Health, Health Care, and Equality of Opportunity: The Rationale for Universal Health Care

Abstract

This article discusses what arguments best support universal health care (UHC), with a focus on Norman Daniels’ equality of opportunity account. This justification for UHC hinges on the assumption of a close relationship between health care and health. But in light of empirical research that suggests that health outcomes are shaped to a large extent by factors other than health care, such as income, education, housing, and working conditions, the question arises to what extent health care is really necessary to protect and promote health, and thereby opportunity. The author argues that, although this challenge to the equality of opportunity rationale is legitimate, it is not sufficiently specified to allow us to adequately assess the extent to which universal health succeeds in protecting equality of opportunity. The article concludes by outlining a more promising strategy for developing a viable rationale for UHC.

Introduction

Universal health care (UHC)—the provision of (basic) health care to everyone, free or at low cost at the point of access—is by many considered an essential component of a just society. While this practice is common in many parts of the world, there is as of yet no consensus on the best principled ethical rationale for provision in this manner.

A principle of equality of opportunity has been proposed as a compelling candidate. This justification for UHC hinges on the assumption that health care makes a very significant contribution to protect and promote health. However, in light of empirical research that suggests that health outcomes are shaped to a large extent by factors other than health care, such as income, education, housing, and working conditions, the question arises to what extent health care is really necessary to protect and promote health, and thereby opportunity.

Critics have used this empirical research as a basis to argue that Norman Daniels’ equality of opportunity rationale fails to justify UHC. In this article, I argue that this line of criticism relies on a set of health measures that are inadequate for an evaluation of the equality of opportunity rationale. It is assumed that the measures of health inequality in question accurately track inequalities in opportunity, but I show that this assumption does not hold. I propose some alternative metrics by which we can better assess the contribution of health to protecting equality of opportunity. Finally, in the last section of the article, I consider an alternative strategy for justifying UHC.

UHC and the Equality of Opportunity Rationale

UHC can be defined as everyone having access to “quality health services that meet their needs without being exposed to financial hardship in paying for them.” ¹ This is a broad definition that leaves it open whether we as a society ought to provide sufficient health care, where everyone has access to a basic minimum, however that is defined, or equal health care, where everyone has access to the same level of coverage and the wealthy are prevented from buying more. Furthermore, it leaves it open what kind of health system—for example, social insurance or public system—might be best suited to achieve this objective.

Nevertheless, a key feature of UHC thus defined is that it entails that health care services are allocated on the basis of need rather than the ability to pay. This mode of allocation gives a distribution that can be characterized as “more equal” than many other goods, and wealth and income in particular. Accordingly, it is considered by some to require a strongly egalitarian or “special” justification. ²

Norman Daniels provides such a justification by accounting for the value of health, and thereby health care, in terms of its importance for opportunity. ³ Health is important for opportunity because impairments to health reduce our functioning and thereby restrict the range of opportunities available to us. Health care, in turn, by protecting and restoring health, is thereby also important for maintaining citizens’ normal opportunity range. This move serves both to explain the oft-intuited “special moral importance” of health care as well as to enable the allocation of health care to be governed by a (modified) Rawlsian Fair Equality of Opportunity Principle (FEOP). ⁴ In the Rawlsian framework, the FEOP is a more stringent or demanding distributive principle than the difference principle (DP), in the sense that it requires equality rather than maximin. ⁵ Thus, the importance of health for opportunity provides a plausible explanation for why health care ought to be distributed “more equally” than wealth and income.

The Social Determinants of Health Challenge to the Equality of Opportunity Rationale

Daniels’ argument is built on the assumption of a close relationship between health care and health. However, as Daniels also recognizes, empirical research suggests that health outcomes are shaped to a large extent by factors other than health care, specifically the “social determinants of health” such as income, education, housing, and working conditions. ⁶ This raises the question to what extent health care is really necessary for maintaining health and in turn, opportunity.

