Consistent with previous studies, respondents in the present study rejected the orthodox economic approach to costs. In the first stage of the study only six participants agreed that among patients who are equally ill those who can be helped at a low cost should have priority, with the majority finding this 'discriminatory' and 'unfair'. After discussion this reaction lessened and support for option 1 increased to nine, with many feeling torn between a commitment to fairness and the need for efficiency. In the second stage of the study, however, the number of participants supporting option 1 decreased to just three, with the majority favouring some form of compromise (trade-off). Two explanations for this change in the second stage are (i) the emphasis placed on exactly how many patients will not be treated on each option, and (ii) the greater range of alternatives on offer (increasing from three to eight).
Four of the eight strategies offered in stage 2 were governed by a clear principle – 'all the money for disease 2', 'the same amount of money for both diseases', 'the same number of patients' and 'first come, first served'. It is worth noting that these four options attracted less support than the 'no-name' option 3. Although the numbers are small, graphically they represent a hump-shaped curve with a single peak at option 3, suggesting it was not the laudable-sounding labels attached to some strategies that attracted respondents (see Figure ). Rather, most situated themselves somewhere between the 'efficiency principle' (option 1) – 'efficiency' understood here as health maximisation – and the 'equity principles' (option 4 and above).
Figure 1 Number of participants favouring the eight options for allocating a hospital budget (N = 40). Option 1 = All money for low cost patients. Option 4 = Same money for both diseases. Option 5 = Same number of patients. Option 8 = First come, first served. (more ...)
(1) Participants' understanding of the task
It is worth noting that participants appeared to have no difficulty grasping the concept of economic cost. For economists, money per se is not the issue, it is the benefit forgone. In the first stage, for example, those who expressed a preference for the most cost-effective option did so on the ground that this would create the opportunity of 'doing more good' or because of a desire to 'help as many people as you can'. Similarly, participants discussed how treating low-cost patients at the expense of higher-cost patients would mean that 'at the end of the day you are treating more'. In the second stage, opportunity costs were actually quantified and the patients who would fail to be treated on each option were clearly shown. Again, those who chose the cost-effective option did so on the ground, for example, that it is 'fairer to treat the most people possible'. Although the term 'opportunity cost' might not have been used, it was the main topic of conversation, and was clearly understood by participants.
It might be thought that respondents must have failed to grasp the notion of opportunity cost or they would not have rejected the cost-effective alternative. That is, if the patients are alike in all respects, the only explanation for participants choosing to treat fewer rather than more patients is that they (are irrational or) failed to fully comprehend the implications of their choice; that is, they failed to understand the notion of opportunity cost. However, there are other explanations for respondents rejecting the cost-effective option, and these clearly came to the fore in the group discussions; for example, 'fairness', a desire not to 'abandon' people, a desire to give everyone a 'chance'. Judging by the reasons participants actually gave, it was not a failure to appreciate the concept of opportunity cost that led them to the reject the cost-effective option, but a willingness to trade-off health gains to secure these other goods.
Related to this, the fact that participants raised issues about health benefits does not mean they were off-target. For example, some respondents questioned not the issue of cost per se but rather the maximand; that is, some seemed to think the rule of rescue should apply (do not abandon small groups of high-cost patients). Meanwhile, others seemed to be arguing for some weighting of probabilities of outcomes (which might reflect a desire for using regret theory or prospect theory rather than expected utility theory), and many were interested in equity concerns and were thus questioning the distribution of health benefits rather than costs per se. However, the study was concerned with exploring the reasons for placing more or less emphasis on costs. It was a qualitative study that sought to know why (as previous studies had discovered) the public places less emphasis on costs than economists would deem appropriate; that is, what other considerations participants consider important, and for which they would be willing to trade-off overall health gains. Hence, it was entirely appropriate for participants to bring up issues such as the fairness of different distributions, health maximisation as an objective, the importance of not removing hope, the rule of rescue, and so on. It would be wrong to think that because participants raised these reasons for downplaying the importance of costs, that they were questioning something other than the relevance of cost.
