Out of favor
in 2007, massachusetts Insurance Commissioner Nonnie S. Burnes announced the end of the auto insurance regulatory regime that had been in place for decades. Burnes proclaimed that the time had come for the state’s “fix-and-establish” method of auto insurance regulation to be replaced by “managed competition.” And she predicted consumers would see greater choice and lower rates from more competition.
Consumer groups and Massachusetts Attorney General Martha Coakley denounced the decision and warned that consumers would on average see higher auto insurance rates than they would under the fix-and-establish system and that insurers would introduce new and unfair rating factors (the characteristics used to determine a policyholder’s premium). Coakley eventually challenged the rate filings of five insurers, demanding that the commissioner reject the proposed rates. The attorney general hired me to review the filings and testify as to the fairness of the proposed rates. Although the commissioner had kind words to say about me personally, she decided my testimony was not relevant to her review of the filings and refused to let me testify. Too bad, because my testimony went to the heart of the problems with “managed competition.”
For most consumers, auto insurance is a necessary evil. You have to buy it and it costs a lot. Some consumers have a vague idea that state officials have set the rates in the past and that those rates have been coming down in recent years, but that’s often the extent of their knowledge.
In Massachusetts, under the fix-and-establish regime, policy forms and rates were established by the insurance commissioner and, consequently, any insurance company operating in the state was required to use those forms and rates. Insurers were able to offer a few discounts — typically, a reduction of 5 percent to 10 percent for membership in a particular group — in addition to the discounts mandated in the rates set by the commissioner.
When insurers set their own rates, they start by figuring out how much overall premium (income) they need to cover claim costs, expenses, and profit. The result is a statewide base rate. The insurers then decide which rating factors they will use to set how much particular types of consumers will pay in premiums.
Rating factors are the characteristics of the consumer or vehicle that modify the base rate. Some of the rating factors seem obvious and logical. If you drive your car a lot of miles, you run a greater risk of being in an accident. Where you live makes a difference: The frequency of auto accidents is greater in a more densely populated area than in a sparsely populated area. If you have had driving violations or accidents in the past, you are more likely to have an accident in the future. Other rating factors include the type of vehicle — more expensive cars cost more to repair or replace than inexpensive ones — and whether the car has anti-theft devices or the driver has taken safety training.
Using these types of rating factors not only makes sense and seems fair to consumers, but also provides clear incentives for less risky behavior. With these types of rating factors, consumers have some control over their auto insurance premiums.
In states other than Massachusetts, there has been a revolution in the nature of insurance rating factors over the last 10 to 15 years. Insurers have introduced many new rating factors that focus on the socioeconomic characteristics of the consumer, with the result that traditional factors like one’s driving record have lost importance — and factors often outside of the consumer’s control have become the primary determinant of insurance premiums. The new factors include credit scores based on a consumer’s credit history, educational achievement, occupation, household structure, and prior liability limits.
At this point, your eyes are glazing over and you are thinking, what are these rating factors and why haven’t I heard of them? Well, you haven’t heard of them because they have not been permitted in Massachusetts. In other states, the most important factors for determining auto insurance premiums are credit history and prior liability limits. These two, like others mentioned above, are largely proxies for race and income. It turns out that insurers’ use of consumer credit information discriminates against low-income and minority consumers. The Missouri Department of Insurance conducted a detailed study of insurance scoring and found the single best predictor of a consumer’s insurance score was the consumer’s race.
“Liability limits” refer to the amount of coverage on the policy and can be broken down into “minimum limits,” the least amount of coverage mandated by a state, and “excess limits,” or coverage greater than the minimum. With prior liability limits as a rating factor, a consumer who previously had a minimum limits policy will pay more than a consumer who previously had an excess limits policy — even if both consumers are otherwise identical. It doesn’t take an economist to see that a rating factor using prior liability limits leads to higher rates for low-income consumers.
When Commissioner Burnes decided to let insurers develop their own filings, she set out some ground rules. To her credit, she prohibited insurers from using certain socioeconomic rating factors, including credit scoring, education, occupation and prior liability limits. She required insurers to maintain the existing limits of differences in rates by geographic area, meaning that rates in certain parts of Boston or Worcester should not jump up just because of geographic location. She also said no consumer could experience an initial premium increase of 10 percent or more.
It is important to put the 10 percent cap on premium increases due to new rating factors in perspective. Had the commissioner maintained the fix-and-establish system, and set the rates using the same input values as her predecessors, rates would have gone down about 10 percent. So, by capping the premium increase to 10 percent over current rates, and assuming a 10 percent rate cut if fix and establish had continued, the commissioner was actually allowing premium increases of 22 percent over what the premium charges would have been under the old system.
