Show Summary Details
Page of

Justice, Resource Allocation, and Emergency Preparedness: Issues Regarding Stockpiling 

Justice, Resource Allocation, and Emergency Preparedness: Issues Regarding Stockpiling
Justice, Resource Allocation, and Emergency Preparedness: Issues Regarding Stockpiling

Norman Daniels

Page of

PRINTED FROM OXFORD MEDICINE ONLINE ( © Oxford University Press, 2022. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Medicine Online for personal use (for details see Privacy Policy and Legal Notice).

date: 28 June 2022


Because disease and disability impair normal functioning, they limit the range of opportunities people would otherwise have, given their talents and skills. Therefore, health—or normal functioning—is of special moral importance from the perspective of social justice because it makes a limited but significant contribution to the range of opportunities open to people (Daniels, 1985, 2008). If, as many believe, society has an obligation to promote equality of opportunity, then it has a social obligation to promote population health and to distribute that health equitably. Meeting this obligation requires a just distribution of all the determinants of health, including public health services and personal medical services (Daniels, Kennedy, & Kawachi, 1999, 2000).

Many health needs arise regularly and predictably and must be planned for accordingly. Other population health needs arise less predictably as a result of natural and manmade emergencies. The obligation to protect population health and distribute it fairly should accordingly also govern the approach to emergency preparedness, including the need for stockpiling medical supplies and treatments for a range of health emergencies. These obligations apply in regional emergencies following natural disasters or more localized forms of chemical and radiologic terrorism; they also apply in national or international health emergencies that might follow natural pandemics or some forms of biological terrorism.

What principles, considerations, or procedures should play a role in planning for population health emergencies, including developing stockpiles of medical resources? The traditional emphasis in public health planning is on health maximization strategies—either minimizing the aggregate impact of certain health problems (e.g., lowering mortality rates, maternal mortality rates, or infant mortality rates) or maximizing certain measures of health benefits (e.g., life expectancy to life years or quality-adjusted life years [QALYs] gained or disability adjusted life years [DALYs] saved). More recently, however, the public health movement has paid more attention to promoting equity in health, both at national and international levels (World Health Organization Europe, 2002; World Health Organization Regional Office of Europe, 1999). In a simple maximizing principle or strategy, distributive effects can be comfortably ignored, and public health officials may have a clearer sense about their obligations, although they still face problems arising from uncertainty. When they realize that obligations of justice involve the pursuit of both equity and population health improvement, however, clarity and agreement on goals are harder to reach. People will disagree about how to resolve conflicts among these objectives. This perplexing problem arises in emergency planning as well as in meeting standard health needs. In both contexts, an appeal to principles must be supplemented with reliance on a fair, deliberative process (Daniels & Sabin, 1997).

Planning for emergencies is not itself an emergency situation. Planners should not be tempted into thinking that all nuance or subtlety about ethical obligations can be finessed because they are considering a crisis situation. Proper planning ensures that due care is taken regarding how to meet moral obligations in actual emergencies. In this chapter, I examine more systematically the conflict between maximization and equity, especially as it arises in the context of stockpiling for emergencies. Because the many disagreements that arise about all of these matters are ethical in nature, questions about the legitimacy and fairness of stockpile design and other questions of emergency preparedness will be addressed. To that end, I shall briefly describe the kind of fair, deliberative process that should be used in planning for public health emergencies. The account of fair process provides a framework for thinking about community participation in planning and the proper communication of decisions.

Emergency Preparedness Versus Other Health Needs

A Continuum Versus Exceptionalism

Justice requires that normal functioning be protected in a population and that the health that results be distributed. This is optimally accomplished if there is a just distribution of the social determinants of health (including basic liberties, education, effective political participation, control over life and work, income and wealth) as well as an equitable public health and medical system. Such a system would emphasize health risk reduction, the equitable distribution of risks, and appropriate forms of medical prevention and treatment for chronic and acute health conditions. The relatively predictable prevalence of standard public health and medical needs means that, on a population basis, there is little uncertainty plaguing health planning. At the same time, different health needs compete for scarce resources. As important as health is, it is not the only important good that must be protected or provided, and so resources for health compete with other important social needs and goals. This means the problem of priority setting for resource allocation is pervasive and unavoidable in public health planning, even when problems of uncertainty about population needs are not significant.

