Can you Insure the DNA Dozen?

There’s an interesting piece that’s been floating around some of the newspapers (Telegraph coverage here; The Australian here) over the last few days about the “DNA Dozen” – 11 scientists and one lawyer who’re having their genome published online in order to demystify the process and tame some of the public’s perception.

Fair dos to them.  My hunch is that genetic privacy arguments are often bunk, and anything that helps storm the citadel is fine by me.  What we think of as our genomes aren’t really our genomes at all, so much as a genome that’s overwhelmingly the same for everyone save for a couple of quirks here and there.  The notion of ownership of the genome, and of privacy, is much slipperier once you begin to get your head around that.  If we’re to make sense of genetic privacy at all, we need to do so with a different set of tools from those that we apply to other kinds of information – using conventional notions about the control of information generates some deeply counterintuitive results.

Anyway: I’m taken by a comment from  Helen Wallace of GeneWatch UK:

Your DNA contains very personal information about you, and in the longer term we can’t be certain this won’t be used by insurance companies. I don’t think scientists should be encouraging this.

It’s not clear whether she’s saying that scientists shouldn’t be promoting the use of genetic information by insurance companies – which is OK, because that’s not what they’re doing – or shouldn’t be making their genomes public – which is unexplained: the objection would here seem to be that people shouldn’t make genetic information public because that’d mean it was public.  GASP!

But the insurance claim is interesting in its own right.

Putting my own (increasingly loony-lefty) politics to one side for the moment, I don’t see the problem with insurance companies getting hold of information.  They didn’t enter the market for the benefit of the customers; they did so to make a profit – and there’s nothing self-evidently wrong about that.  They seek profit by offering a price for insurance on an open market.  For people to be able to withhold their genetic information (which is much less useful than Ms Wallace thinks anyway) implies that they could buy insurance without having to tell the seller some important piece of information – and that looks like cheating.

On the other hand, worries about insurance disenfranchisement seem to me to be overblown.  If there’s a free market in insurance, that requires maximal information.  But it also suggests that there’ll be competition, and so companies won’t be able to penalise people with a certain gene unjustly for fear of being priced out of the market.  Premia ought therefore to fall to a level representative of real risk, rather than prejudice.  If some people pay a higher premium because of their genetic inheritance, then that’s not necessarily unjust: they often pay a higher premium for many insurance products because they represent a bigger risk.  At the same time, a market will grow up to serve that particular section of the community, just as there’re car insurance firms that specialise in different demographic groups.

If you want a free market in goods and services, then I suspect you’ll have to bite the bullet in respect of making genetic information public – but, assuming the market functions correctly, then it doesn’t matter much.  The other option is to abandon the free market.  I’m all ears on that front: I’ll carry the banner with you.  But in that case, publicising the genome doesn’t make any difference either.

  • Keith Tayler

    “Information” has always been a complex issue in the insurance and assurance industry. (I notice that “insurance” is often used when “assurance“ is being discussed and visa versa in many bioethics papers). Fortunately it is not, as you suggest, left to the free market to determine the risk and certainty at the proposal and policy stages. Actuaries still do this work with no thought to a “free market”. In the 18th C there was a free market that price annuities but it soon went bust. The mathematics can be easily distorted to create a bubble market, as has happened in the so called sub-prime mortgage market. Of course there are companies that specialise in different groups, but they are not allowing a free market to set their premiums because the market – no matter how much information is available – can do the job of a actuary. Efficiency, technology, marketing, etc. helps to drive down the cost premiums at the edges, but at some point you hit the actuarial mathematics.

    We have been pooling our risk with others for the last three hundred years or so (all very socialist and mutual – the function of a lot insurance/assurance was to benefit their customers), but too much information undermines this collective and leaves a minority with very high premiums or “uninsurable” (if you have Huntington’s it is not a matter of probability anymore).

    Genetic information is particularly problematic because by knowing one persons information you know a lot about their relatives and future generations. A free market cannot operate with entailment; so, even if we do assume that insurance and assurance is governed by the free market, it cannot operate as a free market under these conditions of entailment.

  • Keith Tayler

    Sorry – I have been told that I have been careless yet again and that I wrote “can do the job of the market” instead of “cannot…”

  • Keith Tayler

    Sorry again – I mean “actuary” not “market”. I think I will go back to bed.

  • John O'Malley

    Hate to get technical here but I'm not sure what great information we are dealing with here. Insurance companies deal in risk and all this data does is give you information on risk. So do you know a lot about their relatives and future generations? Probably not much more that you couldn't have gleaned from death certificates from that family.

  • This is what I'd've said, but I'd've done it less articulately. One thing I'd add, though, is that the power of the problem with learning stuff about relatives and descendents seems to draw largely on the assumption that genetic information is ab initio private – and that's precisely the assumption that's under scrutiny here to begin with. So the worry might turn out to be question-begging.

  • Keith Tayler

    Nobody is doubting that the genetic information on individuals will increase and that this will cause the insurance and assurance industry some big problems.

    The situation can be characterised two ways. Firstly, that individuals can have access to their genetic information but not the insurance company. Governments, especially the US, have passed legislation to this effect, but obviously this opens the insurance and assurance companies to individuals with “high genetic risk“ taking out large policies because they do not have to disclose the information. Genetic privacy could well destroy the insurance/assurance industry if companies are forced to insure without disclosure.

    Secondly, the individual and the company have access to the information. In that case individuals with high genetic risk will either have to pay higher premiums or will be refused cover. This again could destroy the industry as it would increasingly move to “low genetic risk”. The traditional concept of accepting a degree of high risk individuals as mutual, friendly and co-operative company “members” would be lost. Of course actuaries have spent much of their time trying to identify high risk individuals (death certificate information saved the annuity), but genetic information could be too much information for individual risk to be pooled, i.e, the pools could become too small or filled with only average to low risk people (at what point does an average/low risk individual say – “I do not need insurance, I can finance my own risk because it is low).

    The number of mutuals, etc. have dropped in the UK (quite a few in the US), but I believe we should try to keep the ethos of the mutual alive, and who knows there is still time to strengthen it. Limiting the amount of cover and number of policies any one individual can have might enable high risk people to become members of a mutual without the other members becoming dissatisfied. I certainly do not think, as Iain suggests, that it is all down to an information maximised free market and let the Devil take the hindmost (or overthrow the capitalist system and impose…….. Not sure what Iain has in mind for us). Nor indeed is the worry about genetic privacy question-begging. Too complicated to discuss here, but it does require more than an all or nothing solution. In brief, privacy and ownership are two sides of the same coin. I am not prepared to give up my genetic information privacy to a market system that takes ownership of genetic information. Get patents and IPRs out of our collective genetic information and I might share the information (limited control for research and some product development is okay, but the process of enclosure that is taking place has echoes of another age).