One of the fears associated with genetic technologies is that they may be used to increase the inequalities in our society; paradoxically, the same technologies, applied to genetic testing, might lead society to ensuring equal access to healthcare. How it will all play out in practice is hard to predict, but insurance companies are certainly heading for an unknown future. Society might even benefit.

Currently, the opposite is happening, if anything: in Britain, which does have universal healthcare but also has a parallel system of private medicine for those who prefer to pay, there is serious concern that the new technologies may break the bank of the national system. Some of this is due to concerns about profit-gouging as a result of restrictive patents – a breast-cancer screening that cost the British National Health Service (NHS) about $1000 initially and $50 for each follow-up (at no cost to the patient) was being sold in the US by a private company for $2400 and $500, respectively, by a company that planned to extend its patent to the UK. If similar pricing policies were enforced, the NHS probably could not afford universal breast-cancer screening.

But that is only a test for two genes, BRCA-1 and 2. If, as seems likely, there are a dozen or more genes involved in the predisposition for breast cancer, and if a royalty has to be paid on each test for each gene, the entire operation becomes economically impossible. Let alone the possibility of multiple diagnostic tests for predispositions to numerous other conditions, such as heart disease, that might be avoided or postponed by specific preventative measures, some of which are essentially cost-free in financial terms.

They are not, by the by, cost-free overall. If people like eating red meat, for example, then telling them not to causes them to pay a kind of price – to forego a satisfaction. It may seem trivial to vegetarians, or people who just don't like beef, but it's real enough to those who do. And that leads to a serious problem with public policy attempts to change patterns of behavior when they are based on purely statistical evidence. Some people (and let us assume that a future genetic test would be able to identify them with significant success) really should not eat red meat for fear of heart disease; but most people can eat it with reasonable confidence. If we can test everyone, fine; those at risk can be warned, and will probably follow the advice, thereby helping themselves and the accountants of the insurance industry (or NHS) at the same time. If we can't, then what are the chances of persuading everyone to act as if they are at risk? Pretty low. You only have to look at the consequences of the BSE or "mad cow disease" scare in Britain to see that. Sales did drop, but briefly and not by all that much. Even if we ignore the pressure of the cattle industry – the real world does get complicated – Britons, like Texans and many others, enjoy eating beef. It's going to take more than a vague, population-wide statistical report to change their diet.

This is essentially what has happened with tobacco smoking: Statistically, smokers are at increased risk for lung cancer and other ailments, but most smokers do not die of lung cancer. And they know it. At a population level, predictions can be made; at an individual level, they are much less certain, though there are now some indications that yet another genetic test might improve the accuracy. Admittedly, tobacco is addictive (more for some people than others), and there are people who continue to smoke even after diagnosis of related ailments, but I would venture to guess that, given a strong statistical correlation of smoking and death for the individual, as opposed to the weak one that now exists, most such individuals would, reluctantly, stop. The anti-smoking lobby generally acts as if smoking is only an addiction; it's more than that – it's a pleasure. One that many of us are reluctant to give up.

My father smoked very heavily, reputedly 60 a day, until he was over fifty, then quit, and eventually died at 83, almost exactly the same age as his non-smoking sisters, of a heart attack. My mother smoked, though less, almost to the end of her life, and died of multiple organ failures (some of which may have been associated with lung capacity, I suppose) but not cancer. I reckon my chances are pretty good. I'll probably quit again, I do know that I feel better when I don't smoke too much, and that number has been falling over the years – it used to be 20 a day, it's now 15 or less – and obviously it's not 'good for me' ... except that it did really help when I was in crisis, shortly after my father's death, which coincided with several other stress-producing events in my life. You know the classic stress events – death of a close relative, moving, job loss, end of relationship? I had them all, simultaneously. And I fell off the non-smoking wagon, after seven years. And I got through it. I chose not to drink to excess, not to use anti-depressant prescription drugs, not to use illegal drugs, not to starve, not to get fat; I chose to smoke tobacco. And I'm not sure I was wrong. Even on medical grounds.

