SPECIALS
Genetic Insurance

By James Carlisle
April 26, 2001

Health Secretary Alan Milburn has turned the spotlight back onto genetics with a speech at the International Centre for Life in Newcastle. In it, he confirmed that the Government will put a moratorium on the use of genetic tests by insurance firms, if the Human Genetics Commission (HGC) recommends it.

The HGC, for its part, has been investigating the issues surrounding genetic testing and insurance. This has included an information gathering exercise among the top industry and science bods as well as a survey of public attitudes to human genetic information. Only one in nine people in the survey thought that genetic information should be used to set insurance premiums. The feeling seems to be that we're being put upon (again) by insurance companies greedy for profit. The true position, though, is more complicated than this.

What's it all about?

To start at the beginning: genes are the little things in our cells that define how our body builds and operates itself. You can think of them as a blueprint for the body. The genes get passed from parent to child, so that the blueprint for any person is based on the interaction of the blueprints of their parents. The last 10 or 20 years have seen huge steps forward in understanding what they are and how they work. Foolish Doctor David Berger (TMFFoolUK) goes into a bit more detail about this in this Rule Shaker report on Affymetrix (Nasdaq: AFFX).

The relevance to insurance is that by looking at the make-up of particular genes, our life expectancy and the probability of our contracting various illnesses can be assessed much more accurately. So, if this is permitted, then an insurance company could look at your genes and say to themselves "this person is a very high risk" and charge you a very high premium accordingly. Of course, it works both ways. You might be such a low risk that the insurance company will charge you much less for the insurance.

Currently, insurance companies take evidence of hereditary illnesses into account when assessing your risk. So, if you have a family history of a certain illness, any insurance that involves that illness might be more expensive. The reason for these illnesses persisting through families is that a susceptibility to them is carried in the genes. However, the illness may be "dormant" (in other words you might pass it to your children but you won't get it yourself) or "active" (you might pass it on and you've a good chance of getting it yourself). The insurance industry therefore claims, quite reasonably, that for some people with family illnesses, genetic testing can confirm that, in them, the gene is dormant and that their premiums will therefore be lower.

An example

Let's take an imaginary illness called endowmentitis. We'll assume, for the sake of argument, that the average cost of contracting endowmentitis is about £20,000 and that the chances in any particular year of any person over the age of 18 contracting the disease is about 1%. Now, if we ignore the insurance company's profit, the correct price for an adult specifically insuring against contracting endowmentitis would be £200 per year (that is, 1% of £20,000). So, if the insurance company insures 100 adults for one year, then it will receive £20,000 in premiums. However, on average one of its customers will contract the disease and the insurance company will have to pay out £20,000 for the claim.

Now, let's say we discover that 50% of this population carries a particular gene which we shall call the "Foolish gene" and the chance of anyone with the Foolish gene contracting endowmentitis in any particular year is much less than average, at only 0.1%. If we take a million people, then we would expect 500,000 of these to have the Foolish gene and, of these, 500 (that is, 0.1% of 500,000) will contract the disease.

Overall, from a typical population of a million people (half of whom have the Foolish gene and half of whom do not), we still expect 10,000 people to get ill (that is, 1% of a million). So the non-Foolish 500,000 people contribute the other 9,500 cases (the total of 10,000 cases minus the 500 Foolish ones).

So, the average annual claim for those people with a Foolish gene is a mere £20 (that is, 0.1% of £20,000). On the other hand, the average annual claim for those unfortunate souls born without the Foolish gene will be £380. Now, if the insurance company is able to test people to see whether or not they are Foolish, then they would be able to charge Foolish people premiums of only £20 per year, whereas everyone else would get charged £380.

This presents a problem. The Fools would say that they should pay lower premiums, because, otherwise, they are effectively subsidising the non-Fools for their tendency towards endowmentitis. The non-Fools don't like it, they produce the (rather spurious) argument that they're the ones that are most likely to get endowmentitis and therefore they most need the insurance. They therefore say that it is wrong that it is more expensive for them to obtain the necessary insurance. All this is complicated further by the fact that, until a test is done, no-one knows whether or not they are Foolish. As a result, no-one knows which camp they should be in until it's too late and they find themselves in it.

Why do the insurance companies want to test us?

Well, that's a more difficult question than it seems. The answer is probably that they think they want it but, if they thought about it some more, they might change their minds. Insurance companies have a very definite mind set about getting more and more information about probabilities. They are desperate for information. This is not surprising since this is what enables them to compete effectively against all the other insurance companies. However, individual bits of information that are available to the whole insurance industry don't help anyone.

In the example about endowmentitis, it actually makes no difference to the insurance industry whether they have the information about the Foolish gene or not. If a million people are insured (of which half are Foolish and half are not), then the insurance industry is raising £10m (500,000 times £20) in annual premiums from the Fools and £190m (500,000 times £380) from the non-Fools, making a total of £200m. The total annual claims are also £200m (10,000 claims, 500 Foolish and 9,500 not, at £20,000 each). So, the balance of premiums to claims is achieved just as it was before the information was available.

In fact, theoretically it's even worse for the insurance industry. The purpose of insurance is to control risk. It reduces uncertainty. The less uncertainty there is, then the less need there is for insurance. If a gene told you that a person had no chance whatsoever of contracting a certain illness, then there would be no point in them having insurance against it. Similarly, if a person's genes told you that they were virtually certain of contracting an illness, there would also be no point in them insuring against it, because the premiums would have to match the expected cost of contracting it. There is no point paying £20,000 to be insured against something that is certain to cost you £20,000. So, the more certain everyone is about life, the less business there will be for insurance companies.

A better reason for wanting genetic testing arises out of a concern that people will get the test done themselves and then buy insurance depending on whether they need it. You might, for instance, get yourself tested for the Foolish gene and then only buy insurance (at £200 per year) if you didn't have the gene. If you had the gene, you would take your chances (that is, "self-insure") because you'd know that, for you, the premium of £200 per year is too high. If everybody did this then, of a million people taking insurance, 19,000 would get endowmentitis at a total cost of £380m, while the premiums would only amount to £200m. You can see why the industry might be worried about this.

Of course, the insurance company could always ask you whether or not you've had a test and say that the insurance is invalid if you tell a fib. Since the skills and equipment needed to do genetic testing are in pretty tight supply, it might even be possible to keep a register of people who have had the test. That way people can decide for themselves whether to "see their cards" or "play blind". The trouble with this is that people could go overseas and get the tests done in a way that could not be traced.

Private or Public Sector Insurance?

So, if we don't allow insurance companies to use genetic tests, we must either find a way of preventing secretive testing or the insurance companies will have to price in the effect of some people having the test, taking the insurance and not letting on. This would effectively put the premiums up slightly for everybody. If we do allow genetic testing, then the Government (and therefore taxes) will have to pick up the tab for those that have the wrong genes and can't therefore afford the additional premiums. Either way, it is going to end up as the general public footing the bill, whether through slightly higher premiums or higher taxes.

It therefore becomes a question of private insurance versus welfare. The biggest difference between these is that it is easier for private insurance to offer a choice of the level of cover. It enables people to tailor their insurance needs to their existing standard of living. Benefits from welfare, however, do not tend to be flexible in this way. Unless that is, there is compulsory insurance as a percentage of salary and which provides benefits based on contributions. This is, of course, the idea behind National Insurance. Perhaps the solution to this problem therefore lies in an improved version of the National Insurance scheme. One way or another, though, it will take a lot of sorting out.

More: The Fool's Insurance Centre