According to a recent report (The Essentials of Numeracy) by the National Numeracy charity, 49% of the British working-age population has the numeracy skills expected of an 11-year-old. (By the way, National Numeracy's love of figures does not extend to putting a date to their report, which, judging from the sparse references is post-2017.) This puts into perspective Rishi Sunak's inane, fatuous and irrelevant recent suggestion of making maths education compulsory to 18 years of age.

Fear not, dear reader: this is not another jeremiad about the appalling state of maths teaching in the UK. Not directly, anyway.

The strict demarcation of disciplines as taught in primary and secondary education creates vast areas of unexplored knowledge and understanding that fall between the cracks. One such is the understanding of risk and therefore of insurance. Indeed, in spite of the sterling efforts of Sir David Spiegelhalter (Professor of the Public Understanding of Risk at the University of Cambridge), risk and insurance are poorly understood by the general public, media and even by acclaimed critics of economics.

Readers of an earlier piece (The Economics of Organ Donation, 2 November 2022) may recall that if a misguided and plainly wrong critique of economics is required there is no better supplier than Michael Sandel. And, indeed, on the issue of the relationship between risk, insurance and morality, the 'Public Philosopher' does not disappoint. 

For argument's sake, I shall assume that Sandel is a car driver and that like most of us he has a preference for smoothing his consumption profile: faced with the possibility of a large loss, he has decided to trade off the certain loss of current income (his insurance premium) for the uncertain prospect of future bad luck (incurring a costly accident). By so doing, he has transferred risk to his insurance provider, but, according to his own theory (expounded in his bestselling What Money Can't Buy) that market transactions change the nature of the good or service being traded, he has also diminished his own moral stature. Now that he is no longer financially responsible for any damage caused by his driving, he may drive less carefully and more often, take more hazardous routes, etc. He has become a moral hazard for the insurer.

The solution is not an intensive course in communitarian ethics, but some simple economics whereby his insurance contract is designed to minimise his incentive to take less care. Contrary to Sandel's mantra, markets crowding in morality, one might say. The same argument applies to life insurance, where the policy typically excludes suicide and hazardous activities. Notice that it would be inconsistent to regard the risk‐averse behaviour of the insured as virtuous and prudential while considering the bearing of risk by the insurer as irresponsible gambling.

Suppose now that, because of a change in personal circumstances, the insurance calculus is altered and the individual concerned prefers a smaller lump-sum paid now to the larger death benefit paid in the future (minus the stream of premium payments). Provided the insurance market is competitive and therefore the bargaining power of the two parties is unchanged, the mere change in the timing of the accrual of benefits would be morally neutral. But this revised insurance contract is a viatical, which attracts the full force of Sandel's opprobrium ('ghoulish business', 'creepy', 'corrosive effect', etc). Sandel's 'arguments' against viaticals (the insurance contract whereby the insured sells his/her life policy in exchange for an immediate lump‐sum) are worth examining, not for their power but for their weakness.

Sandel starts by admitting that 'the viatical business serves a clear social good – financing the final days of people with terminal illnesses'. Typically, he does not address the question why terminally ill people are put in the condition of having to sell off their assets, but instead states that viaticals 'are wagers on death that give investors a rooting interest in the prompt passing of the people whose policies they buy'. But instead of addressing the issue directly, he refers to a hypothetical straw man ('it might be replied') who raises an irrelevant side issue: 'viaticals are not the only investments that amount to a death bet', thereby offering Sandel the opportunity to state the obvious, namely that 'with life insurance, the company that sells me a policy is betting for me, not against me. The longer I live, the more money it makes. With viaticals, the financial interest is reversed [...] the sooner I die the better'.

Viaticals do pose a problem, a problem of moral hazard, which, as it turns out, has a very simple solution. Just as purchasing motor insurance changes the insured person's incentives to drive safely, similarly, buying someone's life insurance creates in theory an incentive to hasten death and hence cash the policy as soon as possible.

