimination in the marketplace.
Ms. Waters. Thank you. Mr. Hunter, do you agree with that?
Mr. Hunter. Yes. I believe that's correct, and I do believe
that's why people are paying more if they're lower income and
if they're minorities for insurance.
Ms. Waters. Mr. Poe, do you agree with that?
Mr. Poe. Yes, I do.
Ms. Waters. Mr. Neeson, do you understand that?
Mr. Neeson. His answer?
Ms. Waters. No. I have been talking with the three
witnesses who preceded you about the disparity in pricing and
how it impacts minorities. And I wanted an explanation so that
everybody could hear it, to see if you understand it or you
agree with it or disagree.
Mr. Neeson. I know that insurance scoring does work within
races and nationalities. That was based on the FTC study. It
shows--
Ms. Waters. I'm sorry, what did you just say? It works
within? What does that mean?
Mr. Neeson. One of the charts towards the back of the
survey shows that those individuals with better insurance
scores by race had lower loss costs. And one of the things that
we saw--and again, we are not privy to any racial information
of the company; but people that live in perhaps urban areas or
whatever may have prior claims. And what I have seen is that
the--
Ms. Waters. My question to the first person was to explain
the disparate impact on racial minorities from the use of
credit-based insurance scores. She did an explanation. My
question to you was: Did you understand that, what she said?
Mr. Neeson. I did understand that.
Ms. Waters. Do you agree with that?
Mr. Neeson. No, I don't.
Ms. Waters. Thank you.
Let me go on to Mr. Pratt. Did you hear what was explained
by Ms. Rice?
Mr. Pratt. I did.
Ms. Waters. Do you agree with that?
Mr. Pratt. I do not.
Ms. Waters. I beg your pardon?
Mr. Pratt. I do not.
Ms. Waters. You do not.
Okay. And lastly, Mr. Powell, did you hear the explanation
about the disparate impact on racial minorities from the use of
credit-based insurance scores? Do you agree with that?
Mr. Powell. I heard it; I disagree with the conclusion.
Ms. Waters. Okay.
Mr. Chairman, your bill--if I may--I know this is a little
bit unusual--your bill was introduced because of the disparity.
And you said that you had some documentation for it. Would you
repeat that documentation?
Chairman Watt. My documentation is based on the FTC's
report that credit scores in this case are a proxy for race.
And I think--well, that's what we based it on, yes.
Ms. Waters. I see. Mr. Neeson, have you seen the report?
Mr. Neeson. Yes, I have.
Ms. Waters. And you think that the FTC is wrong?
Mr. Neeson. I saw that it showed little proxy effect.
Ms. Waters. I can't hear you.
Mr. Neeson. I heard that it said little proxy effect.
Ms. Waters. What does that mean?
Mr. Neeson. Negligible.
Ms. Waters. What percentage? How much? How little?
Mr. Neeson. I don't know.
Ms. Waters. Mr. Pratt, have you seen the report or read the
report?
Mr. Pratt. I have.
Ms. Waters. Do you think it's wrong? Do you disagree with
that?
Mr. Pratt. I think the report shows that with many
different underwriting factors, if you pull it out on its own
and you don't consider it in the context of the other factors
used in the decision, you might find some kind of proxy effect;
but I think the key point here is that it was a negligible or
minimal proxy effect.
Ms. Waters. Not enough to be concerned about?
Mr. Pratt. Well, I think the insurance industry, and I
suspect all industries, are always concerned to make sure there
is not a sizeable proxy effect. Nobody wants that in the real
market.
Ms. Waters. But if it is a proxy effect, it should be
corrected. Is that correct?
Mr. Pratt. I think that if--I don't believe that the credit
scoring system or the credit reporting system we have today is
an enabler of the kind of proxy that I think we're talking
about here.
Ms. Waters. So the FTC report was wrong?
Mr. Pratt. The FTC report suggests minimal proxy effects.
You might get that with education. You might find that with
geography. You might find that with other factors. And I think
that is what is so key in this discussion is that you can hold
out any individual factor and potentially find some effect that
might speak to race or might speak to ethnicity or might speak
to income. I think that's really the key.
Ms. Waters. Mr. Powell, I saw you shaking your head. That
FTC report is just wrong, right?