Natural disasters, including pandemics, as well as forms of biological, chemical, and radiologic terrorism, add considerable uncertainty to the planning for public health emergencies. The dramatic threat posed by worst-case scenarios for disasters, such as a pandemic on the scale of the 1918 influenza pandemic, may lead some to think that emergency preparedness is not on a continuum with ordinary planning for meeting health needs and that the scale of such threats requires thinking about health needs in an entirely novel way. That is a mistake for two main reasons. First, the best preparation for major emergencies is a properly functioning public health system that makes appropriate allocation of resources for both emergencies and ordinary needs. Second, the considerations involved in thinking about appropriate resource allocation across the range of emergency and ordinary health needs raise common issues that need common solutions. Emergency preparedness “exceptionalism” would be a self-defeating strategy.

Criticism of “biodefense” that has surfaced in response to new budgets and priorities that emerged after the September 11 attacks and the US anthrax threats of October 2001 should be understood in light of these two points about the relationship between emergency preparedness and public health systems. An awareness of the implications of globalization should also be incorporated into the public health system. Biodefense should not be “us” versus “them” but a focus on how “we”—citizens of the world—can address global health issues. That means standard methods and principles for reasonable health planning must be extended to emergency planning rather than abandoned for some new “war” footing.

One of the institutional implications of the argument that emergency preparedness requires good public health system building is that countries with global health budgets are better positioned to think about appropriate resource allocation than are countries with fragmented health systems and multiple budgets with different incentive structures. If emergency preparedness means stockpiling resources that are then not readily available for meeting ordinary health needs, that is better done as a tradeoff within a global budget. Where different budgets address these competing needs in a more fragmented system, unnecessary redundancies or other kinds of inefficiency and inadequacy in decision-making are more likely to be found. Similarly, compliance with restrictions on resources might be easier to achieve if the resource allocation results from a closed budget that puts all people on the same footing; otherwise, compliance is more readily threatened by “gaming of the system” (Daniels, 1986). Specifically, a mixed system with competing budgets and resources might lead to more rampant forms of hoarding and noncompliance with resource restrictions than a system that places competition for resources on a level playing field.

Uncertainty and Planning

In resource allocation decisions for ordinary public health and medical needs, uncertainty plays only a small role at the population level. It is possible to know the prevalence of various health conditions and make fairly accurate estimates of what is needed to meet them during the course of a budget or planning cycle (Murray & Lopez, 1996). An extensive body of actuarial experience is available that allows accurate projections of medical and most public health needs. In contrast, emergency preparedness poses the problem of uncertainty about needs (What exactly will be the outcome of some health emergency at the population level?) and about probabilities (What exactly is the likelihood of a specific emergency arising in a given time period?). Because there is so much uncertainty about both, considerable disagreement arises about appropriate responses.

Since worst-case scenarios are particularly frightening (e.g., major influenza pandemics that might kill millions of people globally and significantly disrupt economies), people who are especially risk averse will want to devote considerable resources to preparing for these events. They might even defend their perspective by suggesting that when the global stakes are as high as these scenarios suggest, an appropriate rule of choice for planning is to minimize worst outcomes. Others will want to put more effort into planning for more frequent and less serious emergencies. They are less inclined to abandon more standard Bayesian rules of choice even without having great confidence in the estimates of either outcomes or probabilities. There is no clear model of rational planning that resolves such disputes about resource allocation in the face of uncertainty. This source of reasonable disagreement about rational choice has similar implications to the reasonable ethical disagreement about tradeoffs between maximization and equity. Both suggest a need for fair process to establish legitimacy for decision-making.

Actually, the two problems may interact. One author has argued that the reservations that planners have about making “social worth” judgments about individuals in medical need in standard, nonemergency contexts may not be appropriate in serious emergency contexts that involve deep threats to the functioning of society and not merely substantial health needs (Arras, 2005). In effect, in planning for worst-case scenarios, health officials may need to modify views about fair treatment and equal respect for persons that should govern standard contexts. In short, if it is rational to plan for extreme scenarios, then ethical restraints may need to be modified as well. This is not to endorse this conclusion but to note that reasonable people may not only disagree about the weight to be given worst-case scenarios in planning, but they also may invoke different or novel ethical considerations in those contexts as well.