Attorney General Coakley challenged five of the 19 insurers who made initial filings last year for rates to be effective in April of this year. Massachusetts law, like the laws in every other state, sets out three standards for auto insurance rates. The rates must not be inadequate, must not be excessive, and must not be unfairly discriminatory. Inadequate rates are rates that are so low that the insurer’s financial solvency is threatened. Excessive rates are rates that produce an unreasonably high profit. Finally, the unfair discrimination standard goes to rating factors: Rates are unfairly discriminatory if they result in different premium charges for consumers who pose the same risk of a claim.
When the attorney general challenged the filings of the five insurers, she argued that their rates were excessive. Commissioner Burnes rejected the attorney general’s challenge. This was not surprising to anyone watching the course of events over the past year. The commissioner wants the new system to be successful, which means attracting national insurers to enter the Massachusetts market. For that to happen, she must demonstrate that rate filings will be reviewed quickly, that insurers will be able to earn the profit they want, and that the attorney general will not cause delay. The commissioner was clear that she, and not the attorney general, is the insurance regulator in Massachusetts. In addition, it would be a bit contradictory for the commissioner to declare the auto insurance market competitive and then agree with the attorney general that some filed rates were excessive.
Coakley hired me to review the rate filings for unfair discrimination. But when it came time to testify on the rate filing of the biggest insurer, Commerce Insurance of Webster, the company claimed that the attorney general had not included anything about unfair discrimination in the petition challenging the rates. Burnes refused to let me testify, saying my comments would not help her decide whether the challenged rate filing met the statutory standards. That was unfortunate because the issue of new rating factors is the dominant problem in the move to managed competition.
Had I testified, I would have pointed out that two of the five insurance company filings I reviewed penalized senior drivers and violated the state law requiring drivers 65 years and older to get a minimum 25 percent discount on their insurance premiums. Here’s how. Two insurers, Commerce and Arbella Mutual Insurance of Quincy, introduced a new rating factor: years licensed. The two companies proposed that drivers licensed for 51 years or more would receive a 10 percent surcharge on their premiums. It is pretty clear that anyone licensed 51 years or more is at least 65 years old. By adding the years-licensed surcharge and not increasing the senior discount, Commerce and Arbella are effectively reducing the senior discount to less than 25 percent.
Suppose that the senior consumer’s premium was $1,000 before the 25 percent discount and before the years-licensed surcharge. That senior should be paying no more than $750, or 25 percent less than $1,000. But with the years-licensed rating factor, the senior driver is now paying $825, or 10 percent more than the maximum the senior should have been charged.
In addition, many of the company rate filings suffered from two other problems. First, for many or all of the new rating factors, the insurers provided no data or other demonstration that the rating factor had any relationship to expected claims, or that the amount of the surcharge or discount for the rating factor was reasonable. As I reviewed the filings, it became clear to me that insurers were picking rating factors and values for those factors because they wanted to market to a particular clientele.
The problem with this approach is that the favored consumers get discounts that must be paid for with higher premiums by disfavored consumers, who are predominantly low-income consumers. The result is a back-door approach to using socioeconomic status as the main basis for setting insurance premiums.
With the new rating factors, consumers who insure multiple vehicles and buy additional policies from the same insurer might see big rate cuts from both current rates and rates that would have been in effect under the fix-and-establish system, while consumers insuring only a single auto might see a 10 percent premium increase over current rates and a 22 percent increase over rates they would have paid under the fix and establish system.
The biggest problem with the new rate filings is the incredible complexity of many new rating factors. To take one example, consider the “rating group” factor in Travelers’ Premier Insurance Co. filing. The factor is an amalgam of a dozen other factors, including single versus multiple vehicles insured, type of vehicle, vehicle age and type of coverages purchased, among others. And the combination of all these rating factors produces a value which is translated into yet another factor used to determine the premium.
There is clearly no need for this monster combination factor — each rating characteristic could be its own rating factor, which would provide some transparency to the consumer as to which factor caused the premium to increase. Rather than clarifying the impact of rating factors on the development of a consumer’s premium, the rating group approach obscures the role of rating factors and allows individual rating factors to have far more weight in determining premium than is otherwise indicated by actuarial analysis.
At best, this approach is unnecessarily complicated and renders the impact of various rating factors opaque to consumers. The consumer can never find out what characteristic was the cause for a premium increase. This defeats the goal of promoting competition by weakening the understanding of rating factors, and, subsequently, the market position, of consumers relative to insurers. The multiple uses of the same risk characteristic also lead to unfairly discriminatory rates because that characteristic is given too much weight in the rating process.Under managed competition, an auto insurer might raise or lower overall rates by 5 percent, but, because of new or changed rating factors, could increase some consumers’ rates by multiples of the overall rate change. With the first round of rate filings, the disparity between the rates paid by favored and disfavored consumers grew. Next year that disparity is likely to grow even more because managed competition has opened the door to abusive auto insurance rating schemes.
Birny Birnbaum is the executive director of the Center for Economic Justice in Austin, Texas. A former Texas insurance regulator who oversaw rate filings, Birnbaum was hired by Attorney General Martha Coakley to analyze the rate filings of five companies in Massachusetts as part of the move to “managed competition.”