Social Pricing in Automobile Insurance

The trend toward greater social pricing in automobile insurance in Canadian jurisdictions could eventually destroy all remaining privately supplied markets, forcing governments to pick up the pieces.

As used in this article “social pricing” is defined as government-mandated subsidization of the premiums of higher risk drivers through the over-charging of lower risk drivers, by deeming as socially inappropriate rating factors used by insurers to segment the higher risk and lower risk driver groups. In forcing social pricing, legislators are motivated, in some cases by misguided egalitarianism, but more commonly by outright political opportunism.

Age, gender and marital status are the rating factors most commonly being outlawed in Canadian provincial and territorial jurisdictions. However, there is a growing list of other rating factors, including credit and payment histories, years licensed and rating based on all drivers in the household, which are being banned as well.

Could this trend actually be a case of appropriate market segmentation being labeled as unfair discrimination? Market segmentation is certainly a form of discrimination. Properly done, market segmentation provides many direct and indirect benefits to automobile policyholders and our society.

Exactly why is the source of these benefits being eliminated by provincial legislators? Somewhere along the line “discrimination” and “unfair discrimination” have come to mean the same thing in automobile insurance and are deemed to be socially evil in all cases.

Discrimination on the basis of age, gender and marital status in matters of, for example, employment has been deemed to be inappropriate for some time and rightly so. Why you may ask doesn’t this same logic apply to insurance? In employment, age, gender and marital status have been found not to be correlated to job performance and hence it is unfair for employers to use these factors in deciding whether or not to offer employment to a particular candidate. Employers do discriminate among job applicants on the basis of education, experience and even personal grooming, among other factors, because these factors do affect job performance.

Unlike in matters of employment, age, gender and marital status can be shown to be statistically significant predictors of future loss results in automobile insurance. To charge different premiums for groups segmented by age, gender and marital status is both economically justifiable and fair in a market economy. For example, in life and health insurance, there are clear statistical relationships between age and gender as rating factors and the occurrence of mortality and morbidity. It is obvious that a market where the 80 year old and the 20 year old would each be expected to pay the same premium rate for life insurance could not function for long, if at all.

We know that young male drivers who are not married form an identifiable higher risk group of potential insureds. As a group, they will have more accidents and as a result higher losses than the same sized group of more mature drivers. Shouldn’t these young male drivers as a group pay their own losses, even though (thankfully) most young drivers will not personally be involved in a loss themselves?

Insurance is a prospective business. That is, we collect premiums today for losses that will occur in the future. We could have a system that charges one and the same premium for all drivers. Few would consider this approach fair for the ‘pleasure use only’ driver or the one with the older, less valuable vehicle. Total fairness would dictate that each driver be charged a unique premium based on precisely the risk he or she presents to the automobile insurance system. Although very desirable, we know that unique personal premiums are not feasible.

Operating in the real world between these extremes, insurers attempt to segment applicants into statistically meaningful groups, whose members prospectively present a very similar level of risk. We do not know in advance which members of each segment will be involved in or cause accidents. Consequently, we collect an equal premium from each member of the segment and then let events unfold as they may during the course of the policy term. At each renewal, each insured’s status is reviewed and if warranted moved to a different rating segment, based on the most current information. That person’s premium could go up or down depending on what rating factors have changed during the last policy term.

The main benefit of effective market segmentation is fairness. We will all pay a premium closer to our theoretically correct personal premium, the more finely insurers are allowed to segment the market. What is needed is more experimentation in creative market segmentation by insurers, not less.

The proper pricing of applicants based on risk appropriately rewards lower risk drivers and gives high risk drivers an incentive to improve their driving record or other rating factors, to lower their prospective risk. At the extreme, high premiums may force the highest risk drivers off the roads, thereby reducing accidents, injuries and loss of life.

Insufficient market segmentation leads to unintended consequences. The higher risk driver saves a bundle on his insurance, which allows him to buy that powerful car he has always wanted; a bad accident looking for a place to happen! Lower risk drivers are not sufficiently rewarded for their good habits and can be expected to worry less about avoiding losses in the future. Supply problems arise, as insurers seek to attract the over-paying lower risk drivers, in preference to the under-paying higher risk drivers. Governments try, with mixed results, to legislate away the supply problems through various forms of “take-all-comers” rules.

In all cases, we can expect the total losses to the insurance system to increase, which eventually will cause everyone’s premiums to increase. Tragically, in auto insurance, higher losses also mean more injuries and loss of life.

In a free market system, as insurers find effective ways to segment the market and give lower premiums to deserving low risk drivers, the premiums of higher risk drivers increase to keep the system in balance. A problem occurs when the highest risk drivers hit the “affordability wall”. Higher risk drivers then raise a ruckus, claiming that regardless of their level of risk, they are entitled to affordable automobile insurance. Are they? Of all groups in society needing a financial “helping hand”, I would have thought that higher risk drivers would be very low on the priority list. If society does want to assist high risk drivers with the cost of their car insurance, help should be provided through the tax system, where a “claw back” of the benefit can be applied to wealthy high risk drivers.

Why are our politicians so anxious to introduce and expand social pricing? It is very much a case of “the squeaky wheel gets the grease”. Under Alberta’s grid system for example, the 10% of drivers that are high risk are now paying thousands of dollars less than they should be, while low risk drivers (90% of all drivers) are in many cases paying hundreds of dollars more premium than they should be. Politicians have silenced a complaining (but undeserving) minority, while hoping that the silent majority does not notice this inequity, which is very much to their financial detriment.

Provincial and territorial governments need to decide whether they want to introduce (or continue) a government-run automobile insurance scheme, in their jurisdiction. If they don’t want to take over the business (or wish to return it to the private sector), then they must allow the privately supplied market to experiment with and develop various forms of market segmentation. Only rating criteria which truly create unfair discrimination should be banned and not those which effectively segment higher and lower risk drivers.