Another consideration about uncertainty is that it may seem reasonable to adopt strategies based on poor-quality evidence when few options are available and the stakes are viewed as very high. An illustration is the World Health Organization (WHO) recommendations regarding treatment and prophylactic use of oseltamivir for H5N1 viral infections in humans (Shunemann, Hill, & Kakad, 2007). Despite weak and indirect evidence for the effectiveness of oseltamivir against H5N1 for either treatment or prophylactic purposes (there is only some indirect evidence from animal models for H5N1 and from uses of the drug for seasonal influenza), the high fatality associated with the illness and the absence of other treatment and prophylaxis regimens led to strong recommendations for its use from an international panel of experts. The risk involved in this recommendation is that too much weight is placed on one set of measures for containing a pandemic, at least in the mind of the public and many providers, perhaps to the exclusion of other public health strategies, such as the simple hygiene measures that are a standard first-line of defense and are emphasized as well in pandemic influenza guidelines. A consequence might be that the public becomes overconfident, thinking that adequate steps have been taken to address a potential pandemic because an antiviral drug of questionable effectiveness is recommended and stockpiled. In view of such effects, again disagreement might arise about the strength of the recommendation, given the limited available evidence about the drug’s effectiveness.

Maximization in Emergency Preparedness Planning

Maximizing What?

Understanding the rationale for maximization and the targets of maximization is based first on an understanding of utilitarianism. Utilitarianism, one of the most familiar consequentialist ethical frameworks, is a theory with two main parts (Sen & Williams, 1982): one part is a theory about ranking worlds on the “better than” relationship; the other is a consequentialist principle that promotes the aim of producing the best alternative world. For utilitarianism, one world is better than another if it contains more aggregate welfare in it than another, where the principle of aggregation is sum ranking. This is determined by simply adding up all the welfare (conceived of as net pleasure minus pain, or net happiness, or, in more contemporary terms, net satisfaction of preferences) of individuals to arrive at an aggregate welfare for the world containing those individuals. Other welfarist theories might chose a different way of ranking worlds—for example, one world might be better than another if the worst-off group in it is better off than the worst-off group in the other world—but utilitarianism uses this simple principle of sum ranking. The consequentialist normative principle, then, is to aim to produce the world with at least as high a rank as any other.

The goal of utilitarianism is to maximize aggregate welfare (however it is conceived), thus capturing a nonmoral notion of goodness. The theory thus builds on the idea that the right thing to do is to produce a world with the most goodness in it, and this has some intuitive force (Sidgwick, 1907/1966). However, utilitarianism does not simply call for maximizing aggregate health in a population. Maximizing aggregate health might have some tendency to promote goodness in a world, but other things contribute to goodness from a utilitarian perspective and might compete with the straight maximization of health. If, for example, some people contribute more to society, adding more welfare to the world than others, improving their health rather than the health of others would contribute to net welfare more than maximizing health in the aggregate. Keeping working people healthy, for example, produces more indirect benefits from investment in health than keeping retirees or the unemployed healthy, so, following utilitarian thinking, health resources would be allocated where they produce the greatest aggregate welfare, which may not be the same thing as where they produce the greatest aggregate health.

When public health approaches maximize health and ignore other goods that may contribute to welfare, they depart from strict utilitarian principles. Such a departure may be justifiable, but not from a utilitarian perspective. For example, some might argue that emergency preparedness and response (EPR) planners must show equal respect for people and interpret that to mean that they should not make judgments about the social worth or social contribution of individuals when assessing their competing claims for health needs. Rather, they must assess the strength of their claims according to their needs alone. Such a view is a departure from a theory that stresses maximizing aggregate welfare.

The article cited previously (Arras, 2005) argues for equal respect in ordinary situations where the social fabric is not seriously threatened. Where the social fabric is threatened, as in some general health emergencies, health officials may consider the contributions of various people to maintaining the social order and assign priorities accordingly. One form of this argument is that there should be less reluctance to embrace broader utilitarian concerns when the stakes are very high in the aggregate, as they are when the social fabric is so threatened. In effect, the argument states that planners can afford to be concerned about equal respect when the stakes are modest, but when they are very great, they should focus on avoiding great harms.

A critic of this view might counter that the more privileged and powerful people during a time of crisis will always assert their greater social contribution and steer benefits to themselves. The consequence will be that relaxing scruples in time of crisis opens the door to serious errors about real social utility. Even if public health planners focus on avoiding great harms, including threats to the functioning of key social institutions, they are not likely to succeed if they ignore the principle of equal respect. Claims from well-situated people about their societal importance are likely to be confused with more objective assessments of what kinds of tasks are essential to protect.