Finally, it should be noted that the perspective adopted in the present study did not change between stage 1 and stage 2. In both stages participants were asked to give their personal
views about a social
issue. That is, the researchers were in effect saying to participants: "We want your personal
view about how this social
judgement – this judgement affecting different groups of hypothetical patients (with different costs associated with their treatment) – should be made". There was no change from stage 1 to stage 2 in this regard. In the second stage participants were asked to imagine they were members of a hospital committee responsible for allocating a budget. But this does not signal a change of perspective; for example, from the personal to the social. In both stages participants were asked to adopt what Nord calls a 'caring-for others perspective' [[14
] pp. 8–12] – to make judgements about other people (i.e. hypothetical patients) – but to give their own personal view.
(2) The preservation of hope
One of the main focuses of the study was on uncovering the reasons for participants' choices. In the second stage a number of participants expressed the view that it was essential not to remove the "chance" (DG 5) of being treated or to "discriminate completely" (DG 3) against Disease 1 patients. Some had strong feelings that life should not be seen as "easily expendable" (DG 1). In contrast to those who selected option 1, these respondents were concerned with ensuring that high-cost patients are not left in a 'hopeless' position. This is consistent with the results of two studies of organ transplantation undertaken by Ubel and Loewenstein [15
]. The authors conducted a hypothetical 'organ allocation exercise' and found that participants wanted to allocate some organs to those with the worst prospects of survival. Participants explained their decision with comments such as "everyone deserves a chance" and "needy people deserve transplants, whatever their chance of survival" [[15
] p. 1052] (see also [16
]). To deny a person with a serious illness any hope of a cure, even when the probability is low, is to add an extra dimension of anguish to their remaining life over and above the suffering caused by the illness itself. A number of participants in the present study were prepared to sacrifice health gains to prevent this.
(3) The 'essential contestability' of fairness
Throughout the discussions we saw participants grappling with the concept of fairness itself: in these cases it was not always a conflict between fairness and efficiency, but between fairness as treating the greatest number of people, versus treating the same number of people from both groups, versus giving the same amount of money to both groups. All of these, including the efficiency option, were seen as being fair in their different ways, with none having ethics unequivocally on its side. It is central to the concept of fairness that it involves treating people with "equal concern and respect" [[19
] pp. 272–273], but different interpretations of what this means in practice are possible. In this sense fairness is an "essentially contestable" concept [20
]; that is, while there is sufficient agreement on its basic meaning to enable discussion and debate, there is disagreement about the application of this concept in concrete cases – to actions, policies, institutions etc. This was evident in the group discussions reported here, where a range of incompatible options were justified on grounds of fairness with equal conviction. Interestingly, only the most 'arbitrary' option – first come, first served – was not explicitly defended on grounds of fairness.
(4) Reluctance to trade
It was obvious that some participants found the conflict between the need to allocate resources efficiently and the ethical imperative to treat people equally very confronting. One response to this was to shift the focus onto other factors, such as the age of the patients, their smoking behaviour, or their different potentials for improvement. This was despite being asked to consider the patients 'equally ill' and 'alike in all other respects'. The moderators responded in this situation by asking participants which strategy they would adopt if
there was the same number of smokers in each group, if
there was the same distribution of ages, the same prospects for recovery, and so on. For some participants this had the effect of clarifying the task and bringing the discussion back to the issue of cost. For others the resistance remained. The moderators did not force these participants, recognising that refusing to accept the terms of the exercise, or rejecting the assumptions on which it is based (e.g. limited resources) are just alternative responses to the difficult choices participants were required to make [22
]. However, while different participants exhibited different levels of resistance, the majority eventually came to the conclusion that there is a need to take some account of costs.
