A View From The Press Box

This will open a Pandora's box of problems, not the least of which will be allegations of invasion of privacy.

If my broker asked me, I would tell him or her that it is none of his or her business--nor is it the business of the insurance company, since they cannot use race, religion, national origin, gender or sexual preference in making an underwriting decision.

Insurance decisions should be based only on the risk the insurer is being asked to take.

SAM RESPONDS:
I hear you, Barry. But how have such personal objections this been handled/overcome in the banking situation Mr. Squires cited? Surely they must face the same resistence, but the law is the law, no?

BARRY HAS THE FINAL WORD:
I understand your position, Sam, but "the law is the law, no?" does not mean it makes sense. As Mr. Dickens once said: "The law is an ass."

I see no reason for the insurance industry to comply with the same silliness their brothers in the banking industry have fallen prey to over the years of the "politically correct."

The law that hurts is not a benefit and many, as Rev. M. L. King said, should be disobeyed.

I do not want insurance carriers to have information that can be used to discriminate based on race, religion or national origin. They should discriminate on one ground only: "Is this a risk the insurer is willing to take?"

As the California Court of Appeal stated:

"An insurance company is entitled to determine for itself what risks it will accept ... It has the unquestioned right to select those whom it will insure, and to rely upon him who would be insured for such information as it desires as a basis for its determination to the end that [198 CalApp3d Page 181] a wise discrimination may be exercised in selecting its risks. (Robinson v. Occidental Life Ins. Co. (1955) 131 Cal. App. 2d 581, 586 [281 P.2d 39]

Race, religion, national origin have no place in making the wise discrimination. It is time, Sam, that the law is applied and insurers make wise discrimination on the risks they take based on facts about the risk only.

This is an absolute no-win situation for insurers regardless of how the numbers turn out.

If it turns out that there is no discrimination (and I haven't seen anything to suggest it in my 16 years of P&C insurance with 3 companies), consumer groups will then acuse insurers of HAVING the information to develop discriminatory underwriting models and using it to find ways not to insure the poor or minority folks.

If the opposite is true, even unintentionally, insurers will have the "I told you so" card played against them.

Add to that the fact that many potential insureds will refuse to answer the question and get incensed about having to be asked, and you can see that there is no way for any insurer to prevail.

I suppose that it would be a bit too obvious to point out that if Prof. Squires is correct in that he's simply asking for the same data to be collected by insurers as is collected by mortgage lenders, then it should be possible just to obtain the data from both industries and merge the two datasets, avoiding annoying innocent consumers further.

Of course, obtaining and merging such datasets would be a royal pain in the tuckus, and insurers will fight to prevent any linking of such data.

But I think the proposal would have a slightly better chance of becoming reality if it didn't involve checking consumers' skin pigmentation at renewal, or sending out questionnaires asking, "what's in your wallet?"

Sam, you made me laugh at your comparison of the NAIC to evolution!

I am wondering if the willingness to disclose that type of information depends on the relative "desire" for that product.

In other words, I WANT to get that mortgage, but insurance companies just have their hands in my pockets, so I have no motivation to cooperate.

I like Charlie's idea of merging the two sets of data. After all, so many of the banking giants are in the mortgage and insurance biz these days!

How would people of blended races be categorized?

It seems like more and more members of society come from racially varied backgrounds.

Wouldn't that nullify race data to some degree?

SAM RESPONDS:
Good question! It seems that those of mixed heritage, such as Barack Obama or Halle Berry, categorize themselves as black, perhaps because that is how they appear. That would probably be the best way to maintain the records, since people (insurance underwriters/agents?) would react to the race they perceive the indvidual to be.

Few ideas have done more to poison race and gender relations in this country than the pernicious doctrine that rights belong to racial and gender groups rather than to individuals.

That doctrine, of course, provides the foundation for the disparate-impact theory of discrimination, and Sam is surely correct to suggest that facilitating disparate-impact lawsuits is the motivation behind the Squires proposal.