Would I have reached for a pack of Camels knowing that I personally had genes predisposing me to be among the minority of smokers who are likely to develop lung cancer? Probably not. In those circumstances, I'd probably have gone for the Prozac or some other prescription medication. Would I have done it if I knew that I did not have those genes? Are you kidding?

There is another problem with genetic testing. Not only do most current tests only predict a statistical probability of developing disease symptoms, they also do not eliminate the possibility of contracting almost any given disease even though the patient is not in the high-risk group. There are exceptions, of course, like Huntingdon's disease, which is now highly predictable in both directions, but most diseases will strike a significant number of people in the low-risk group, as well as a larger number in the high-risk population. And they are more likely to affect people who abuse their bodies. Back to beef: Some people really should not eat it; but no one should eat too much of it. That's just basic dietary advice, and most people follow it. It's human nature, however, that some people are probably going to binge if they think, even incorrectly, that this pleasure is cost-free.

Even so, the potential benefits of testing are real. Cheap, universal genetic tests should lead to better and cheaper healthcare in the long run. Those most at risk for, say, breast cancer should be screened more often than those at less risk (though the low-risk group should also be checked). That's just common sense, an appropriate allocation of resources.

If we manage to get there. The cost question could be a barrier, but let us assume that, in the fairly near future, the tests become cheap enough that cost is not an issue. Under the present American system, it is virtually certain that insurance companies will routinely make coverage conditional on the results of such tests. They already refuse to cover pre-existing conditions; it's not much of a stretch to imagine that a significant probability of a particular disease would be considered a pre-existing condition. After all, the potential for payment is exactly what the insurance companies are trying to compute. You can hardly blame them for turning down an applicant with, say, a 60% chance of incurring an expensive ailment, can you? Makes perfect commercial sense.

But we are not just talking about one disease here. We are, at least theoretically, talking about many different diseases that are statistically predictable but not preventable (fantasies of eliminating 'faulty' genes before birth being certainly further off than testing). So what happens to insurance? One possibility, the one that the industry has effectively started towards, is that no two policies will be the same – every insured person will have a unique policy that exempts the company from paying for the diseases that person is most likely to get. Which basically means that everyone is at least partly self-insured. For the very rich, this might not be a problem, but for the vast majority of people this a scary prospect indeed.

And there is a further complication: If the tests show you are not likely to get any of the major diseases, why waste your money on insurance against them? The answer of course lies in the statistical nature of the tests – you probably should buy insurance – but it wouldn't take many people deciding not to participate to change the pool of subscribers enough that insurance rates needed to go even higher; which in turn would discourage customers and exacerbate the vicious spiral of increasing insurance costs combined with ever-larger numbers of the uninsured. A scary prospect for society at large.

The trouble is that the essence of insurance is the lack of knowledge of what will happen to any given individual; but the essential point of using specific genetic tests on individuals is precisely the opposite.

One political maneuver would be to forbid insurance companies from basing their acceptance of customers on the results of tests; but if you let people have the tests anyway, that seems wildly unfair. It does nothing to discourage the supposedly healthy from opting out, while actively encouraging the potentially sick to sign up. And the tests are a good thing for society at large – they focus medical resources where they are needed – even if they seem to be a bad thing for the individuals who pay higher premiums, or for the companies who lose good customers, from their point of view, and gain bad ones.

No, the only logical answer is universal healthcare. Society needs to ensure that, while the benefits go to the individual patients, the risks are spread around everybody. The system could, I suppose, be administered by private companies, thus giving the present players a role in the new approach, but it has to include everyone. And the main reason this may happen is that significant numbers of the middle class will find themselves either being denied insurance or being faced with the appalling choice (if indeed it is allowed) between getting insurance and getting the most appropriate medical care.

Of course, if greedy, short-sighted and socially irresponsible companies keep the cost of genetic testing so high that most people cannot afford it, perhaps the insurance system will stagger on for another generation. And people will suffer unnecessarily from diseases that could and should have been treated earlier.

We wouldn't let that happen. Would we?