Financial Incentives, Coercion and Psychosis

In a recent editorial in the British Medical Journal, Tim Kendall draws attention to a recent study that suggests that modest financial incentives can significantly improve adherence in people treated with depot drugs for schizophrenia and other psychoses in the UK. This study looks set to reignite the debate regarding the moral permissibility of offering financial incentives as a part of medical care. Whilst those who support this practice point out that we already offer non-financial rewards to many patients, others have criticised the practice as, among other things, amounting to coercion. In this post, I shall contest this particular objection to the practice of offering financial incentives to patients as part of medical care.

Prior to addressing this particular argument, let me sketch the contours of the problem in a little more detail. As Kendall points out in his editorial, whilst non-adherence to treatment programmes is a problem across the “whole of medicine”,[i] it is particularly difficult to achieve adherence to anti-psychotic drugs, despite the fact that adherence to such drugs not only seems to directly improve treatment outcomes, but also seems to enhance the efficacy of other forms of treatment of psychotic illness (such as cognitive behavioural therapy) .Two reasons why it is particularly difficult to achieve adherence in this case are that many patients suffering from psychotic disorders do not accept that they are ill, and many anti-psychotic drugs are associated with serious unpleasant side effects.

Whilst acknowledging the need to guard against the risks that offering financial incentives in this case might pose to the doctor-patient relationship, and stressing the need to research other ways to improve adherence, Kendall suggests that offering financial incentives to patients suffering from schizophrenia in order to achieve adherence may be morally permissible. There are various lines of objection that critics have raised against this view. Perhaps most salient amongst these objections is the claim that offering financial incentives to adhere to a treatment plan is coercive, and thus invalidates the patient’s consent. This sort of objection is apparent in the comments of Joanne Shaw, the former vice chair of NHS Direct, on the guidelines published by the National Institute for Health and Care (formerly Clinical) Excellence in England and Wales in 2007. These guidelines recommended that physicians provide financial incentives in order to achieve adherence to treatment programmes. In response, Shaw wrote:


Paying people to take medicines sends a signal that they need to be compensated for doing something that is not inherently in their own interests. It is coercion by carrot rather than stick, but coercion none the less.  [ii]


However, is it really true to say that this is an instance of coercion? I shall suggest that it is not. To begin making this argument, it is illustrative to consider a paramount example of coercion: Imagine a patient who is suffering from a broken shoulder, and who refuses to consent to undergoing what their doctor believes is a necessary surgery, because the doctor has told him that the surgery has a very small risk of complications that might lead to paralysis of the affected arm. Here, the doctor could coerce their patient to undergo the surgery by threatening them with a very bad consequence if they refuse to consent; for instance, the physician might threaten to refuse the patient any further pain relief until he agrees to consent to the operation.

Part of what seems problematic about the physician’s behaviour here is that they seem to undermine the patient’s autonomy. The coercion is understood to invalidate the patient’s consent, because the physician has subjugated the patient’s will to their own by unduly reducing their patient’s options in order to achieve the outcome that the physician believes is best; the patient can no longer choose to maintain the status quo option of continuing to receive pain relief and not undergoing surgery. The physician’s conditional threat makes them believe that these two events are now mutually exclusive.

This, I suggest, is an example of a coercive threat, and it is uncontroversial to suppose that such coercive conditional threats can, and normally do, invalidate the consent that the patient gives as a result. In contrast to this though, Shaw (and others who support her arguments) claims that offers can be just as coercive as threats; presumably, this is what she means when she says that offering patients financial incentives is “coercion by carrot rather than stick, but coercion none the less”.

However it is somewhat problematic to claim that offers can be coercive in the same way as threats. Although threats and offers share many structural similarities, they are crucially different in one respect. Whilst credible conditional threats take away the victim’s belief that they are free to maintain their status quo situation, conditional offers will normally leave the status quo option intact; that is, it is normally possible for the recipient of an offer to maintain their status quo situation simply by refusing the offer. Accordingly, whilst threats serve to subtract a desirable option from the victim’s choice set, namely the status quo option, offers will normally serve to supplement this option with a further one. This difference seems to have implications for the autonomy of the recipient; if we assume that threats take away a valuable option, whilst offers add a valuable option, there seems to be grounds for claiming that the former will undermine the recipient’s autonomy whilst the latter will not.

Of course, there are some exceptions to this. Some offers may change the status quo situation; this is a common line of objection to allowing people to have the option of voluntary euthanasia. Moreover, sometimes offers can form a part of a coercive situation; if I were to illicitly reduce your freedoms so that your status quo situation following my intervention was now a poor one, and then made you an offer that gave you an alternative to maintaining your new status quo situation, this would, I suggest be coercive. For example, if Jones stole Smith’s wallet, and then offered to return it if Smith performed some degrading task, this would be an example of an offer constituting part of a coercive situation. Here though, the offer per se is not coercive; rather it is the illict reduction of freedoms that preceded the offer, and which were put in place in order to make the offer attractive, that makes the situation a coercive one.

With this in mind, why does Shaw claim that offering financial incentives in order to achieve adherence is coercive? In the passage quoted above, she seems to be suggesting that such offers are coercive because an implication of the offer is that the thing that recipients of the offer are being given an incentive to do is not inherently in their own interests. Even if we grant to Shaw the contentious point that adherence to a treatment programme may not be in the interests of some of those suffering from psychotic illness (perhaps on the basis that anti-psychotic drugs have several unpleasant side-effects), this does not entail that such an offer is coercive; this is because it is still possible for patients to reject the financial incentive, and to continue with non-adherence; their status quo situation is still intact. The financial incentive just makes one of the options more appealing than it was.

Of course, the fact that the offer is not coercive does not entail that it is a moral one; for example, a non-coercive offer might still be exploitative if it seeks to capitalise on an extant relationship of inequality. For example, suppose that I stumbled across a man dying of thirst in the desert. Suppose that I have a backpack full of water bottles, and offer him one bottle for £1’000’000. Here, I have given the man an incentive (i.e. a bottle of water) to do something not inherently in his own interests (giving me £1’000’000 of his money); and for Shaw, this means that the offer is coercive. But this seems implausible; it seems ridiculous to claim that the man could not autonomously consent to agreeing to this transaction because he had been ‘coerced’ by my offer. I have undoubtedly treated him unfairly, and I may be an appropriate subject of all sorts of moral insult; however, although it might be fair to say that I have exploited the man in this example, it does not seem that I have coerced him, and thereby rendered his consent invalid. This is an important distinction, since exploitation does not undermine the validity of consent in the same way as coercion. In order to remedy the wrong that coercion does, I would need to (at least) reinstate the freedom that I had taken away from my victim, in order to reinstate their initial choice set; however, in order to remedy the wrong of exploitation, I would instead need to remedy the pre-existing inequality between myself and the exploited party, or give them a far fairer deal by giving them a better incentive, or demanding less for my proposed incentive.

To conclude then, whatever may be wrong in offering those suffering from psychotic illnesses financial incentives to adhere to their treatment programme, it is not coercive. If opponents of this practice are to object to it adequately, then it must be on the grounds that the offer is somehow exploitative, or that the practice leads to other bad consequences (such as the deterioration of the patient/doctor relationship) that outweigh the benefits of adherence.


[i] Kendall, “Paying patients with psychosis to improve adherence”, BMJ 2013;347:f5782


[ii]Shaw, “ Is it acceptable for people to be paid to adhere to medication? No”, BMJ 2007;335:233



  • Facebook
  • Twitter
  • Reddit

Comments are closed.


Subscribe Via Email