Personal Responsibility in Health and Health Care: Luck Egalitarianism as a Plausible and Flexible Approach to Health

Introduction

Allocating medical resources in accordance with needs informs much contemporary thinking regarding the provision of health care (Jensen and Petersen 2017). The increased prevalence of noncommunicable diseases has, however, brought to the fore the alternative rationing criterion of personal responsibility (Golan 2010; Leichter 2003; Minkler 1999; Reiser 1985; Schmidt 2009; Sharkey and Gillam 2010). The recent Global Burden of Disease Study illustrates why: half the major causes of death are noncommunicable diseases (Lim et al. 2013; Murray et al. 2014). Other studies indicate that these diseases claim large proportions of health care budgets (Long, Reed, and Lehman 2006; Rasmussen et al. 2004). Allocating health care resources based on personal responsibility is part of a wide range of policies and policy proposals. In Germany, the degree of out-of-pocket payment related to dental care increases for those who have failed to meet a specified frequency for checkups (Schmidt 2008, 209). In West Virginia, families can be disadvantaged by their health insurance company if they miss appointments (Daniels 2011, 267). In the Netherlands, personal responsibility as a rationing criterion is a possibility provided by the letter of the law (Tinghőg, Carlsson, and Lyttkens 2009, 203). In Florida, obese people can be denied treatment qua being obese (Eyal 2013). In the United Kingdom, a number of Clinical Commissioning Groups within the National Health Service (NHS) have proposed to restrict elective surgery for patients who either smoke or are obese (Pillutla, Maslen, and Savulescu 2018). In many countries, people with alcohol-related end-stage liver disease must undergo six months of sobriety before they are allowed on the waiting list for an organ transplantation (Dom et al. 2015). Others suggest that this group should be considered ineligible for transplantation (Sample 2005). These proposals seek, in various ways, to make resource allocation sensitive to personal responsibility. The underlying thought is that those who are responsible for their diminished health have a weaker claim on resources that could have aided others.

Responsibility-sensitive policies are controversial, and an extensive literature has emerged which is critical toward personal responsibility (Andersen et al. 2013; Bærøe and Cappelen 2015; Brown 2013; Denier 2005; Fleck 2011; Friesen 2018; Vincent 2009; Wikler 2002). This article contributes to this discussion by applying luck egalitarianism, a prominent responsibility-sensitive theory of distributive justice, to health and health care. This article first presents luck egalitarianism in general, and as a theory of distributive justice in health. It then describes the luck egalitarian framework employed here. The most notable features of this are that it integrates concerns for distributions other than health, is pluralistic in its concern for other values than distributive justice, and is flexible regarding how to give lower priority to those deemed responsible. Lower priority is used as a shorthand for institutional measures, which treats the claims of some as weaker than the claims of others. These three features set the framework apart from how luck egalitarianism is often discussed in the literature. The framework is then applied to three health-related cases: oral health, allocation of livers for transplants, and travel insurance. For each topic, this article discusses how we can assess responsibility through an exemption scheme inspired by Roemer. This discussion also highlights important factors mitigating responsibility. Then it is evaluated whether we have reasons to believe that those responsible for their health disadvantage will be unjustly disadvantaged in other spheres of life. Finally, this article poses the question of how we can give lower priority and whether doing so conflicts with other important values. Specifically, it is considered whether the proposed institutional measures are too intrusive in the assessment of responsibility and too harsh toward those deemed responsible. Based on this analysis, it is concluded that luck egalitarianism is a much more attractive and flexible theory than much contemporary discussion allows for.

