Use of Credit for Rating and Underwriting Insurance
By Werner E. Kruck CPCU, ASLI
May 10, 2002 - Few issues have so polarized the insurance industry and its customers as the current debate on whether it is appropriate to use credit scores in the rating and underwriting of automobile and homeowner policies. While efforts by the insurance community are focused on proving the high correlation of loss to credit score as justification for its use, consumer groups are equally aggressive in their attempts to discredit credit scoring as ineffective, irrational, and unfair to certain segments of the population. What appears on the surface to be a battle of statistics and mathematics is, in reality, an emotion filled battle for both sides.
To hear the insurance industry tell it, credit has perhaps the highest correlation with loss of any variable that they have ever seen, so they should be able to use it. The benefits, they say, are that many people will get reduced rates because credit will be used to isolate the minority who cause a disproportional amount of the claims. Laws to restrict or prohibit the use of credit then, are truly “anti-consumer” then, in the claims of industry spokespeople.
The consumer activists charge that use of credit represents a “black box” approach to selection and rating which is discriminates unfairly and is being used as a method to “redline”, or prevent insureds in low income urban areas from obtaining insurance on the same basis as the remainder of the population. While those arguments have been around for several years without getting much traction with regulators, the newest outcry of the battle is from individual consumers who are expressing their outrage to insurance regulators and legislators regarding the manner in which credit has been used in their specific circumstances.
Taking the most heat are the insurance agents, once again caught in the middle between the customers they are trying to serve and the insurance companies that simply expect them to find a way to explain things to insured that they cannot.
The big underlying is just what controls society should have to direct the pricing and selection of customers by insurance companies, and what rights insurance companies should have to choose their customers and set their prices. From a purely libertarian perspective, other than specific discrimination against race and creed, insurance companies should be able to do their own thing, and if customers don’t like it, they can go to another company and vote with their feet. In the U.S., our current social contract recognizes at least the legal ability of citizens, through their government, to place limits and restrictions on pricing and selection by insurance companies. In a sense, then, consumers are arguing about the type of controls need to be placed on the use of credit, and insurance companies are fighting the consumers’ right to controls.
Does the Correlation Make Sense?
The biggest argument that the industry uses is that the correlation of loss to credit is higher than any other variable that they have ever measured. This is actually a true statement, but not relevant to the debate since it simply explains the industry’s motivation, not its justification. The problem relying on correlation, as Robert Hunter, consumer advocate, points out, is that it could be used to justify all sorts of irrational schemes. A study in California about 30 years ago showed that people with dark hair had the worst driving records, and an Australian study indicated the same for Geminis compared to all of the astrological signs. Despite the established high correlations, no one is suggesting we use them. Indeed, the difficulty with credit is that nobody truly understands what causes the correlation.
Instead of admitting that they do not know, people like Joe Annotti, spokesman for the National Association for Independent Insurers make statements like “Our impression is if you’re willing to take risks on (the financial) side of your life, those behavior patterns are going to probably translate through to the rest of your life, including repairs on your car or driving fast.” Note the word “probably” and translate the statement as “we don’t know why, but we want to use it anyway.”
Minnesota Insurance Commissioner Jim Bernstein recently noted the inconsistency of the industry position in holding “behavior patterns” responsible for the correlation. The industry, he said, insists that, "individuals with higher credit scores are more responsible with their financial behavior and that this translates into their driving or their responsibilities as homeowners—but we have seen no other evidence that that is true." He has continued to press on how a random event such as a rainstorm, hailstorm or windstorm can "know who has a better credit score and who doesn't," since the industry claims that the correlation holds true even for weather related losses.
Is use of credit discriminatory?
It is a fact that all rating factors discriminate at the level of the individual. Rating segments insured into groups that have higher incidents of overall loss, but that is not true for the majority of members of the group. The legal question is whether a rating variable is “unfairly discriminatory”, in that it violates society’s rules for putting specific groups at a disadvantage.
In the case of credit, there is actually more evidence to show that it is not unfairly discriminatory than there is to the contrary. To start out with, all of the variables used to compute the credit score are objective items that the insured could be said to have control of. Typical are payment delinquencies, outstanding credit card balances, collections, foreclosures and bankruptcies. Contrast this with typical rating factors such as gender, age, and marital status. I can choose to pay off my credit cards, but charging me a higher premium because I am a male limits my options.
Why do we accept something that on the face of it is more discriminatory and clearly beyond our ability to change? First, because we understand it to be true – we believe it. We know that males tend to be more aggressive, and we also know that when there is a male and female in the car at the same time, more often the male drives, so there is more exposure to loss.
Also, we accept that fact about ourselves. I am not only male, but I feel pretty darn good about it, and I can even take pride in the fact that I have higher premiums because of the innate traits of the “Male Club.” When you tell me I have poor credit and have to pay more for insurance, regardless of my driving record, you have just judged me as an individual, and I am offended, whether you are correct or not. Moreover, I want to know how that has any possible connection with my driving.