This line of criticism of the equality of opportunity rationale is exemplified by the argument put forward by Gopal Sreenivasan, ⁷ and more recently Shlomi Segall. ⁸ Sreenivasan characterizes the equality of opportunity rationale as a two-step argument: The first step establishes that a fair share of health is necessary to ensure an individual’s fair share of opportunity, and the second that access to health care is necessary to achieve one’s fair share of health. Moreover, Sreenivasan argues, a “fair share” of opportunity must be understood in relative terms, as a share of opportunity that is “more or less equal” to the shares of opportunity that others in one’s society have. Thus, what is at stake is the impact of health care not on health or opportunity as such, but rather its impact on the distribution of health and opportunity. Accordingly, it should be possible to test the extent to which health care contributes to equality of opportunity by assessing how much it contributes to reducing inequalities in health. Sreenivasan and Segall both argue that, because the social determinants of health have a greater impact than health care on the level of health inequality in a population, UHC does not plausibly contribute to maintaining fair equality of opportunity. Sreenivasan substantiates this point with reference to the introduction of the National Health Service in the UK in 1948. Contrary to general expectation, UHC did not succeed in reducing health inequalities between socioeconomic groups, and 40 years after its introduction, these health inequalities had actually widened. ^{9 , 10} Thus, if our objective is to protect citizens’ fair share of health (and thereby opportunity), we would be better off spending our health budget targeting the social determinants of health rather than on UHC. Accordingly, it appears that a principle of equality of opportunity cannot justify UHC.

If it is right that health care makes only a modest contribution to reducing health inequalities, does it thereby follow that its contribution to reducing inequalities in opportunity is similarly modest? I argue that unless one takes a simplistic linear view of the relationship between health and opportunity, it is still possible for health care to be an important contributor to fair equality of opportunity, even if its contribution to reducing health inequalities, as measured by the sheer quantity of health gains that it yields, appears to be less than that of social and material factors. I will show that it is not straightforward to assess the potential importance of health care for the distribution of opportunity merely on the basis of its magnitude of impact on the distribution of health.

Health Inequality and Inequality of Opportunity

The line of criticism outlined in the previous section moves from a premise about health care’s relatively modest contribution to reducing inequalities in health to a conclusion about health care’s relatively modest contribution to reducing inequalities in opportunity. However, claims about a given health intervention’s level of impact on the reduction of health inequalities are based on sets of health data that represent the level of health inequality before and after the intervention. Thus, this move must rely on the further assumption that the relevant health data are a good indicator of people’s level of opportunity.

More specifically, the move presupposes that quantities or shares of health, as represented in the relevant data, reliably “track” quantities or shares of opportunity. Unless quantities of health correspond to quantities of opportunity in this systematic manner, information about the level or magnitude of health (inequality) does not straightforwardly translate into information about the level or magnitude of opportunity (inequality). Consequently, it would not be warranted to draw conclusions about the contribution of health care toward the reduction of inequalities in opportunity on the basis of its contributions toward the reduction of inequalities in health.

In the following, I examine whether it is reasonable to assume that quantities of health or health inequality track quantities of opportunity or opportunity inequality. I argue that this depends on how well the relevant health measure captures those aspects of health that matter for opportunity. I further argue that health measures used by the critics of Daniels are inadequate in this respect because the relationship between health and opportunity is complex in ways that cannot be captured by these measures.

Measuring Health and Health Inequality

Assessing “how much” health citizens have relative to each other, before and after some health intervention, requires that health be quantified or measured. Measuring health, or health inequality, is a complex and difficult task that involves many methodological and normative choices, and a range of different measures exists. ¹¹ Health is a complex and multidimensional concept, which seems to resist being reduced to a single comprehensive definition. Any quantification of health, whether it concerns health states or lifelong health achievements, is necessarily a reductive enterprise. A further complicating factor pertains to the measurement of health inequality—as opposed to mere health—because inequality itself can be quantified in different ways. ¹² Thus, there will often be more than one possible answer to the question of “how much” inequality a given distribution contains.