Mr. Powell. That specific result I would take issue with. I
do not believe that it would withstand any sort of objective
scrutiny, based on the way it was calculated. If I were
reviewing that as an academic peer reviewer, which is a role
that I take on frequently, I would not accept that as something
that could be stated as a conclusion, based on the measurement.
Ms. Waters. I see. Given your academic and intellectual
review of the study, could you respond to this committee with
you conclusion, based on the study that you have alluded to?
Mr. Powell. Based on the FTC study--
Ms. Waters. Yes--
Mr. Powell. From the results that they present, I would
conclude that there is not a detectable proxy, that the result
they get is invalid, and they all but say that in their report,
that--
Ms. Waters. Would you present that to this committee? Could
we ask you to give us your conclusion in writing, based on your
review and your study?
Mr. Powell. I would be pleased to, yes.
Ms. Waters. I'm not simply asking for the conclusion, as
you are giving it now--
Mr. Powell. Oh, yes. Sure--
Ms. Waters. But because of your intellectual study, I would
like to see how that is set forth. Thank you.
Mr. Powell. I'd be happy to.
Ms. Waters. I yield back the balance of my time.
Chairman Watt. I thank the gentlelady.
The gentleman from Texas, Mr. Green, is recognized for 5
minutes.
Mr. Green. Mr. Chairman, because Representative Boren needs
to leave, may I switch places with him, please?
Chairman Watt. I would be delighted to have you switch
places with Mr. Boren.
Mr. Green. Thank you.
Mr. Boren. Thank you, my good friend, Al Green, and Mr.
Chairman.
I just have one question, and starting with Mr. Neeson,
going to Mr. Hunter, I would like your response on this. As
Kevin McCarty testified in our earlier panel, there are
inherent weaknesses in the credit reporting system. Though
reports vary, there is no question that many credit reports
contain mistakes, and it is a lengthy process to correct
mistakes, and on the credit report of our constituents and your
consumers.
Additionally, the methodology used in credit scoring is
opaque to customers, leading to greater confusion and hurdles
in obtaining and maintaining a good credit score. Some business
practices in my State of Oklahoma allow a consumer to obtain a
policy with their current credit score and their premium with
that company will not go above the pricing floor due to this
credit score change, due to any credit score change. In fact,
the consumer's better credit gets a proportionate decrease in
the premium.
So basically, this. If you start getting bad credit--after
I go in to meet with my insurance agent, and I get a premium
let's say on an automobile, if I have bad credit after that, my
premium can't go down because my credit rating goes down. If my
credit rating goes up, I actually save money. And so that is
kind of a unique thing that is happening in Oklahoma.
What do you all think about that practice? And is that
something that our committee needs to kind of look at, at the
Federal level? Starting with Mr. Neeson, going to Mr. Poe.
Mr. Neeson. Thank you, Congressman. Again, the industry's
extremely competitive and the pricing algorithms for each
company vary dramatically, not only in the factors used and the
approach used.
For example, as you mentioned, there are a number of
companies that use credit at the initial issuing of the policy
and then either don't use it later or only use it as an
improvement factor. There are also regulations by different
States that may or may not require review of credit over, you
know, different years.
But the short answer to you is yes, many companies do look
only for the improvements, so that it can be used in a positive
fashion, again, to retain customers. It's very hard to sell new
customers; they want to keep them, so they can continue to have
those customers as customers.
Mr. Poe. Thank you, Congressman. Actually, I think that
probably a bigger impact of the use of education and occupation
is far greater than the discussion on credit scores, because if
you study the use of education, whether you have a 4-year
college degree, or whether you have a masters degree, or
whether you work in a white-collar high-paid traditional
occupation, it is far greater in the impact of every person, in
particular minorities and lower-income people.
So to be honest with you, even if you adopted some sort of
practice like that, dealing with credit scores, you would not
escape the inevitable impact that education and occupation has
in our industry. So I don't think it would make any difference,
to be honest with you.
Mr. Boren. Thank you. Ms. Rice?