Maximizing What Measure of Health?

A public health approach that aims at maximizing health in the aggregate, leaving aside other goods—including in times of emergency—must make choices about what measure of health it aims to maximize. Should we save the greatest number of lives? Should we save the greatest number of life years? Or the greatest number of QALYs? Each choice has ethical dimensions that will provoke disagreement.

Considering the choice between maximizing numbers of people saved versus number of life years saved, the latter appears to give priority to saving younger people rather than older ones; the former implies that age should not matter and that life is valuable independently of how much of it has already been experienced. In maximizing lives saved over life years saved, EPR planners might be charged with being insensitive to the greater “need” the young have for more years than the old, or how much worse off they would be dying young than would those who are older, or, as one conception of fair allocation over the life span would have it, that it might be prudent for everyone to favor using scarce resources to maximize life years saved if they did not know how old they are and had to choose. Arguments about fairness pull in both directions, and reasonable people may disagree about them.

In some situations, maximizing lives saved might be preferred, but in reality it proves easier to save more lives in areas of denser population than in more sparsely populated areas. This might lead to concentrating resources in ways that give a much greater chance at being saved to some groups than to others. Maximizing numbers of lives saved is being done at the expense of giving people equal, or even fair, chances at being saved. Some would argue that maximizing numbers of lives saved is a form of favoring “best outcomes” over “fair chances.” Others, however, would prefer giving more weight to fair chances than would happen with always favoring best outcomes. Therefore, this particular form of maximization immediately encounters one of the standard distributive problems that has been labeled an “unsolved rationing problem” (Daniels, 1993).

Favoring the maximization of life years saved over lives saved presents another version of the same problem. Here, the best outcome is saving the most life years—that is, favoring the young over those who are older (other things being equal). But this choice means that older people lose all chance—including a fair chance—at any significant benefit so that the young can produce a best outcome. In addition, if maximizing QALYs (or saving DALYs) is favored, planners may readjust what counts as a best outcome, but they still encounter the same distributive problem.

It might be thought that a strategy of saving the most lives would be defensible both on consequentialist and nonconsequentialist grounds because a life saved can be considered a good outcome by a consequentialist, and a principle treating all lives as equally worth saving might seem to be fair to all people since all have claims to be saved. However, a consequentialist might argue for considering the indirect benefits of saving some lives as compared with others, so that more valuable outcomes might result if those lives were saved that contribute more to society. Saving a life is a good, but other goods must be considered as well. Similarly, the nonconsequentialist might complain that some people have stronger claims on being saved than others, perhaps because their dying at a younger age makes them worse off than someone who has lived a considerably longer life. In short, reasonable people will disagree about even the principle some consider most obvious in emergency situations, namely, a principle calling for saving the most lives.

Division of Labor: An Argument for Maximization?

Each choice of what measure of health to maximize produces some version of broader concerns about fairness in the distribution of health benefits. Still, there is some plausibility to thinking that the responsibility of people dedicated to protecting the population’s health, including in emergencies, should be concerned with maximizing some measure of health. This is plausible because of the efficiency that results from a proper division of effort or responsibility. If those in public health are charged with maximizing health, they are likely to produce more health than if they are charged with more complex judgments about how to trade health against other goods. Fairness across sectors or spheres, then, becomes a problem for some broader social agency—a democratic political process, perhaps—that is responsible for arriving at an overall fair social policy. But experts within each sphere should be charged with the task of doing what they know how to do best, namely, maximizing the good that sphere is charged with producing. One possible implication of that argument is that public health agencies should pick a measure of health and then aim to maximize it and not be troubled by the nuanced concerns about fairness. Unfortunately, the ethical disagreements noted here would mean that such a strategy would be challenged and would raise questions about the legitimacy of a strategy that deliberately ignored concerns about fairness.

Does Helping Worst-Off Lead to Maximizing Aggregate Health?

A further, traditional support for adopting a maximizing strategy is the hope, often expressed by proponents of public health, that the most successful strategy for promoting aggregate health is to improve the health of those with worse health. This group can make the biggest gains and thus contribute the most to maximizing population health in the aggregate. In eradication efforts for some infectious diseases, substantial gains in measures of aggregate health have resulted from efforts to help the poor with endemic conditions.