The framers of our Constitution, who warned against the “mischief of factions” and who carefully crafted the Bill of Rights to protect the liberty of individuals—not groups—would be appalled at the notion that a perfectly rational insurance practice (e.g., risk assessment) that is neutrally applied to each individual and utterly blind to race, gender or income, could somehow be regarded as “unfairly disadvantaging one racial, gender or economic group” (to quote Sam) because of statistical disparities in the way it affects particular groups.

Insurers are right to worry that the Squires proposal would force them to violate their customers’ privacy. But we should also take a more expansive view of the issue, because the Squires proposal aims ultimately to turn the concept of unfair discrimination on its head.

For discrimination to be unfair, it must be intentional and directed against identifiable individuals because of their race, gender or income. It is simply not possible to unintentionally and unfairly discriminate against a group.

We are told constantly, especially during the current presidential election campaign, that our society must “move beyond race.” We will never be able to do so if people like Prof. Squires continue to insist that law and regulation treat people not as sovereign individuals, but as interchangeable units in demographic categories defined by race, gender and income.

My vote goes with Barry.

The law is that no one can discriminate (or make an underwriting selection or rejection) based on race, religion, national origin, gender or sexual preference.

Add to this the redlining laws that also restrict geographic discrimination. If the industry collected such data on an individual risk, an underwriter could be tempted by a person's potential bias to discriminate. It is better NOT to go down this path.

I am probably opening up a can of worms, but don't insurers unintentionally discriminate against lower-income people all the time?

Insurance companies inspect the homes that they insure. If the homeowner does not maintain the home, they are given a list of "recommendations" to fix the home or the insurance policy will be cancelled.

If a homeowner is lower income, they may not be able to afford the repairs required by the insurance company. Their home coverage is then cancelled. Their agent then has to write the home with a sub-standard carrier--for more money in some cases.

There is nothing wrong with this. It is effective underwriting and I fully encourage this. The better the pool of homes being covered, the less likely there will be claims. There are lower premiums for everyone.

I think that requiring insurance carriers to gather and disclose race, gender, income level, etc. will be a detriment to all.

When it is determined that insurance carriers are unintentionally discriminating, some rights group will sue and carriers will be forced to change their underwriting requirements to be less discriminatory. This will lead to higher premiums for all.

Maybe I am making too much of this, but I just see this as a very bad path.

Bob Detlefsen misses one of the most elementary points of virtually any introductory social science class, and that is the fact that for many individuals their opportunities are limited by their race.

It is not a matter of group rights. Their individual rights are limited due to their race. The "group rights" argument is a straw man often trotted out by those who simply choose not to understand the nature of racial inequality in the U.S. and elsewhere.

And the fact that insurers do not collect information on the race of insureds is hardly evidence that race does not enter into underwritng and pricing decisions, either by intent or impact.

Knowing the location of a house, and the racial composition of neighborhoods, is sufficient for racial bias to often creep in, again either by intent or effect.

For example, if an insurer will not provide coverage to older homes or varies covereage by age of housing, this will have a dramatically racial effect because racial minorities are heavily concentrated in neighborhoods where the housing stock is older, though not necessarily any less insurable.

This is one reason why several insurers agreed to stop using age in their marketing in the settlement of several fair housing complaints a few years ago.

As a former political science professor, I have to wince at Gregory Squires’ pronouncement that “for many individuals their opportunities are limited by race.” This well-worn truism is a fine example of the intellectual laziness that has given social science a bad name.

Which individuals? Which opportunities? How were their opportunities limited? Were they limited by race or by some other factor(s)? How do we know?

He then declares, “Their individual rights are limited due to their race.” These are not “facts.” They are not even testable hypotheses. They are merely political slogans.

Regarding the example he gives, if the age and value of a home are correlated with claim costs, the use of these underwriting criteria is fair and reasonable regardless of whether they produce a “racial effect.”

Squires continually conflates “intent” and “impact,” completely ignoring my point that unfair discrimination results from intentionally treating individuals unequally, not from equal treatment of individuals that produces unequal results among groups.