Ms. Rice. I think I agree with Mr. Poe that first of all,
for consumers sort of coming into the system, you are going to
be disproportionately affected, just by sheer virtue of the
fact that you are using a scoring mechanism, you are going to
be disproportionately negatively impacting African-American and
Latino customers coming into the system. So to say that we are
going to disparately impact you coming into the system, but
you're not going to have to pay a higher premium beyond the
higher premium that--the inappropriately higher premium that
you paid coming into the system, is not an adequate answer.
Mr. Boren. Okay. Mr. Hunter?
Mr. Hunter. Well, if I put myself into the position--I
don't agree with the use of credit scoring, as I've indicated--
if credit scoring works, then the system you just described
makes no sense. I mean if credit scoring really works, and
somebody gets a worse score, their rate should go up. And if it
gets a better score, rates should go down. And it's not a zero-
sum game. Because if the score doesn't go up, that means
there's less money coming in from the credit scoring system,
which goes into the base premium. That means the people with
thin files and all are going to pay more. Today people with
thin files pay too much because the neutral rate has off-
balance built in from inadequate credit scoring collections
like that.
And so the neutral people are going to have pay more, if
you don't raise them on the people who are getting worse. But I
don't think the whole system should be used at all. But if you
use it and you really believe in it, then it makes no sense to
cap it.
Mr. Boren. Thank you all so much. I yield back.
Chairman Watt. The gentleman from Texas, Mr. Green?
Mr. Green. Thank you, Mr. Chairman. Let's start with Mr.
Neeson. Mr. Neeson, do you agree that persons who are more
wealthy tend to elect to have higher deductibles?
Mr. Neeson. I've not done any study on that, so my opinion
might be that that would be the case. But--
Mr. Green. If this is true--
Mr. Neeson. I don't have that information.
Mr. Green. I understand. You don't have the empirical data.
But it seems to suggest--my question would seem to suggest that
if you have more money, you might have a $1,000 deductible as
opposed to a $250 deductible. You have not found that to be the
case? Persons who have more money tend to take out higher
deductibles?
Mr. Neeson. I have observed that people who want to manage
their prices, their costs of insurance, take out higher
deductibles. Our agents often encourage customers to have
higher deductibles, so that they can manage the expense of
their insurance better. That's what I have observed.
Mr. Green. And this means, of course, that these persons
with higher deductibles are prepared to pay a higher amount of
money for any infraction, for an accident.
Mr. Neeson. I think--
Mr. Green. Or they should be, because if you have a $1,000
deductible, you're going to pay the first $1,000.
Mr. Neeson. Or they would get a loan to pay for it. Or--
Mr. Green. Right--
Mr. Neeson. If they think they aren't going to have a
claim.
Mr. Green. But generally speaking, this would benefit a
person who has the money to pay that $1,000 deductible,
wouldn't it?
Mr. Neeson. Very wealthy people may not even choose to
purchase physical damage on their cars.
Mr. Green. We're not talking about very wealthy, we're
talking about people who are more wealthy than some other
people.
Mr. Neeson. I don't--
Mr. Green. See, I'm not a very wealthy person, but when I
was--I've had the privilege of being poor, and to have acquired
some amount of status in life. And when I was poor, I had the
lowest deductible I could get and I used my insurance every
chance I could whenever something happened. But when I gained a
little more status, then I decided I wanted to get a $1,000
deductible because I'll pay the first $1,000 to keep you from
going up on my policy. That's what I did. Does that make sense?
Mr. Neeson. That does make a lot of sense--
Mr. Green. I hope it makes sense, because that's what your
agents encourage us to do when we can afford it.
Mr. Neeson. The premium would be higher for the lower
deductible, so a lot of people do use higher deductibles.
Mr. Green. And do you agree that generally speaking,
minorities in this country--just as a matter of fact--tend to
be the persons who are less wealthy than others? Generally
speaking?
Mr. Neeson. I'm listening to you. I don't have any
information on that.
Mr. Green. You don't have any observations? Have you kind
of looked around?
Mr. Neeson. Certainly.
Mr. Green. Have you not noticed? You read the newspaper?
Mr. Neeson. Yes.
Mr. Green. Okay. All right. So it's a fair statement, I
think.
Mr. Neeson. Yes.
Mr. Green. Well, let's just see how your colleagues feel.