(5) The importance of deliberation
Another central aim of the study was to afford participants the opportunity to deliberate and discuss the issue of treatment costs, and, in this way, to question a crucial assumption underlying the current practice of "implicit rationing" [24
]; that is, minimise cost per unit of benefit. The course of each discussion group was similar – at the beginning, some people rejected the relevance of costs altogether, and others looked for ways of avoiding the need to make hard decisions. After discussion, however, most participants came to see this as "ideal" or "utopian" (DG 3) and began to recognise the need to take some account of cost – the need to be "objective" (DG 2) and face "reality" (DG 3). We are therefore inclined to agree with Payne, Bettman et al. [25
], who argue that preference elicitation is best viewed "as architecture (building a set of values) rather than as archaeology (uncovering existing values)" [[25
], p. 244].
Because participants had the opportunity to explore their own views and to hear alternative views expressed and defended, it is hard to explain the results of the present study as due to a lack of reflection, at least in comparison with other studies. In particular, the discussion group methodology allowed participants to engage with the task in a way that aids comprehension [26
], that allowed them time to consider all of the alternatives carefully [31
], that afforded them the opportunity to seek clarification of the task [32
], and to construct considered views rather than simply self-report pre-existing preferences [[33
], p. 47].
Elaborating on the final point, the deliberative approach has the potential to move participants towards 'strong evaluation' rather than 'weak evaluation'. Weak evaluation presupposes no more than that the subject be "a simple weigher of alternatives" [[34
], p. 23]. In particular, it requires no more than the expression of personal preferences, and does not require subjects to go beyond the "self-interest" perspective [[14
], pp. 43–47]. By contrast, strong evaluation presupposes that subjects have the ability not just to reflect upon alternatives, but also to reflect critically upon their own preferences, and to assess them as selfish, intolerant, generous, biased, and so on. Strong evaluation encourages subjects to adopt an 'other-regarding' or social perspective:
Groups and individuals thus enter the allocation decision process, not as simple weighters expressing irreducible preferences, but as strong evaluators, capable of recognizing the challenge that certain preferences make to our preconceptions of what health, human nature and the human community are or should be [[35
], p. 251].
This does not eliminate the possibility that subjects might be confused about their own values or the matters of fact on which they are based. Nor does it eliminate the possibility of ethically questionable preferences being expressed. No procedure for resolving complex social problems can guarantee that. However, if techniques can be developed that facilitate strong evaluation rather than weak evaluation it lessens the likelihood of this happening, and provides a more secure foundation for public participation.
(6) Comparison with the earlier Australian study
The present study differs from the earlier Australian study by Nord, Richardson et al. [6
] in several ways. For example, although the earlier study involved a budget allocation exercise, those who had previously chosen to allocate to the cheapest patients or who had voted for a 'first come, first served' strategy were excluded. This meant that 47% of subjects did not complete the budget allocation exercise. Also, only five strategies were available in the earlier study, with the result that spending the same amount of money on both illnesses was not an explicit option, although this was the second most popular strategy in the present study.
Perhaps most importantly, the present study involved discussion groups, with an emphasis on reflection and deliberation; respondents chose their preferred method for allocating the hospital budget, for example, after in-depth discussion of the pros and cons of the various alternatives. In their paper, Nord, Richardson et al. [6
] looked at a number of different reasons for downplaying the importance of costs (e.g. respondents may have anticipated their own emotional response if they were seriously ill, or may have felt a duty to those with a serious condition [6
]), but they are the researchers' reasons rather than their subjects'.
Despite these differences, the same tendency for people to disregard costs in prioritising health care that was observed in the earlier study was also evident in the present study. In the decade that preceded 1995, Australia had implemented Medicare, a Commonwealth-funded health insurance scheme, providing all citizens free and universal access to health services. However, from 1995 to the current period the nation experienced a different social and political environment where government support for Medicare lessened, a greater emphasis on subsidised private health insurance emerged, and fiscal conservatism gradually became a bipartisan feature of political party health policy announcements [36
]. Despite the differing social milieu, however, substantially the same attitude of Australians towards their public health system arose in the present study, along with further detail about the reasons for this.