Luck Egalitarianism in Health

In contemporary debates over distributive justice, luck egalitarianism most clearly stresses the importance of personal responsibility. Luck egalitarians consider distributions as just, when they reflect people’s exercises of responsibility and conversely that distributions reflecting luck are unjust (Arneson 1989; Barry 2008; Cohen 1989; Knight 2009, 2013; Lippert-Rasmussen 2015; Stemplowska 2009). Inspired by Lippert-Rasmussen (2001), this article defines the core luck egalitarian claim as follows: distributions are just, if and only if people’s comparative positions reflect nothing but their comparative exercises of responsibly. As luck egalitarians emphasize personal responsibility, it is unsurprising that we can identify both past and present attempts to apply luck egalitarianism to inequalities in health and health care (Albertsen 2015a; Albertsen and Knight 2015; Cappelen and Norheim 2005; Le Grand 2013; Roemer 1993; Segall 2010a, 2012). These attempts have invited a large literature disputing the plausibility of luck egalitarianism in health. Some critiques point to the implausibility of luck egalitarianism in the context of some specific disease or behavior (Feiring 2008; Fleck 2011; Mailly 2005; Nielsen and Andersen 2014). Others claim that luck egalitarianism is in general an unattractive theory of health (Ahola-Launonen 2016, 2018; Bærøe and Cappelen 2015; Buyx and Prainsack 2012; Cavallero 2011; Levy 2018, 2019; Nielsen 2013; Nielsen and Axelsen 2012; Schmidt 2009). Each of these approaches has a distinct weakness. The first branch fails to reflect sufficiently over what lessons the discussion of a specific case provides for luck egalitarianism in health as such. The second branch pays insufficient attention to case-specific details. This article departs from how luck egalitarianism is often discussed in the context of health by proposing the following approach.

When evaluating a health disadvantage, we should ask the following ¹ :

The first question reflects the luck egalitarian commitment to the moral importance of personal responsibility. The second question follows from an integrationist commitment to take disadvantages in other spheres into account, and the third reflects a commitment to pluralism. If the answer to the first question is affirmative, and those to the two following are negative, we have luck egalitarian reasons to introduce institutional measures assigning lower priority to the person. How we are to assign lower priority is a separate question. The framework is compatible with a wide range of measures to provide lower priority. This framework departs in important ways from contemporary debates over luck egalitarianism in health. Specifically, it moves beyond three narrowing sentiments, which are unfortunately all too frequent in contemporary debates over luck egalitarianism in health.

Three Cases

The approach presented above will be applied and discussed across three very different cases: oral health, distribution of livers among potential recipients, and health-related travel insurance. The cases are selected because they differ regarding the severity involved and whether they are inside or outside the health care system. Allocating livers for organ transplants is a rationing decision inside the health care system, which involves very severe outcomes for those who do not receive a liver. Oral health serves as a case with less severe consequences located inside the health care system. ⁷ To limit the extent of the discussion, this case will be considered by assessing caries and the surgical removal of wisdom teeth. The third topic is travel insurance. When people fly for leisure, scuba-dive, ski, or scale high mountains, they sometimes find themselves in need of medical care and specialized transportation back home. Travel insurance is discussed because it deals with severe cases outside of the health care system.

While space precludes giving a detailed account of the various laws regulating these areas across countries, some initial observations might be helpful to understand why these cases are fruitful for the discussion of responsibility-sensitive policies. Although clearly related to an important part of our health, oral health services is often institutionally located outside the traditional health care system. This means that even in more or less universal welfare states, costs associated with oral health treatment is financed through out-of-pocket payment and private insurance (Holst, Sheiham, and Petersen 2001). This approach to oral health places a large burden on those who are disadvantaged through bad oral health, irrespectively of their personal responsibility. Regarding allocation of livers, the case is particularly interesting because a large proportion of those who need a new liver have such a need because of past alcohol consumption (Stell, McAlister, and Thorburn 2004). Alcohol-related liver disease contributes significantly to the global burden of disease (Rehm, Samokhvalov, and Shield 2013). Most countries treat transplant candidates equally, irrespective of how the need arose. In assessing this, it is important to note that we do not have efficiency-based grounds for preferring one group over the other (Burra and Lucey 2005, 496; Trzepacz and DiMartini 2011, 222). This means that if screened properly, we cannot say that a person with a past of alcohol consumption will fare worse than a person without such a past. Regarding travel insurance, the most common institutional arrangement is that travelers cover or insure against all costs acquired while traveling, regardless of whether they are responsible for the disadvantage, which gives rise to these costs. ⁸ The following sections addresses these three cases employing the already presented framework. It reflects over the best institutional measure to assess responsibility and eliminate the influence of luck on people’s comparative positions. It then reflects over how to implement lower priority and discusses whether such an arrangement is vulnerable to the objections regarding harshness and shameful revelations.