Finally, one has to wonder about a scholar who tries to rebut an argument on the grounds that those advancing it “simply choose not to understand” what he claims to understand.

What he really means is that those who disagree with him are wrong because they don’t share his view of the world.

Bob Detlefsen can choose to deny the reality of racial discrimination in his world if he wants to do so. But if others follow, we do so at our own risk.

Paired testing by fair housing groups has consistently shown that whites and non-whites, as well as residents of white and non-white neighborhoods who pose identical risks, receive very different treatment from insurers.

Similar research has documented similar outcomes in employment, home purchase and rental, auto sales and elsewhere. There is no secret about "which individuals" or "which opportunities."

Obviously, not every racial disparity is attributable to intentional discrimination alone. But racial inequality is also not simply the outcome of intentional acts by identifiable individual actors.

The concept of disparate impact has been established by the courts. There is no "conflation." These are different forms of discrimination. But they both constitute discrimination nevertheless.

A range of public policies have historically treated different races differently, and some continue to this day. New Deal legislation creating Social Security, labor organizing rights, and other benefits left out domestic and agricultural workers. It was no accident that those excluded occupations were those where African-Americans were concentrated.

The FHA in its early years insured mortgages almost exclusively in all-white neighborhoods. Early underwriting manuals observed that "If a neighborhood is to retain stability, it is necessary that properties shall continue to be occupied by the same social and racial classes."

These are just some of the policies that have led to wealth disparities today. So, for example, when financial services utilize credit scores in their underwriting practices, they perpetuate into the future racial inequalities that originated in earlier public policies and private practices.

For further information I recommend Ira Katznelson's book, "When Affirmative Action Was White," Massey and Denton's "American Apartheid," and Kenneth Jackson's "Crabgrass Frontier."

As co-author of "The Housing Divide: How Generations of Immigrants Fare in New York's Housing Market," and an associate professor of sociology, I fully concur with Greg Squires's recommendation for the insurance industry to collect data on race, gender, and income status.

I'm not sure what world Bob Detlefsen is living in if he doesn't think race continues to play a role in shaping the opportunities and lives of individuals.

This year marks the 40th anniversary of the Fair Housing Act. As stated by Kim Kendrick, HUD Assistant Secretary for Fair Housing and Equal Opportunity, "Forty years after the passage of the Fair Housing Act, an alarming number of families are still being denied housing and still need the protections this landmark law offers."

Housing is inextricably linked to the opportunity structure (i.e., schools, access to jobs, healthcare) in American society. If blacks and Latinos are being denied access to this structure, it has implications not only for their lives but also for those of their children.

The insurance industry should step up to the plate and make their data available. Discrimination in all aspects of housing needs to be confronted in order for ALL people in this country to have equal access to the American Dream.

The problem with bromides such as "race continues to play a role in shaping the opportunities and lives of individuals" is that their very banality prevents us from knowing what is being alleged.

For example, when Geraldine Ferraro recently proclaimed that Barack Obama would not be a leading contender for the Presidency if he were white, she was affirming the racial determinism implicit in the Squires-Friedman postulate that race shapes the lives and opportunities of individuals.

Strangely enough, Ferraro's assessment was widely denounced. Do Squires and Friedman think her critics live in a dream world?

Prof. Friedman exhorts the insurance industry to "make their data available." Apparently she doesn't realize that insurers don't have the data that she and Squires demand, because they don't identify their policyholders and policy applicants by race, ethnicity or income level.

Hence, they cannot possibly discriminate on the basis of these characteristics. They prefer instead to discriminate on the basis of risk, since that is how they maintain solvency.

The Squires proposal would force insurers to become race-conscious when currently they are race-blind.

Neither Prof. Squires nor Prof. Friedman have explained how insurers' failure to inquire about their customers' race and income is responsible for people being "denied access to the housing structure" because of their race and income.

I am, thank G_d, not a sociologist--but I know insurance.

Insurance underwriting is, by definition, discrimination. The insurer has the right and obligation to select risks that will make a profit for the insurer the underwriter represents.