Although this may hold for some public health programs, it is not in general true. Worst-off groups may be very inefficient converters of resources into health compared with somewhat better-off groups, so one cannot assume that health maximization and giving priority to those worst off pull in the same direction in general. Indeed, various social science studies have found that many people are willing to trade maximization of aggregate health for making more modest gains that improve those who are worst off. In short, health maximization does not follow, at least in general, from targeting the worst-off parts of the population.

Maximization Versus Equity: The Case of Stockpiling

An illustration of how reasonable disagreement arises about key choices in emergency planning is the development of stockpiles of key medical resources. I believe there is a lack of consensus on fine-grained principles that could resolve some disputes about these choices in the goals and design of a stockpile. In the absence of such a consensus, a plan will have legitimacy and be perceived as fair only if it is developed in a way that ensures accountability for its reasonableness. The issues of legitimacy concern more than simply who decides: legal authority, and perhaps moral authority, for decision-making may be clear, but how decisions are made also affects legitimacy. In the following paragraphs, key choice points about stockpiling will be discussed. I will describe the conditions that a fair deliberative process must meet if it is to establish the legitimacy and fairness of the choices made.

Stockpiling and Public Goods Versus the Concept of “Shared Responsibility”

A policy debate that cuts across many of the specific issues regarding stockpiling is that of “shared responsibility.” Shared responsibility is the claim that the federal government does not have sole responsibility for providing protection to people in the context of medical emergencies, such as an influenza pandemic. Rather, since benefits accrue to many people and many institutions, responsibility for providing those benefits can be divided across federal, state, and local levels, both public and private. Ultimately, shared responsibility could devolve to individuals as well, who might, for example, have the responsibility to acquire and fill a prescription for prophylaxis or treatment use of an antiviral drug.

A basic consideration of social justice is the role of the government in providing for public goods. In contrast to liberal egalitarian views that assign to the government broad obligations to protect the well-being of groups and individuals, even theories of justice that assign a modest role to the government generally agree that it has a special obligation to protect the provision of public goods. Providing security against infection in the context of pandemics is one such obligation; police powers of quarantine derive from this obligation to provide security, not only from external aggression but also from infectious disease as well. If individuals can be prevented from becoming infected, they will not infect others. Being kept free from infection is thus not just a good for the individual but also affects the security of others. If individuals are left to protect themselves, and some fail to act responsibly or cannot protect themselves because of limits on their resources, this then exposes third parties to substantial harms. It is this feature of protection from infectious disease that necessitates assigning primary responsibility to society—to government—for providing the public good of health protection in the context of infectious disease.

Some specific obligations might be assigned to specific institutions in order to ensure the effective provision of the public good. Such obligations could fall to various federal, tribal, state, and local public agencies, as well as to various private organizations. That assignment of responsibility would require the force of law or regulation, and it would have to be backed up with sanctions and strong forms of accountability. It could not be left to the private or voluntary assumption of “responsibility” because people may find excuses for why other things take priority over this responsibility. If, for example, certain private corporations failed to purchase and provide oseltamivir prophylaxis to their employees, they might pay the private penalty of losing part of their workforce. Those employees are then likely to infect others as a result of their not being “responsibly” protected. These externalized harms point to the need for stronger assurance that the public good of protection is provided. The state cannot escape its obligation to provide a public good through attempting to pass on assumed responsibilities. However, with the appropriate legal and regulatory framework, the obligation can devolve to appropriately accountable entities, both public and private.

Assigning those obligations, however, will raise important issues of justice. If private entities are required to assume such obligations, then it is fair to expect them to be able to meet those obligations. For example, small-scale private purchases of antiviral drugs are likely to be much more costly than large-scale purchases at government rates. Individuals, groups, or institutions without the resources to meet those obligations cannot be required to do so without appropriate redistribution of resources. Blanket appeals to individual, corporate, or even community responsibility are not a substitute for careful assessment of the fairness of assigning obligations to those entities, given their actual capabilities to meet the obligations.