Assessing Responsibility

As a theory of distributive justice, luck egalitarianism is compatible with a wide range of views on what responsibility is and what it takes to be responsible for an outcome or an action (Knight 2006). Many luck egalitarians remain neutral about what the best theory of responsibility is (Arneson 1989, 68; Cohen 1989, 934; 1993, 28; Temkin 2011, 57). Some luck egalitarians have, however, offered specific views on how we should understand responsibility. Segall (2010a, 19–24) suggests that we should hold people responsible for the outcomes of actions we could reasonably have expected them to avoid. Specifically, this means that we cannot reasonably expect people to avoid “actions that agents generally have a vital interest in exercising, and/ or actions that society has a vital interest in having exercised” (Segall 2012, 330). ⁹ Roemer (1993, 147) pursues the luck egalitarian ideal that

society should indemnify people against poor outcomes that are the consequences of causes that are beyond their control, but not against outcomes that are the consequences of causes that are within their control, and therefore for which they are personally responsible.

Roemer suggests what he terms a political rather than metaphysical view on responsibility. To assess responsibility, society decides what it deems people to be responsible for in a given context. Building on this, people are divided into types understood as “a subset of the population, having approximately the same value on the factors” (Roemer 1993, 150; 1998, 7; 2001, 449; 2003, 261; 2012, 168). Within each type, people are equal in their circumstances understood as the degree to which they are subject to influences beyond their control. They differ, however, in their effort: their “autonomously chosen action—within the individual’s control—which, if expended in greater amounts, will increase the degree to which the individual acquires the objective” (Roemer 2003, 261). In assessing responsibility, Roemer (1998, 11) suggests that we compare the expended effort of those within each type. When comparing between types, we should compare the degree to which people differ from the median of their respective type (Roemer 1993, 151; 1998, 18; 2012, 168). According to Roemer (1993, 149), we have achieved our egalitarian goal if and when “all those who exercised a comparable degree of responsibility are equal, regardless of their circumstances.” Roemer’s approach to responsibility has been extensively discussed in the literature. ¹⁰ A particular important concern is that countries vary in what they would consider it reasonable to hold people responsible for. For this reason, we may want to evoke a different standard for sorting people into types. The discussions of responsibility, which follows, are inspired by Roemer’s work, but does not profess to pinpoint how people in actual countries would deem other people responsible for. Instead, it follows the guiding principle of listing as circumstances, what most theories of responsibility would not consider people responsible for. Using this approach, let us see how an approach employing Roemer’s conceptual framework of effort, type, and circumstances can help us assess responsibility in the cases at hand. ¹¹

Consider first oral health as illustrated through wisdom teeth and caries. ¹² In terms of assessing responsibility, wisdom teeth illuminate that some specific disadvantages can as a whole be considered something which people are not responsible for. There is no need for an individual assessment nor for dividing people into types, as nothing indicates that the complexities and pain related to wisdom teeth can be avoided through individual effort. Caries, however, arises in a complex mix of behavior and bad luck. The prevalence of caries relates to oral hygiene and food consumption (Moynihan and Petersen 2004; Sambunjak et al. 2011). However, natural factors influence prevalence and vulnerability. This includes Sjögren’s disease (Mathews, Kurien, and Scofield 2008), diabetes (Sandberg et al. 2000), and cancers treated with chemotherapy (Michelet 2012). Socioeconomic factors also heavily influence the prevalence of caries (Chankanka et al. 2011). In general, the literature demonstrates the prevalence of social determinants in oral health (Petersen and Kwan 2011; Watt 2012).

