The World As I See It

At current pace of development in genome sequencing, it is conceivable that in the next decade or so, affordable technology will enable nearly everyone to learn about his/her susceptibility to certain diseases due to hereditary genes.  This is certainly a breakthrough in the medical field, where treatments can be tailored to individual’s genetic make-up.  It maximises the desired effect while minimising the side-effects.  But its impact on the industry of healthcare insurance would be less promising.

To keep premiums low, it is important to have risk-pooling at a large extent, which is the way Singapore is heading.  At the extreme there is UK’s National Health Service, where everyone is a compulsory participant.  I am not endorsing NHS in Singapore, as the cost of bureaucratic meddling could easily offset the intended benefit (low premium).

But the rationale for insurance is uncertainty and risk-pooling to begin with.  Each of us pays a small amount (or relatively small amount) in expectation that insurers will pay a big sum on the occurrence of an unlikely (or unwanted) event.  Knowledge in the link between genes and diseases greatly reduces the extent of imperfect information, and hence the extent of uncertainty, resulting in a smaller range of risk to pool - only those at the higher end.

People who have less to worry are willing to pay less in premiums, or not to buy at all.  Only those who have reason to expect a claim will buy, and are willing to pay more.  “Bad risk” people self-select into the customer pool, and premiums have to rise to accommodate higher frequency and severity of claims.  As such, “good risk” people face even less incentive to buy at higher premiums, where their money go to pay for others’ claims. Consequently premiums have to rise again and again, until the purpose of insurance is defeated. Adverse selection due to near perfect information leads to market failure in insurance, leaving customers worse off.

Whether to have genetic privacy law looks difficult.  The law is to avoid genetic discrimination by the insurers. If there is this law, we have insider information on our genes and our susceptibility to associated diseases, but our insurers don’t have. It can lead to market failure as described above. If there is no such law, insurance does not function well too. The cost can be prohibitively high for people with certain known risks, hitting those not well-to-do especially hard.

Such is the difficult impact of genome research on healthcare insurance.  As for now, I don’t favour a genetic privacy law, because the common practice of checking family history is equivalent to what genetic information is trying to reveal.

A more thorough analysis on the topic can be found in the latest issue of The Economist.