For many of these measurement issues, what is the “best” approach will ultimately depend on what aspects of health are relevant in a given context, and one’s reasons for measuring health or health inequality. If there is not one, universally best way to quantify and measure health, then there also cannot be one, determinate answer to what a “share” or quantity of health is. It does not follow from the general claim that health is important for opportunity that every possible quantification of health has the same bearing on opportunity. In order to determine whether quantities of health track quantities of opportunity, this needs to be considered with respect to how health is quantified or measured in any given case.

Health and Opportunity

Arguably, not all conceptualizations and measurements of health (or inequalities in health) are equally good trackers of opportunity (or inequalities in opportunity). Life expectancies or mortality rates, the measure used by Sreenivasan in his example of the introduction of the National Health Service in the United Kingdom, are standard measures for population health, and are also frequently used to measure health inequalities. However, the length of lifespan (whether actual or expected) is a fairly crude measure of health. Many forms of ill health do not affect mortality, such as disabilities, injuries, or nonfatal chronic diseases, and therefore, the extent to which a life has been affected by any of these conditions would not be captured by a mortality-based measure. But clearly, these forms of ill health affect opportunity. Two individuals, each with a lifespan of 70 years, could have vastly unequal levels of opportunity, depending on other aspects of their health such as disability and nonfatal morbidity in the course of their lives. Conversely, two individuals having unequal lifespans is consistent with their having equal levels of opportunity. Thus, inequalities in health as measured by the length of lifespan do not appear to track inequalities in opportunity particularly well. ¹³ Given that, on this particular measure of health, unequal shares of health may be consistent with equal shares of opportunity, and vice versa, we simply cannot determine to what extent different health interventions have impacted on the distribution of opportunity only on the basis of their impact on the distribution of health.

Health expectancies are another common measure of population health and health inequality. This type of measure combines data on mortality and morbidity into a single indicator, expressing how many years an individual or group can be expected to live in good health, free from disability and disease. ¹⁴ In capturing not only quantity of life years but also their (health-related) quality, these measures offer richer information and a more accurate picture of people’s overall health. Accordingly, we would expect that such measures provide a more accurate picture of people’s level of health-related opportunity.

However, even more fine-grained measures of this kind could mask inequalities that could potentially be relevant. As a summary or whole-life measure of health, health expectancies express the expected total health achievement of life as a single number. It does not tell us anything about what I will call the health trajectory, and how that total amount of health is distributed across that life.

To see how this could be relevant, consider the following example concerning health inequalities between men and women, as measured by “Healthy Life Years,” one example of a health expectancy measure used, for example, by the EU. Overall, men and women tend to have comparable healthy life expectancies, with women generally doing slightly better than men. ¹⁵ However, the underlying pattern of ill health is more dissimilar. Generally, women live on average longer than men, but spend a greater amount of time in ill health, both in terms of absolute number of years with ill health as well as the ratio of number of years in ill health to number of years in good health. ¹⁶ So while men and women have fairly equal shares of health overall, their respective shares of health nevertheless differ with respect to the relative distribution across quantity and quality of life years.

What are the implications of the different health trajectories of men and women for their respective levels of opportunities? Different answers are possible, depending on how one conceives of the relationship between health and opportunity. One possible view is that there is a fixed, linear relationship between health and opportunity, such that a unit of health, conceived as, for example, one healthy life year, always corresponds to a unit of opportunity, however that is conceived, regardless of the overall distribution of health in terms of quality and quantity of life years. In this view, men and women have (more or less) equal shares of opportunity in our example.