In the liver case, we must also consider the social and natural factors affecting alcohol consumption and differential vulnerability. There is considerable evidence that social factors such as alcohol abuse in the family count as a risk factor (Rhee et al. 2003). Furthermore, poor social circumstances, unemployment, and poverty are prevalent characteristics among those who abuse alcohol (Allen et al. 2017; Hoy 2017). Natural or genetic factors affect vulnerability to alcohol and the individual propensity to become addicted (Buscemi and Turchi 2011). Consider finally travel insurance. Here the picture is somewhat different. In assessing responsibility, it is important to note that people succumb to disease for many reasons while traveling (World Health Organization 2012). As we isolate this to those who travel for fun and leisure, we are unlikely to find the same patterns of social inequality and genetic vulnerability as the two areas just discussed. If what we assess is responsibility for the decision to travel (and to a lesser extent the responsibility to undertake risky activity while traveling), we have a quite different pattern than in the other areas. ¹³ The social profile of the travel insurance case highlights the source of an important objection to this way of assessing responsibility. If we deem people to be less responsible when their actions match the actions of those who are like them, could well-off people cite the fact that many well-off people travel as a reason to consider them less responsible for doing so? This is arguably not an implication of Roemer’s approach. The reason for this is that there would have to be a certain strength to the effects of circumstance or social class on the choice to travel without insurance. One would suspect perhaps that there is a difference between the distributions of the decision to travel and the choice of travelling uninsured. Furthermore, and more importantly, there is arguably a difference between traveling uninsured when you can hardly afford the insurance and traveling uninsured when you can, which blurs this comparison.

This concern aside the main obstacle for implementing a system based on Roemer’s distinction between circumstance and effort is when sometimes we cannot observe past effort. This is true for caries in the context of oral health. We cannot know peoples’ past behavior in terms of oral hygiene and consumption of sugary food. Thus, we cannot directly assess their effort and compare that effort with others in the same type. For the transplant case, it is in one sense easier, as we can readily distinguish between those with an alcohol-related need and those who need a new liver for different reasons. However, we cannot directly asses the past drinking behavior of those in the former group. Travel insurance provides an example of a case, where we know people’s past behavior because we can assess whether they decided to travel and whether they took out insurance for potential diseases and costs associated with their vacation.

When we cannot assess people’s past behaviors, we are faced with considerable difficulty. If we choose to consider everyone with caries and alcohol-related end-stage liver disease as responsible for their health disadvantage, we would not pay sufficient attention to the already presented factors potentially undermining responsibility. In light of this, the most workable system would be to create a system of exemptions. Based on the epidemiological facts regarding what affects vulnerability and behavior, we can for each of the relevant health disadvantages create a list of exemptions. Within such a system, people with a given health disadvantage can be considered not responsible for this if they can cite one or more responsibility undermining factor. There is a discussion to be had about what it would take for something to get on the list of exemptions. Although a clear set of criteria cannot be presented here, Roemer’s approach provides us with the thought that the list could consist of a number of different categories with different weight. If a person fits some of the criteria on this list, we can deem them not responsible and would have luck egalitarian grounds for not giving the person lower priority. On the contrary, luck egalitarian compensation is required. For those who are deemed responsible, the next step would be an evaluation of whether any disadvantages in other spheres of life offset this.

Offsetting Unjust Inequalities

Following from the second, integrationist, element in the framework, the next step evaluates whether, among those deemed responsible for their health disadvantage, there are some who are unjustly worse off in others spheres of life. The focus will especially be on socioeconomic factors, which may, even if they do not remove responsibility for the health disadvantage, count against giving lower priority. For those deemed responsible, and not unjustly worse off in other spheres, we would have luck egalitarian grounds for introducing lower priority. This section sets out ways to develop luck egalitarian policies, which ensures that the extent to which people’s comparative position reflect luck is neutralized or minimized. In doing so, it becomes clear that the second element in the framework involves a number of important commitments.

The first commitment is to assess whether those deemed responsible for their health disadvantage are in other spheres, worse off in other ways, which are unjust. Prime candidates to include as possible offsetting factors would be low income, little education, or unemployment. Even when we cannot demonstrate that such factors undermine the responsibility for the person in questions’ health disadvantage, the presence of such bad luck in the person’s life makes it likely that the best luck egalitarian policy is one, which does not give lower priority to them.