So credit is discriminatory, just like all of the other rating factors, but perhaps not as acceptable by people that are judged to be of higher risk. That might indicate the likelihood that use of credit would affect a company’s relationship with its customers, but not that it should be illegal.
Is credit used purposefully to unfairly discriminate?
Consumer groups charge that companies are using credit history as a way to charge groups higher premiums to weed out certain customers, particularly minorities and the poor. In a nutshell, this is the core of the issue, because if true, is not only illegal, but opens companies to major class action lawsuits. The difficulty is that if consumer groups can show that the use of credit really does unfairly discriminate against a minority group, called “disparate impact”, they usually do not have to prove that the company was aware of the impact and that it used it on purpose to get large judgments. Simply the fact that people were damaged unfairly by the companies’ indiscriminate use will normally be enough to produce large judgments, and companies have usually settled quietly to avoid bad publicity.
While there are studies that have been done that appear to indicate that the use of credit is not discriminatory to minorities and poorer folk, there have been no studies that clearly show discrimination. The consumer activists are responding to the studies by attempting to disparage and invalidate them, and to give specific individual examples of situations and rationale that seem to be clearly wrong. Once again, for the rest of us, it is hard to make a determination when both sides are arguing on a different basis.
How do companies use credit scores?
Until the last few years, companies used credit scores mostly to help in underwriting new business. Companies set thresholds so if scores were beneath a certain level, they would do additional underwriting. Under the terms of the Fair Credit Reporting Act of 1970, if a company turns down an applicant solely on the basis of credit, they must notify the applicant that they have been refused based on their credit history and provide them with the information needed to obtain a copy of their credit information and pursue correction if necessary. To avoid such reporting and the follow-up responsibilities under the Act, many companies made it a standard rule never to deny coverage solely on credit. For underwriters, who are charged with both profitability and compliance, that has been translated as “when you need to turn someone down for credit reasons, find another legitimate reason for issuing the denial.”
Over the past two years, more companies have begun using credit aggressively to purge their books of their “unprofitable” business by non-renewing policies based solely on credit, or by charging higher prices designed to discourage that business through a process called “tiering.” With tiering, the company has multiple sets of rates at various premium levels that differ from each other only by the range of credit scores that qualify for each premium level. Several years ago, companies were hesitant to try this, but as a few did and they were permitted to by state insurance departments, the trickle turned into a flood. Today, especially in non-standard auto, it is extremely difficult for a company to be profitable without tiering, since they would end up with a disproportionate number of insureds with high credit scores and inadequate premiums.
How does credit score use affect “comparative rating?”
Most personal lines agents subscribe to “comparative raters” which are software programs developed by third party vendors that give the agent a quote for all of the companies they represent, and the agent only needs to enter the data from the application one time. Once the agent selects the company, many comparative raters will even print out the final application for the applicant’s signature. Since credit score formulas are proprietary and are not all calculated the same way, comparative raters are no longer effective since without the correct credit score for each company they cannot select the rating tier of the applicant. Agents are now forced to submit a separate online application to each company they believe will be competitive to get the comparative quotes. Naturally, this has significantly increased their workload, and reduced the effectiveness of competitive rating for consumers whose agent may never be aware of the best deal for them.
Doesn’t the use of credit reduce rates for the majority of insureds?
This is the argument that the industry keeps trying to hammer home. The high correlation of credit and risk means, in essence, that they are able to target a smaller universe of policies with a higher percentage of the overall losses than with any other method. By definition then, a greater number of people fall outside the higher rated or non-acceptable classes. The thinking is that if the majority would realize that this is a good deal for them, even at the expense of the few, the industry could generate overwhelming support for the use of credit.
The weakness in the approach is that they are telling people that good credit saves them money at a time when premium levels in general are rising, so that few people in any risk category are seeing actual decreases. It is not as effective an argument to tell people that they would have an even greater increase if they had bad credit, but that is closer to reality.
Who gets the high rates, and is it fair?
Without quantitative studies to show definitively who is hurt most by the use of credit in rating and underwriting, opponents of credit use are raising awarenes of common sense situations that have yet addressed adequately by the industry.
The Missouri Insurance Department estimates that the number of people who do not have sufficient credit history to qualify for a credit score may be as many as 1 in 10! This group looks like it may have a disproportionate representation of segments such as Hispanics, Muslims, farmers, senior citizens, young people, and those who pay cash and do not use credit. When insurance companies get a “no credit”, their response varies from throwing it into the worst credit bucket, to individual review. Common sense tells you that companies are probably not giving this class the benefit of the doubt and rating them as preferred.