However, the linear view seems too simplistic. First, it seems fair to say that “shares” of opportunity is not just a matter of the amount of opportunity one enjoys, but also the variety and quality of those opportunities, and that being able to enjoy the full range of different kinds of opportunities will be important. An individual who lives to the age of 40 in full health and an individual who lives to the age of 80 but with a significant disability will both enjoy smaller shares of opportunity than someone who lives to the age of 80 in good health. However, a life in good health cut short at 40 is likely to be deficient in opportunity in a different way from a life of normal length but with significant impairments. Different aspects of health—such as functionality and length of life—plausibly matter for opportunity in different ways. Second, health is likely to have different levels of impact on opportunity at different ages or life stages, because different kinds of opportunities will be available or important at different stages of life. ¹⁷ For example, there is a case for giving more weight to certain opportunities in early adulthood to mid-life, such as the opportunity to pursue a career, because one’s career trajectory will strongly influence income, which is another very important determinant of opportunity. Illness and disability that negatively affect one’s employment opportunities are likely to have a negative accumulative effect on overall opportunity levels because of the synergistic effects of health and income. ¹⁸ Accordingly, the same health shortfall (in terms of the “quantity” of ill health it reflects) could have a greater or lesser impact on one’s level of opportunity, depending on the life stage in which the health condition occurs.

If we reject the linear view of the relationship between health and opportunity and accept that this relationship is much more complex in ways that I have briefly sketched above, the implication is that the underlying health trajectory, in addition to the total amount of health one enjoys, can plausibly make a difference to one’s overall level of opportunity. Accordingly, in the above example, men and women could have unequal levels of (health-related) opportunity, even if they have equal shares of health. Consequently, even a more fine-grained measure such as Healthy Life Expectancy is not by itself a sufficiently accurate indicator of a person’s level of opportunity.

The Equality of Opportunity Rationale Revisited

I have argued that it is not necessarily straightforward to assess the (relative) contribution of health care to protecting equality of opportunity on the basis of its (relative) contribution to reducing health inequalities. The preceding discussion shows that we need different, and perhaps more sophisticated measures of health and health inequality to better evaluate the equality of opportunity rationale. Further conceptual and empirical work is needed to provide a more nuanced, specific, and empirically grounded account of the different ways in which health affects opportunity across the life course. (Daniels’ account is fairly brief here and does not go much beyond broader statements about how impairments to normal functioning reduce the range of exercisable opportunities. In my view, further specification would benefit his account.) This will give us a better understanding of the extent to which different kinds of health inequalities are indicative of inequalities in opportunity. This would also be valuable for the purpose of identifying health inequalities that are of greater concern, in being strongly linked to inequalities in opportunity.

Nevertheless, at the current juncture, we can identify some health measures that would be more appropriate for evaluating the equality of opportunity rationale. One possible measure is inequality in child and infant mortality. While inequalities in health as measured by overall life expectancy do not track inequalities in opportunity particularly well, as argued above, extreme inequalities in length of lifespan are an exception. Clearly, a very short life contains much less opportunity than a life of normal length. By reducing inequalities in child and infant mortality, we are reducing an important source of inequality of opportunity.

A different strategy could focus on inequalities in the extent to which health affects employment, rather than look at inequalities in health outcomes directly, given the substantial effects of employment on income. Thus, in order to assess the impact of health care on inequality in opportunity, we could look at inequalities in early exit from the labor market due to poor health or inequalities in (long-term) sick leave, both of which will result in a loss of income. If health care can be shown to contribute to reductions in these inequalities, this would constitute a significant reduction of inequality of opportunity.

A further route by which the introduction of UHC has a clear and very substantial impact on opportunity is in the prevention of financial bankruptcy. This is a known problem both in high (notably the United States) and low-income countries. The WHO framework for defining national priorities when moving toward UHC in countries where this is still lacking lists financial risk protection as a key criterion for determining whether an intervention should be prioritized. ¹⁹

Lastly, a different route could improve on the summary or whole-life measures by measuring health inequality at different life stages. This approach would also give a more accurate picture of the overall health trajectory. ²⁰

The Efficiency Argument for UHC

The equality of opportunity rationale builds on a very direct link between a particular explanation of the special moral importance of health and health care, and the justice requirements for a particular mode of allocation of health care. Moreover, it is presented as a strongly egalitarian rationale for UHC. An alternative, more pragmatic approach grounds UHC in arguments from efficiency. This approach highlights a different set of features as truly distinctive of health care, as opposed to its special moral importance.