However, refraining from giving lower priority for adverse health outcomes cannot exhaust the luck egalitarian response to the identified factors. A commitment to neutralize the effect of luck on people’s comparative positions must go beyond this. This second commitment is equally important and has broader ramifications. If we identify various ways in which individuals or social groups are worse off than others through the factors just mentioned, we must actively address these factors. We have reasons pertaining to luck egalitarian distributive justice to do so. The luck egalitarian policy, properly understood, would also be committed to neutralize and eliminate such factors adversely affecting people’s position. This is true independently of their adverse effect on people’s responsibility for health. In terms of travel insurance, perhaps the most important element here would be to counter-balance lack of knowledge and misconceptions regarding travel insurance. It would not be good, from a luck egalitarian perspective, if people traveled uninsured under the false belief that they were already covered by their national insurance or national health care service. For the two other cases, equalizing access to education, jobs, and resources are likely candidates for luck egalitarian policies.

The third commitment following from the luck-neutralizing is to examine whether there are aspects inside the health care system, which disproportionally lets people’s social circumstance affect people’s position. This is clearly relevant in the organ allocation case. Factors influencing the allocation process should not reflect differential luck. In a recent discussion of this, Rosoff (2017, 178) points out that seemingly clinical and well-justified criteria for organ allocation can deepen existing inequalities, by benefiting those of a specific class or race. Furthermore, some criteria employed in the system can themselves be a matter of luck. Thus, we could very well have luck egalitarian grounds for opposing the significant role played by geography in the current U.S. system (Organ Procurement and Transplantation Network 2012).

Summarizing this section, and the one which preceded it, we may describe the role of luck as a filtering mechanism affecting both decisions on whether to give lower priority and broader societal policies. Social factors people are not responsible for provides reasons to not give lower priority, both when they mitigate responsibility for the health disadvantage and when they unjustly disadvantage people. Some factors, such as poverty, are both unjust for how they affect health and for how they affect people’s overall advantage. Furthermore, the luck egalitarian commitment to mitigate the influence from luck provides us with reasons to eliminate these factors and/or mitigate their impact on people’s comparative positions.

Lower Priority in Light of Nonjustice-Based Considerations

The above two sections fleshed out how we can assess whether people are responsible and raised the issue of how we should in an integrationist fashion identify factors which unjustly disadvantaged them. These factors can both provide us with a reason for not giving them lower priority and furthermore give us luck egalitarian reasons to counteract and neutralize such unjust disadvantage. The above framework maintains that for those where the process above did not provide us with reasons not to give them lower priority, we would have a pro tanto reason to do so. Whether we, in the end, give lower priority or not depends on whether there are reasons beyond distributive justice to refrain from doing so. In the pluralistic spirit of the framework employed here, this section addresses this. It considers whether we can introduce measures of lower priority without being too harsh or too intrusive. This discussion over permissible ways for introducing lower priority is conducted by looking at each of the three areas under analysis. It is important to stress that decisions of priority setting should be made through overall principles ensuring that equal cases are treated alike.

Consider first oral health. Denying treatment is not the only option. It is not even a very plausible one for the pluralist luck egalitarian approach. Why should we maintain that those responsible for their bad oral health should go untreated? Why not introduce a system that lets them decide for themselves, whether they want to be treated at their own expense. A system of co-payment has two additional benefits. It provides another instrument to mitigate social inequalities in health by introducing some form of public reimbursement for some social-disadvantaged groups, and it allows for partial responsibility, where those who can cite only a few or minor mitigating circumstances pay a fraction of the full amount paid by those without such.