Gathering data on race, religion or national origin will not help. Better to gather data on the type of people prone to arson-for-profit, insurance fraud, insurance claims frequency, good housekeeping, and professional risk management.

Insurance should only be concerned with insuring one race--members of the human race.

It is appalling how often people say that insurers could not possibly discriminate because they don't even collect information on race, etc.

This flies in the face of consistent findings from paired testing evidence provided by several non-profit organizations demonstrating that EQUALLY QUALIFIED applicants who happen to be white and non-white are treated differently. That is, similar risks are often treated dissimilarly.

On occasion legal action has been taken to provide appropriate compensation. Fair housing attorneys, the Justice Department, and other authorities have taken such action. Insurance regulators, however, have been missing in action.

Even in the absence of such information, knowing the address of a property is often sufficient to know the racial composition of the area. When an agent recently indicated that a company he wrote for asked, "do the kids in the neighborhood play hockey or basketball," does anyone really think that company was interested in the recreational habits of the locals?

We are told time and time again that insurers only discriminate on the basis of risk. If only this were the case.

I have had occasion to look at a number of homeowners insurance situations to determine whether or not there has been discrimination. It has been my experience that agents almost immediately know the race or national origin of their customers in a call, and they act accordingly, and soemtimes discriminatorily.

Agents who know their communities often think that they know by property address, neighborhood, or sometimes even phone numbers or surnames the race or national origin of callers.

I have seen too many situations where agents, thinking that they know race or national origin, treat people differently because of those factors, and because of assumptions that they make about people based on race or national origin.

Keeping data by ZIP code will help prevent this type of discriminatory behavior. Agents who don't discriminate have nothing to fear.

And as someone who has worked for government agencies, i can tell you that data based on race and national origin are collected all the time by the government, and by federally-funded housers too.

There are effective techniques out there to prevent violation of personal privacy while also advancing the principles of equal opportunity and non-discrimination.

It does not violate federal or state laws to collect data about race or national origin; it DOES violate laws to consider race or national origin of individuals or of neighborhoods in making decisions.

And race and national origin are still too often considered in homeowners insurance decision-making.

As an act of disclosure, given the subject matter, I should mention up front that I’ve met Dr. Squires, respect him, and admire his work. So, it should come as little surprise that I wholeheartedly support his proposal.

I’m glad to see the subject spark healthy debate and to read such passionate responses. It would be disappointing to see any issue of this magnitude generate apathy, and while I may shake my head in disagreement at some opposing views--I shudder at the self-defeatist comments stating, in effect, “sounds nice, but it’ll never happen.”

As Dr. King so famously noted, “Lukewarm acceptance is much more bewildering than outright defiance.”

Let’s be real here--that the industry fails to collect or release data on the race, gender, or income level of their policyholders in no way suggests that insurers fail to take such characteristics into account--if only subconsciously.

Were this not the case, why would insurers criticize the proposal on the grounds that it might reveal “trade secrets?” If race, gender, or income level prove utterly immaterial to insurance companies, what “trade secrets” would be exposed?

We’d be hard pressed to find a marketing firm in the country that fails to collect demographic information on race, gender, or income bracket. Ever fill out a warranty card? You’re telling me the insurance industry is the exception?

The ad agencies they hire certainly take race into consideration when designing their campaigns. Perhaps you’ve noticed that the Allstate spot currently running on BET or during NBA games is not the same one they play on the Style network or CBS. Check out:

http://www.adgabber.com/video/video/show?id=546804%3AVideo%

http://www.adgabber.com/video/video/show?id=546804%3AVideo%3A71931

Personally, I’m not willing to accept the promise of lower insurance premiums as hush money to sit idly by while insurers engage in discrimination on the basis of race and gender. It seems a classic divide and conquer tactic to force members of the compressed and dwindling middle class to side with big business for fear of higher premiums.

Call me an idealist, (I’ve been called far worse) but I tend to value justice over cheap consumer goods. Most among us oppose child labor, for example.

Is it really acceptable to suggest that we should eliminate child labor laws, minimum wage laws, environmental protection laws, thus and so simply because these policies may raise the price of goods? Don’t values take precedence over"value"?