Apart from the fact that this 'rooting interest' is quite difficult (not to say illegal) to turn into practice, the ensuing moral hazard problem is easily solved by making the policies anonymous: I do not need to know the identity of the investor who buys my policy, and he/she does not need to know mine (only my medical history, etc).

Sandel concedes that a viatical 'is merely creepy, not morally objectionable', only to raise a new objection: 'perhaps the moral problem lies not in any tangible harm to me but in the corrosive effect on the character of the investor. Would you want to make a living betting that certain people will die sooner rather than later?'

Sandel appears to be oblivious to the fact that 'the investor' most likely to buy the viatical is the insurance company currently providing life insurance to the terminally ill, i.e., the economic institution least likely to be 'corroded' by actuarial calculations involving death.

The change of terms whereby a life policy is turned into a viatical involves another important factor, of which Sandel seems to be blissfully ignorant. A diagnosis of imminent death is, to state the obvious, a crucial piece of new information and we would expect the parties directly affected (the insured and the insurer) to adjust their behaviour, by applying the very same principles underlying the buying and selling of (virtuous) life insurance. The time of death, which at the time of taking up life insurance was uncertain (random) has now become (virtually) certain. Therefore, we would expect a rational agent to smooth his/her income profile and shift income from the future cash‐rich state (death payment) to the current cash‐poor state.

Who is best placed to accommodate this change of policy? The current insurer, of course, whose economic calculus has changed in the opposite direction and who is best placed to supply the immediate lump‐sum and carry the burden of (reduced) risk about the precise time of death. According to Sandel, this is a morally reprehensible 'death bet', whereas it is nothing of the sort. This points to the basic misunderstanding of insurance and gambling at the heart of Sandel's 'argument'.

To quote directly: 'We commonly think of insurance and gambling as different responses to risk. Insurance is a way of mitigating risk, while gambling is a way of courting it. Insurance is about prudence; gambling is about speculation'.

If the 'we' in the above quote refers to people without any understanding of insurance and gambling, then the statement is true by definition. Any second‐year student of mine who wrote the above rubbish would get a fail mark, because it shows the most basic mistake one can make when talking about insurance and gambling. Insurance and gambling are two sides of the same coin: taking up insurance means swapping uncertainty for certainty and therefore providing insurance must mean swapping certainty for uncertainty, i.e. gambling.

Let me see if I can out‐Sandel Sandel and think of an insurance contract even more morally objectionable than a viatical.

What about the following scheme: 10 of you pay me £1,000 each upfront and at the end of each year I will pay out a fixed proportion of the £10,000 capital, say 5%, to be shared equally among all surviving members. By joining this scheme you are effectively betting on the earlier death of your fellow investors.

Not sufficiently morally repulsive? Fine. What about this twist? You still pay me £1,000 upfront, but instead of betting on you outliving your fellow investors, you can select a stranger, presumably a healthier and younger individual, whose life you nominate as being 'yours'. Surely such a repugnant scheme does not exist or if it does it must surely be the brainchild of some immoral lowlife in Wall Street. Nope.

What I have just described is one of the oldest methods whereby governments raise public finance and in fact was one of the main sources of public revenue under King Louis XIV. It is called a 'Tontine' (after the Neapolitan Lorenzo Tonti who sold the idea to the Sun King in 1689) and technically it is a 'life contingent group annuity with benefits determined by survivorship'. Please notice the date (1689), a time when economics did not exist as a word, let alone as a social science.

If What Money Can't Buy had been produced as an economics‐related article and had gone through the rigorous peer review which characterises economics as a discipline, Sandel would have been spared the embarrassment of accusing economics and economists of devising ghoulish death bets and we, the readers, would have been spared his inanities on the ethics of insurance (See The free flow of knowledge, 23 November 2022). But popular bestsellers like What Money Can't Buy or lectures on YouTube are not peer‐reviewed and allow scholars like Sandel to pontificate on subjects they know very little about.

Dr Manfredi La Manna is a Reader in Economics at the University of St Andrews

Return to homepage