Do you agree that minorities tend to be poorer than some others
in this country? If so, raise your hand.
[Show of hands]
Mr. Green. Okay. Everybody seems to agree with this, Mr.
Neeson.
Mr. Neeson. There are wealthy people of all races.
Mr. Green. Yes, there are. But minorities don't tend to be
in a disproportionate number of the more wealthy people of all
races. Do you agree?
Mr. Neeson. Yes.
Mr. Green. Okay. There are some things that we just have to
agree to, we can take notice of, without having to have
empirical evidence. So if this is the case, then probably
minorities are going to be persons who are not going to have
the higher deductibles because generally speaking, you have to
be prepared to pay that deduction before you can get your car
back and make it road-worthy again.
Let me go on another point quickly. You mentioned teenaged
drivers increasing the rate paid by some multiple. What was
that number again? Teenage drivers or a teenage driver coming
onto a policy?
Mr. Neeson. I've--three or four times.
Mr. Green. Three or four times whatever the current premium
is? Now this is somewhat enigmatic for me, because I was born
into a family that happens to be paying a high premium because
of my father or my mother having a low credit score, and now
because of my birth--I have no record of driving poorly, I have
no credit history, but their premium will go up three or four
times, some multiple, just because I was born into the family.
Is this true? Of course it is.
Mr. Neeson. The age of the driver? That had nothing to do
with credit.
Mr. Green. No, but you're going to increase the premium
some multiple, based upon what the mom and pop are already
paying, right?
Mr. Neeson. That's correct.
Mr. Green. Okay. So this driver has no history, has no
credit score, but that driver is going to increase the family's
premium some multiple simply because he was born.
Mr. Neeson. At least with Westfield, the insurance score is
based on the parents, so he would benefit from the better
insurance score of the parents--
Mr. Green. I understand. But if the parents don't have--
suppose they have a poor insurance score, then the parents will
pay more because the child was born.
Mr. Neeson. Because of the age. And you would probably
agree that youthful drivers do present a higher likelihood of
future--
Mr. Green. I do--but the question is should the multiple
that the parents pay be increased, based upon that driver being
born into the family, when the multiple is already high? You
see, if you neutralize that driver, then it would be okay. But
now what you're saying is that family is going to pay some
multiple because that driver was born, and that multiple is
based upon the credit score of the parents, not the driver.
Mr. Neeson. I do know that the majority of people do have
better insurance scores. And so the parent would likely benefit
from that. If we add--
Mr. Green. But those that don't, does it benefit those who
don't have better credit scores?
Mr. Neeson. They would be paying higher, yes.
Mr. Green. They would be penalized?
Mr. Neeson. Yes.
Mr. Green. Okay. Thank you.
Thank you, Mr. Chairman.
Chairman Watt. I thank all of the members and the witnesses
for their participation in this hearing.
Let me just ask Mr. Neeson one question, if I may. A public
policy that says one should not be charged a higher insurance
rate because of their race, that seems reasonable, then? Okay.
So--
Mr. Neeson. For automobile insurance?
Chairman Watt. Automobile or homeowners'. So if we just
passed a law that said, ``Thou shall not discriminate in rates
based on race,'' and gave individuals a private right of
action, would that be preferable to what we have proposed here?
Mr. Neeson. I know of no company that uses race for
pricing--
Chairman Watt. I didn't ask you that. I said, would that be
preferable to what has been proposed here? I mean, just a
straightforward prohibition on using anything that
discriminates, and give individuals the right to enforce it.
Mr. Neeson. I as a person, as an individual, feel that it
would be wrong to charge by race for automobile and homeowners'
insurance.
Chairman Watt. Okay. Thank you.
I appreciate it.
The Chair notes that some members may have additional
questions for this panel, which they may wish to submit in
writing, as well as for the earlier panel. Without objection,
the hearing record will remain open for 30 days for members to
submit written questions to these witnesses and to place their
responses in the record.
We thank every single one of you for your participation
today. We have been called for votes, and the hearing is
concluded anyway, so we came out just in time. Thank you so
much. The hearing is adjourned.
[Whereupon, at 1:09 p.m., the hearing was adjourned.]
A P P E N D I X
May 21, 2008
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]