In short, the concept of “shared responsibility” seems to fit more naturally with the provision of individual goods, not public goods. Where public goods, like protection of a public against infection, are involved, the more appropriate concept from the perspective of justice is “obligation,” not the weaker notion of responsibility; and where the obligations are to be shared, there must be adequate attention paid to the fairness of the distribution of the obligations. Appealing to a notion of shared obligations to provide a public good does not avoid questions about distributive justice; it simply raises those questions in another form. If the issue is how to divide the burden of paying for protection against a pandemic, for example, then society cannot distribute obligations to pay for parts of that burden without being sure it is just or fair to do so. The danger of the language of “shared responsibility” is that it may smuggle into the “sharing” a notion of a different kind of good (i.e., a private good that one can choose to have or not have, when in fact the decision not to have the good imposes risks on third parties that are not consented to and are arguably unjust).

This does not constitute an argument against federalism, which is thought of as a division of obligation between federal and state governments. Nevertheless, one caution does apply to the issue of federalism; that is, there are externalities to state-level actions that may require strong forms of federal regulation. If one state fails to purchase drugs needed for treatment, for example, and symptomatic residents infect other people, the consequences extend beyond state boundaries. Indeed, some metropolitan areas straddle state lines, and ongoing interstate commerce, including the movement of workers, means that treatment and containment efforts must be coordinated across state lines. Where there is a risk of state-level failure to contribute to the provision of the public good of protection against infection, there must be stronger federal control. This does not mean that obligations cannot devolve onto states for purchase of some medical supplies, but only that there must be strong forms of accountability for federalist divisions of obligation. Here, too, “responsibility” may be too weak a term.

Goals or Objectives of Stockpiling for Pandemics

The goals or objectives of stockpiling medical resources for a pandemic are often coupled with specific prioritization strategies for their use. Thus, if a vaccine or antiviral drug were in short supply, it might best be used selectively to try to contain an outbreak; if it is plentiful, it might best be used to protect everyone. If it is being used to protect everyone, the remaining prioritization questions have to do with the sequence of the administration or distribution effort, and such choices may involve different degrees of risk for different groups. This combination of goal selection and prioritization issues can be used to illustrate how ethical disagreement can surround many stockpiling choices and, ultimately, to argue the need for fair process. Although a national stockpile may seem to avoid the prioritization issues because states may be responsible for determining how to prioritize the use of resources released from a national stockpile, it is nevertheless important to understand the distributive issues that arise as a result of the combination of resource scarcity and prioritization decisions, regardless of how authority for making them is currently divided.

A preliminary point that cuts across these cases is that the scarcity of a resource in a stockpile may result from quite different factors. It might result from an initial decision that underestimates what the needs will be, from a budget limitation, or from problems in the market for the drug or device (i.e., too few producers or too little being produced). Sometimes, the decision that creates the scarcity may be criticized on ethical grounds; sometimes the scarcity is not itself the result of bad decisions, but the prioritization decisions that are then made because of the scarcity may themselves be subject to criticism.

Vaccines and Their Uses

To illustrate ethical disagreements that might focus on vaccine use, I present hypothetical cases and distinguish among them. First, suppose that a vaccine for a potentially pandemic avian influenza is available early in an outbreak. If it is available in quantities capable of protecting the whole population, the only prioritization issues concern staging of the mass vaccination campaign. Such prioritization questions may involve some additional risks to those who are among the last groups to be vaccinated. Suppose, for example, urban populations were to be vaccinated first, in part because the risk of early outbreak was estimated to be higher in more densely populated areas. Still, given population mobility, including workers commuting to cities and others visiting or shopping there, some risk remains to nonurban populations that is ignored by the prioritization. People in the areas where immunization campaigns are delayed could argue that they are being denied a chance at a substantial benefit (early protection) because a best outcome (i.e., preventing the most cases) is achievable by giving priority to urban populations.

Put in this form, their complaint would be a version of the classic “best outcomes versus fair chances” distributive problem. The nonurban people who are at some risk, even if at less risk than urbanites, can complain that they are being asked to forgo all chances at early protection so that others, at admittedly higher risk, can obtain a better outcome. Reasonable people might then disagree about how much weight or priority to give best outcomes versus giving more people a fair chance at some benefit. Even if they agree that always favoring best outcomes is unfair to those with less than best outcomes, they may not agree on how to weigh the two considerations: fair chances at some benefit versus better outcomes. Some propose a proportional “lottery” that takes both factors into account, although just how the proportions are established remains a point of controversy (Brock, 1988; Kamm, 1993).