One version of an efficiency argument that has been put forward is James Tobin’s “specific egalitarianism.” ²¹ Specific egalitarianism is the view that “certain specific scarce commodities should be distributed less unequally than the ability to pay for them.” ²² While the name “specific egalitarianism” may suggest that the view is grounded on egalitarian values, Tobin’s argument for this mode of allocation is only weakly egalitarian. Specific egalitarianism is a scheme for local, as opposed to general, reduction of inequality, effectively “limiting the domain” in which economic inequality is allowed to operate. It is introduced as a less demanding alternative to the redistribution of income, on the grounds that it can secure a more equal distribution of specific goods, without the efficiency and disincentive costs believed incurred by the high taxes required for general redistribution.

Within its broader remits, specific egalitarianism may take somewhat different forms and be justified on different grounds, depending on the particular good to which it is applied. Tobin’s case for specific egalitarianism for health care also rests on a further contingent argument that health care is in scarce and inelastic supply. ²³ Tobin’s concern is that unrestricted overconsumption by the rich will leave the poor with less than the basic minimum in a free market. By isolating these goods from the market, they can be distributed more equally, ensuring that everyone gets enough. Thus, the general argument for specific egalitarianism is an efficiency argument, albeit an efficiency argument that is at least partly based on a moral concern for the absolute welfare of the worst off.

However, the supply of health care today is not plausibly scarce and inelastic to the degree Tobin had in mind at the time of writing. If this key assumption is relaxed, the isolation of health care from the market would no longer be required to ensure the basic minimum for all. Different arrangements would be possible: for example, health care could be entirely market-based, ²⁴ but with subsidized access for the poor. Or public provision could ensure the basic minimum for all, free of charge, but this could coexist with a private market for the provision of additional health care services. As such, as applied to today’s context, Tobin’s argument is fairly indeterminate in that it does not give us reason to prefer one type of health system over another.

A more detailed efficiency argument is put forward by Joseph Heath. ²⁵ This is also a more determinate argument in that it specifically makes the case for state involvement in the funding of health care. This is a two-part argument that focuses on first, certain unique features of health care needs, and second, the market failures associated with health care.

Health care needs are generally unpredictable, unequally distributed, and potentially very costly. In these respects, health care differs greatly from other basic human needs such as food. Patterns of health care usage tell us that some people will need (a lot of) health care, but with some exceptions, we do not know who they are. The costs of health care services, moreover, range from trivial to very high. Individuals would not know how much to save to cover future health care needs, and the costs of saving enough to cover future catastrophic health care expenses would be prohibitive. One way to solve this problem would be through the pooling of risks of need for costly health care through a system of health care insurance.

However, as is well-known, a private market in health care insurance is subject to several forms of market failure, including adverse selection and moral hazard. Of course, various market failures exist for many other goods too. But the point was effectively made by the British economist Alan Williams, who famously compared health care to the duck-billed platypus, that health care is unique in that it is associated with three distinct sources of market failure, which is more than for any other commodity. ²⁶

These features of health care make a strong case for some form of state involvement, which allows us to correct these market failures. State involvement would take the form of either a regulated insurance system (often called “social insurance”), or a tax-funded national health care system like the UK’s NHS. ²⁷

The efficiency argument demonstrates that we do not need to rely on the special moral importance of health care in order to justify UHC. However, we do not need to see the two argumentative strategies as mutually exclusive. A more general lesson to emerge from this discussion is the case for embracing pluralism and complexity both in the evaluation and justification of policy. A policy can have many effects and be valuable for many different reasons.

Lastly, further questions remain about which health system is the better choice for securing UHC. Different countries have organized their health care systems in different ways, usually relying on a mix of public and private funding. These different systems may vary with respect to, for example, the level of choice and level of inequality of coverage that they allow. Thus, the question remains how much inequality in access to health care, above the basic minimum, is morally justifiable. A further important empirical question is which system, if any, is better at reducing health inequality. ²⁸