Lotteries are usually only proposed as a way of allocating scarce public resources in special cases, when there seems to be no other fair approach. For example, lotteries may be appropriate when citizens have equally strong needs or claims and there are no decisive reasons to choose one way or another. Or, as Peter Stone argued in his 2011 book “Luck of the Draw”, lotteries may be appropriate if they are the only way to prevent bad reasons from influencing decisions. http://ndpr.nd.edu/news/27864-the-luck-of-the-draw-the-role-of-lotteries...
In such cases, lotteries can help to achieve “horizontal equity” – the equal treatment of equals. In his provocative blog, however, Adam Wagstaff proposes lotteries as a general approach to healthcare resource allocation. Although I have come across this idea before, I have not previously seen it spelled out in such practical detail – and Wagstaff is to be commended for that. The basic idea is problematic, however, since some citizens have greater claims on public healthcare resources than others – for example, greater health care needs or, more controversially, a larger historical contribution to social insurance. Differences in healthcare needs or insurance contributions are often seen as good reasons for differential allocation of healthcare resources. Lotteries therefore violate the principle of “vertical equity” – the appropriately different treatment of people with different needs or claims on resources. Wagstaff’s proposed solution to this problem is to weight lotteries in proportion to people's claims - e.g. in proportion to cost-effectiveness, or need, or contribution, or whatever other source of claims. However, although a "claims-weighted lottery" can in theory deliver vertical equity “ex ante” (i.e. before the lottery is drawn) it will inevitably violate vertical equity “ex post” (i.e. afterwards). That is, lotteries would result in some people with greater needs for healthcare resources – the ones who are unlucky in the lottery – receiving fewer resources than people with lesser needs – the ones who are lucky in the lottery. So lotteries violate the fairness principle that health care should be distributed according to need rather than luck.
There are also powerful political barriers to healthcare lotteries. Policymakers dislike explicit lotteries - there aren't many votes in the campaign pledge "when elected, I will make decisions by tossing a coin". So do clinicians - if they had to implement patient level healthcare lotteries they would have considerable discretion to fiddle them by altering their diagnosis and reporting behaviour, and would face pressure from patients to do so. And patients dislike explicit healthcare lotteries most of all – especially those who are unlucky, and even more especially those who attribute their “bad luck” in the lottery to being too honest, too inarticulate, or too poor to persuade or bribe their doctor to fiddle the lottery.
If explicit lotteries were politically acceptable, one might expect to find them used already in cases of “horizontal equity” when people have equal needs and hence equal claims on resources – for example, in cases of allocating treatment for patients diagnosed with the same disease and needing the same treatment. But they are not. Rather, health systems typically use waiting times and waiting lists to ration scarce resources in this kind of case. There is thus considerable political resistance to use of explicit lotteries even in special cases where they can help to achieve horizontal equity and do not grossly violate the principle of “ex post” vertical equity. Things can of course change, and ideas can gain public acceptability over time. So this is perhaps an idea worth pursuing.... though more details are needed not only on the practical details but also on the philosophical underpinnings and the strategy for overcoming political resistance.
Richard Cookson, University of York, 31 May 2013