I would also humbly suggest that if anyone is attempting to presumptuously act as the arbiter of reality, it’s the person who seeks to invalidate the countless perspectives of those who either experience or acknowledge race and gender discrimination. That the fact you feel you’ve never encountered racism or sexism represents a reflection of privilege.

I wouldn’t expect someone to recognize that they’re NOT being followed in a store, NOT being pulled over, NOT being sold a mortgage at a higher rate than they qualify for, NOT being harassed because of white skin privilege, for example.

That the world functions, for you, the way you expect it to function is entirely consistent with what one should expect of a racist, sexist society. Thus, it’s not that your perspective is INVALID--only that your perspective, one that minimizes or occludes the operation of racism and sexism is not UNIVERSAL.

Now, if you can’t directly experience this discrimination for yourself, and you don’t wish to take people at their word when they report instances of discrimination on the basis of race, gender, sexuality, etc., how can this experience become “testable” if we can’t gather any meaningful data on it?

We’re talking about the burden of proof. How can one prove that the industry is not racially discriminatory? Because they say so? Because "you" say so? Forgive me if I’m less than convinced.

Collecting demographic data allows us to demonstrate the presence or absence of de facto discrimination. So, if you want something empirical, if you need further evidence as proof that racism and sexism persist in 2008, why not support measures such as this that prevent industries from hiding behind a veil of presumed race neutrality?

If we’re so confident that these companies are behaving ethically, that no discrimination is taking place, why not prove it?

If a company claims its food products are “organic,” we expect that they substantiate this claim for the benefit and safety of American consumers. If an insurance company claims to treat customers equally regardless of their race or gender, shouldn’t we have the ability to verify that claim as well? I don’t think a little accountability is too much to ask.

Thanks to everyone for their insight and relative civility.

Sam,

You talk about opening Pandora’s Box in the 4/21 issue of NU, and you refer to the article by Greg Squires in the same issue. (See Sam's April 22 blog entry for Prof. Squires' op-ed article.)

It seems to me that Mr. Squires quotes statistics that could be used to refute the theory that homeowners insurers discriminate against minority or poor homeowners.

He says “between 1993 and 2000, the share of loans made to blacks increased from 3.8 percent to 6.6 percent; the Hispanic share increased from 4 percent to 6.9 percent; and the share of low-to-moderate income borrowers increased from 19 percent to 29 percent.”

Correct me if I am wrong, but I believe none of those loans were made without proof of a property insurance policy. Lenders just do not let the policyholder go bare with respect to insurance coverage.

Mr. Squire himself states that these loans were made. It seems quite reasonable to assume that property insurers did not get in the way of these trends, but rather supported them in step with the banks.

What other conclusion can you reasonably make?

Given that, what justification is there for imposing yet another regulatory requirement on an industry that is heavily regulated already?

Additional regulations have real costs. Ultimately, the consumers pay the costs. What is the justification for adding more costs to the homeowners transaction?

Gosh, just what we need, another layer of beauracracy overseeing what to my mind's eye is an academic approach to social engineering--well-meaning or not.

I have been in the P&C business for over 30 years and a bank director for over 20. I am well familiar with reporting for such issues as CRA, and it takes time and money to do it properly.

The reality is that results are skewed when an area has little diversity, either ethnic or economic.

Conversely, they are equally skewed when an area has a concentration of ethnic and economic factors that lead to a lack of need or desire for financial services.

The information garnered can be massaged to show results that don't reflect the state of the industry.

At the "touch-labor level"--our level as agents--we will be reticent to engender resentment from insureds, prying further into their lives than they want, when many are buying what they perceive as a "commodity," not a service.

Agents, I'm afraid, will tend to be "creative" when completing the newly required information.

Indureds will be bombarded with additional sales material and satisfaction surveys based on the newly accumulated and analyzed information.

Hold fast, companies! Don't give into another idea to "open" the indusrty to the new wave of social management.

Private industry has done a good job, with a few bumps in the road, to police itself.