People who have financial difficulties are not all deadbeats. Catastrophic events, serious illness with astronomical medical bills, extended joblessness, and any other infrequent but serious setbacks can cause bankruptcy or poor credit on the part of honest and hardworking families. Does the credit score mean that people going through a bad time are more likely to have a claim, or is it more that the credit score process is incapable of separating the perennial deadbeats from the temporarily misfortunate?
The crime of identity theft is increasingly creating havoc for individuals. It takes a huge amount of effort and a number of years to clean up that kind of mess, but in the meantime, should they be surcharged? There are documented stories of people being non-renewed under this circumstance, despite having a “preferred” loss history over multiple years with the company!
The credit scoring matrix normally differs for auto and homeowners. Since the use of credit is to predict “future” losses, an individual could be in a preferred tier for one product and told that they are a non-standard risk for another due to their credit. Why does this make sense?
A study by the nation’s largest mortgage insurer, MGIC Investment Corp., last year concluded that some borrowers with the highest credit scores face much more serious risks of delinquency and foreclosure than borrowers with low scores. The difference is location. They concluded that no matter how high your credit score, you’re more likely to get behind on your payments and lose your home if you happen to live in an area with a weak local economy and declining property values.
While there are studies that claim to prove that credit scores do not have a correlation with income, it defies logic to believe that a lower income family is just as capable of dealing with economic setback without loss of credit as a wealthy one.
The “type” of credit one uses has an impact on the credit score. It is clear that in lower income urban communities, many people prefer cash and check cashing establishments over checking and bank relationships. Credit scores may penalize them for that choice, even if they make their payments on time.
Finally, credit scores change over time, while policies are typically written for 6 or 12 month time periods. Can insureds have their credit reviewed during the policy term? Will they get a discounted rate on renewal if their credit improves?
But I have been with my company for years without claims?
To be sure, there are a handful of companies whose processes and culture value and respect those clients with a long history and excellent loss experience, but it is a fallacy on the part of most insureds that their company will naturally give them significant additional consideration for their loyalty and outstanding record. When a loss free insured of multiple years with a company is non-renewed for credit, the usual response is outrage, which reflects recognition that the trust they had in the institution has been breached, even if it was unfounded. There are few companies whose customer service people can even tell from their system how long an insured has been with them, or what their loss experience has been beyond the current year.
Bottom line is that unless you are with one of a few exceptional companies, your loyalty is relatively worthless in practice, and insurers continue to underestimate the impact of this natural consumer expectation. Trust broken, even if only assumed, is not easily restored. Not by a company, and not by an industry.
What is it that the companies really want?
The companies want a nice and simple cost effective way to segment their pricing and underwriting so that they can offer a price to all customers and know that they will collect an appropriate premium. Credit offers them the best method they have ever seen. Without credit, they must use more external data sources such as Motor Vehicle Reports, prior loss reports, additional driver reports, etc, which are expensive, time consuming, and plagued with inaccuracy. Customers want instant and binding quotes, and insurance companies require instant information that they can rely upon to satisfy that demand.
Also, at this point, insurance companies are actually fighting to preserve the status quo. According to a 2001 Conning & Co. study, 92% of the largest personal auto companies use credit data in underwriting or rating, and more than half have been doing so since 1998.
What do consumer activists really want?
While some of them are truly motivated to do good, many of these organizations make their money through a legal “extortion” process by which they file suit, create a publicity crisis for the company, and then settle out of court, usually including a “consultancy” contract to assist the company in making sure that they are truly reformed and following the spirit and letter of the settlement.
What do the rest of us really want?
We would just like to be communicated with, respected, and treated fairly by insurance companies. Given how companies have used credit recently, it may be asking a lot. Maybe insurance companies could tell us the truth about what they are doing and why, and then, as consumers, we can make informed decisions. Not that it is necessarily going to be easy for companies to level with their clients.
One insurance company tried to be up front about the use of credit on renewals, and explained to insureds what impact their credit score had on their rate. Agents petitioned the company to remove the information from the renewals because angry customers were calling them and complaining that their credit was excellent and they deserved the best rate! They did cease the communication, but not the use of credit.
Another company mentioned credit in passing to its renewal customers while focusing their attention on simplification of its major rating factors, none of which had the near the impact on their premium that applying credit did. This is a prime case of intentional misdirection, and is the type of doublespeak and dishonesty that ultimately breaks trust.
Form a practical standpoint, the use of credit will have to fit into a business process that can be clearly communicated to and accepted by consumers. With the majority of companies exhibiting typical lemming-like behavior as if there is comfort in numbers, one senses that there exists significant opportunity for companies to focus on their total value proposition to their customers, and not just the underwriting. A company concerned with meeting the needs of its business partners and customers while being open and honest about its business practices, will find loyalty and consideration are returned many times over.
Copyright
©2001-2002 freeinsuranceinfo.com.
All rights reserved.