• Ethics

Sunstein and Thaler's book Nudge: Improving Decisions About Health, Wealth and Happiness,1 described a fictional species, Econs. Econs are perfectly rational humans. They optimise their pensions, they read the small print, and they never procrastinate. They are the citizens governments dream of.

But real-life citizens are more Muggles than Econs. We get loans from our own bank instead of shopping around, and our pension-plans, ignored, gather dust. Financial incentives for beneficial behaviours have been suggested as one of a range of “nudge” strategies that push Muggles to behave more like Econs-trying to find, but not cross, the line that separates laziness from choice. This issue looks at a wide range of ethical issues surrounding the use of such incentives in healthcare.

Financial incentives are potentially powerful ways of improving a range of health outcomes. But they risk forming part of an over-simplified nudge narrative. For this reason, I think this special issue is timely and important.

The nudge narrative has its own fictional characters. Take Gandalf. Gandalf is the perfectly rational policy-maker. He has a flawless understanding of both the evidence—which is itself clear and uncontested of course—and of human psychology. He creates an incentive structure which will meet the goal without unintended effects, and yet he never lets outside influences, such as financial interests and ideological agendas, affect his policy-making.

Unfortunately, real-life policy-makers are Muggles too. In 2001, a tax regime aimed at lowering carbon dioxide emissions created a financial incentive for customers to buy Diesel cars. It was effective. We are now suffering enormous damage to human health from the resulting build-up of nitrogen oxide and soot. A 1993 report did contain the relevant information about non-carbon pollutants in Diesel, but was ignored.

In health, evidence is often complex, confusing, and difficult to come by, and reactions to incentives can be unpredictable (e.g. low-fat food campaigns successfully changed behaviour, but appears to have increased rather than reduced obesity). Policy-makers are also hampered by competing demands and interests, and financial incentives of their own.

When the US government tells its citizens what is healthiest to eat, they have first consulted with lobbyists who speak up for the various industries involved in promoting that food, and whose interests are primarily financial.

And when the British Government had to address the risk posed by BSE to human health, it had to balance its duties to both the consumer and to industry. The official report concluded that whenever officials “perceived the possibility of a significant risk to human health” they “gave this precedence over consideration of the interests of the livestock industry”.2 Yet what counts as “significant” when it comes to risking contracting such an awful and deadly disease? Gandalf would know, but Muggles do not.

This might have implications in how we approach incentives, such as, as Gheaus suggests (see page 139), ruling out financial penalties for bad behaviour and focussing on rewards instead.

Respect for autonomy is perhaps the most common ethical concern discussed in the context of financial rewards, and it is discussed by Dawson,3 Healy (see page 146), Wild (see page 139), and Krubiner (see page 168).

Autonomy is a cornerstone of bioethics. But we should be careful not to create a third fictional nudge narrative actor, Mill-lite. Mill-lite is perfectly autonomous, but his autonomy is, as Wilde famously described ignorance, “like a delicate exotic fruit; touch it and the bloom is gone”. Dawson critiques the ‘Nuffield Ladder’ which ranks interventions in increasing order of their interference with autonomy on a number of fronts. But there is one more problem with the ladder. As Griffiths and West explain:

“by placing the provision of information at one … the Nuffield Ladder implies that such measures must come at a price to liberty….[but] The provision of information should instead be represented as an intervention that enhances autonomy.”4

Griffiths and West caution against over-simplifying Mill's harm principle both in separating liberty–costing, from liberty-enhancing interventions, and in considering to whom the harm principles apply.

In this issue, Won et al (see page 183) analyse the ethics of trials involving financial incentives for adolescents using drugs and alcohol to accept long acting contraception. Won et al recognise that financial incentives may enhance autonomy by “increasing the number of options available” but caution that to “minimise risks of coercion, even the highest incentive value in the trial should be small”.

But it might also enhance autonomy by enhancing their motivation towards something they already autonomously desire5 (see discussion in Krubiner (see page 168)). For example, only a minority of adolescents actively desire pregnancy (see (see page 184)). Most might prefer to avoid it but a variety of factors interfere, one of which may be limited motivation. If I promise myself a treat if I complete a boring task, I have autonomously chosen to complete the boring task. The incentive is enhancing my motivation to complete the task I have chosen. The more I value the incentive, the greater the enhancement, but my autonomous choice remains. That isn't to say that coercion is not a risk that Won et al are right to identify, but it is (as they recognise) more complex than a sliding scale of dollars and pounds. Dawson criticises the Nuffield Ladder for focussing too much on the cost of interventions to liberty and not enough on their interaction with other values. But we might also ask whether well-targeted incentives actually preserve freedom: Ulysses was made more free by his chains.

Finally, broader society is no Utopia. In 1834, the New Poor Law aimed to incentivise work by offering the able bodied poor support only through workhouses. This is ‘credited’ with creating the deserving and undeserving poor,6 a false dichotomy which persists today. Wild, Grill and Voigt discuss the risk that incentivising “healthy” behaviours create a stigma and categorise the sick into the deserving and the undeserving.

Given that many of the problems health incentives seek to address are strongly linked with poverty, there is a risk that incentives are in fact functioning by alleviating the underlying poverty problem. Krubiner reports a trial of cash incentives for contraception use. These qualitative studies found, “the money helped alleviate economic pressures to engage in transactional sex” (see page 170). Are we still facing the same old problem from 1834, and before: the belief that poverty is a fault in an individual's motivation, not a societal problem.