The problem may be characterized in another way. The urban population has a higher risk of early infection. In this regard, it is “worst off” than the lower risk nonurban population. The prioritization strategy might then seem to be a case of giving priority to those who are worst off, which is an equity consideration favored by many. However, giving the worst off complete priority, regardless of the benefits forgone by others, seems unacceptable to many, but giving no priority to the worst off is also unacceptable to many. But how much priority should we give them? (I have previously labeled this unsolved problem the “priorities problem” [Daniels, 1993].) To illustrate, the objection from the nonurban population might be recast as follows: giving complete priority to the urban population and ignoring providing any early protection to those at (admittedly) lower risk, but still at some risk, is a case of giving too much priority to those who are worst off. Perhaps it is possible to identify the subgroups of the nonurban population at comparable risk to subgroups of the urban population and to extend as much protection to them. Reasonable people will disagree about how to weigh the alternative strategies. Even if the example is changed to suggest that the vaccine is not in abundant supply, but is in limited supply, this brings up variations on the same distributive problems.

Antiviral Treatment and Prophylaxis

Other distributive problems arise in the context of stockpiling as well, and these may be better illustrated by the example of antiviral drugs. Suppose there is a supply of antiviral drugs, such as oseltamivir, which can be used for treatment, but the drug is not in adequate supply to treat everyone who could potentially benefit from it. Suppose, for example, it is less effective for persons with more advanced symptoms than it is for those with recent exposure who could take it either as treatment (with early symptoms) or as prophylaxis (if not yet symptomatic). Suppose, however, that those with the more serious cases sometimes benefit from treatment, perhaps avoiding more serious morbidity or even death, and suppose further that this benefit is produced only at higher dosages, thus decreasing the benefit that might be given to others with less serious disease or as prophylaxis.

Again, there are two ways to describe this kind of example. If health officials give the drug to those with serious illness, despite its lesser effectiveness, they are giving greater priority to those already worse off. Those who would then have to forgo earlier treatment or prophylaxis could complain that too much priority has been given to those who are worse off, and health officials have ignored important claims that others at risk have on resources. If, however, they favor “best outcomes” and then reserve the drug for use by those who will get more benefit from it, then the already seriously ill persons who give up any chance at benefit can complain that they are being sacrificed to the better outcomes of others. It is not fair that they give up all chance at any (significant) benefit so that others may benefit more.

Either way the distributive problem is described, it involves ethical disagreement. Moreover, reasonable people will take different positions on these disagreements. There is no way simply to dismiss one or another position as irrational or based on irrelevant considerations.

These “unsolved rationing problems” hardly exhaust the reasonable ethical disagreements encountered in choices about how to use stockpiled resources. One further general problem focuses on the question of whether certain groups, identified by their roles or functions, should be given priority in access to treatment or preventive protections. There is probably widespread support, for example, for giving priority to those who are expected to face exceptional risks from contact with people who can infect others. Many people would support giving priority access to preventive measures to the first responders of various kinds (e.g., emergency medical technicians and fire and police personnel) and health care workers (e.g., physicians, nurses, or other technicians), who are at higher risk but whose expertise is necessary to providing care during a pandemic. However, despite the support many give to such a prioritization, others worry that it violates basic concerns about showing equal respect for all persons, despite their social contribution. The so-called “God” committee that once functioned at a Seattle hospital has long been held up as an example of what should not be done in medical settings. The committee made life-and-death choices about access to renal dialysis early in the development of that technology that were based on questionable criteria about social worth (stable job and family, church attendance) (Levine, 2009). The complaint against singling out people for special medical priority because of their social role or contribution is that it risks importing questionable social worth judgments into a setting where medical need should be the sole criterion for access to services. If that is the criterion, then (or so the argument goes) persons should be shown equal respect.

Sometimes it is argued that a principle of reciprocity implies that special priority should be given to those who are expected to take special risks. The principle of reciprocity, in this view, allows people to favor giving priority to some people over others without violating their concern for equal respect. They are not judging the social worth of health care personnel (for example) but acknowledging the special burden they face by offering them compensatory protection. Note that this argument from reciprocity would not give comparable priority to those performing other essential tasks but who are not at extra risk and who are not expected to act in the face of those extra risks. A further problem with the appeal to the principle of reciprocity in this way is that, as stated, it balances risk taking with risk reduction. In addition, some might think a principle of reciprocity can be applied more broadly and can balance other considerations—perhaps providing rewards for other social contributions. If so, the principle both opens the door to challenging the concern for equal respect and also to a much broader justification of other priorities. Any socially essential task might be thought worthy of reciprocity in the form of special protection. The problem here is general: there is not a clear idea of the scope of the principle being invoked as a principle of reciprocity, and so reasonable people will have disagreements about what it implies. Far from resolving the disagreement referred to earlier, it may add to it.