For liver allocation, it has been suggested that luck egalitarianism would recommend giving lower priority to all patients with alcohol-related end-stage liver disease (Albertsen 2016a; Arneson 2004, 20; Knight 2013, 159; Segall 2007a, 177; 2010a, 29). The system of exemptions stresses the need for an individual assessment, something not always acknowledged in the literature. What kind of lower priority should be introduced in the liver allocation case? There is, of course, no need to outright deny treatment, but the solution of co-payments makes little sense here. ¹⁴ The shortage of organs is the major obstacle for performing transplants. The most plausible solution is that those deemed responsible receive lower priority in the allocation of livers. In doing so, the pluralist luck egalitarian framework would introduce responsibility alongside other criteria such as urgency and capacity to benefit. How should we conduct such weighting of values? Lexical ranking with responsibility as the first principle is unsatisfying. This would mean that a person deemed responsible would have to wait until every nonresponsible person is treated, no matter how sick or unable to benefit. On the contrary, introducing responsibility merely as a tiebreaker offers too little. The idea of allowing responsibility a role only when people are equally disadvantaged and equally capable of benefiting would mean that responsibility is relegated to an unimportant afterthought. A third suggestion would seek to correct for both these issues. If we understand capacity to benefit as a threshold, one has to be above to be considered for transplantation and measure urgency of need using the MELD score. This score describes the likelihood that one will die within ninety days, then we can rank each patient on the scale from 1 to 40 which MELD covers (Martin et al. 2007). If those deemed responsible are subtracted two points on this score, we significantly reduce their chance of receiving a transplant but ensures that those who are treated can benefit and that those who receive lower priority are not disproportionally needier than those advantaged by the policy.

Consider then travel insurance. Two different luck egalitarian approaches seem to be worth discussing as responsibility-sensitive alternatives to current schemes. One holds that the public should cover the costs associated with diseases, which an individual is not responsible for, acquired while traveling. In this case, only conditions for which the individual is responsible can be discussed, and for these perhaps they should be asked to shoulder some of the associated cost. From one perspective, this is a luck egalitarian adjustment of current practices, as it filters out diseases for which the individual is not responsible, concerning itself only with those for which the person is responsible. The adjustment adheres to the principle that a disease covered under a luck egalitarian health system at home is also covered abroad. Although there is, from a luck egalitarian perspective, something to be said for such an arrangement, it does allow for problematic cost displacement. Although costs covered under this scheme are, by hypothesis, associated with an illness for which the person is not responsible, they could be considered responsible for the traveling itself. If the decision to travel adds additional costs to the health care system when treatment and transport is covered, there may be luck egalitarian reasons to introduce a scheme in which people who are considered responsible for the decision to travel will be held (partly or fully) responsible for this decision and the costs associated with illness on their travels.

Insights for Policy

Although responsibility-sensitive policies are often firmly associated with deservingness and right-wing ideologies, the above shows that applying a responsibility-sensitive theory of justice to health ends up in a somewhat different place policy wise.

Instead of policies which gives lower priority to specific groups, or which deprioritizes whole groups of patients, luck egalitarianism is much better understood as requiring individual assessments. Whereas the focus is often on holding people responsible and assess responsibility, the above discussion also shows how luck egalitarianisms commitment to mitigate the influence of luck on people’s lives can be very important for policies. It provides a justice-based reason to eliminate social determinants in health, but also for broader policies of redistribution and poverty-mitigation.

The analysis also shows that a luck egalitarian responsible-sensitive approach to oral health would differ from existing practices in most countries. It would include that treatment for some forms of poor oral health, such as those associated with wisdom teeth, should be free of charge. Luck egalitarians should support co-payment, but only for those where an individual assessment concludes that they are indeed responsible for their own poor oral health. In most countries, such an approach would likely increase rather than decrease the public coverage of oral health disadvantage. Whereas for the other areas analyzed, there might be more room for personal responsibility than current policies (though this depends on a number of empirical facts not fully assessed here), the policies are little like those depicted by critiques of luck egalitarianism and should be implemented alongside distributive measures little like those recommended by the political right.

Conclusion

A pluralistic luck egalitarian theory, which focuses on health, is integrationist, and allows for a wide range of ways in which we can hold people responsible, proves a much different beast, than the current literature allows for. Rather than a harsh, intrusive, and generally implausible theory in the context of health and health care, it is a flexible theory. A theory readily able to address important social and natural circumstances differentially affecting people’s health. And which, properly construed, is much less vulnerable prominent critiques regarding harshness and shameful revelations. Furthermore, the anti luck sentiment of luck egalitarianism must be taken serious and its wide ramifications for redistributive policies acknowledged.