The stockpiling of ventilators to meet a “surge” in need during a pandemic raises further ethical issues. Some of these issues arise at the federal level, but others may primarily be a matter for states to address, given that states will have to formulate policies for the use of resources released to them from a national stockpile.

The first key ethical issue that arises at the federal level is how large a stockpile of respirators to create. Ventilators, unlike vaccines or antiviral drugs, cannot be stockpiled in quantities adequate to meet the estimated need for them in pandemic worst-case scenarios. The cost for such stockpiling would simply be too great; arguably, the gap in production capacity for ventilators adds a market-based reason for thinking there is no possibility of meeting the needs of a worst-case scenario. But even if stockpiling for a worst-case scenario is not possible, people will disagree about what level of investment is reasonable, and there is no principled way to determine that level. Any budget constraint can seem arbitrary, given that lives may be lost if the stockpile is smaller than it need be. Yet variations in aversion to risk may be one of the key factors shaping people’s willingness to invest in stockpiled ventilators.

A second ethical issue at the federal level concerns the judgment about how many ventilators to release to various states as an epidemic unfolds. Giving too many ventilators to states with early outbreaks will be unfair to states that face later outbreaks, but withholding ventilators from those already in need because others may come to need them elsewhere will be challenged by those who need them early. Some people will definitely die because the ventilators are not distributed to some areas, whereas there is only a risk that others will die elsewhere later in the epidemic. If there is a high likelihood that the epidemic will spread, withholding of ventilators will seem geographically more equitable than if the spread is less certain. This, too, is an issue on which reasonable people will disagree and for which there is no principled way to arrive at consensus about a fair distribution.

At the state level, where the supply of ventilators does not meet the need, ethical issues will arise about how to ration their use (New York State Workgroup on Ventilator Allocation in an Influenza Pandemic, 2007). A standard triage procedure would reserve their use for those most likely to survive if they are given ventilators and to exclude those most likely to survive without them or to die with them. Of course, making this judgment is difficult. It is especially difficult if a new patient is more likely to survive if given a ventilator than someone already on it. Switching in such cases may be the strategy that would save the most lives, but it also means abandoning a patient already in treatment, and many health care professionals would find that ethically unacceptable. Even if people agree that switching is acceptable under some conditions, they may disagree about the details of the conditions that make it justifiable.

At both the federal and state levels, then, ethical disagreement will occur among reasonable people who face choices about ventilator stockpiling and uses. Consensus on principles that can resolve these disagreements is lacking. In many contexts of justice, where there is a lack agreement on principles that can resolve a dispute, planners must rely on a fair process and accept the outcome of that process as fair. I discuss this reliance on procedural justice in emergency preparedness further in the following section.

Private Stockpiling

Private stockpiling, if appropriately encouraged and regulated under the right conditions, could supplement public stockpiling in a way that is not unfair and might be efficient. However, if the private stockpiling is done in ways that compete with public stockpiles, it can produce an inefficient and unfair means of dealing with emergency situations. Just what measures should then be taken to address the private hoarding depends on the effect of both the hoarding and efforts to curtail it.

Worries about private stockpiling or hoarding arose in the context of early discussions of an avian influenza pandemic and shortages of oseltamivir. As an example, a prominent physician e-mailed all his friends at the height of the scare about human H5N1 infections during 2007, recommending that they secure supplies of oseltamivir for their families. Obviously, if wealthier, better connected people have a significant impact on a short supply, the chances of stockpiling it for fairer forms of distribution will be undercut. In addition, such stockpiling at retail prices is inefficient. Arguably, it is ineffective as well, since there will be so little clarity about how to use the drug (and no evidence of its effectiveness).

Since 2007, as plans for a national stockpile have advanced and a surge in production has been accomplished, there is less concern about the adequacy of supply of oseltamivir, although federal supplies remain limited because of funding limits, and prophylactic use of federal supplies is not recommended. In the context of a surge in infections and media attention, however, the interest in private stockpiling, temporarily abated, may reemerge. Some of that interest would be based on the public perception, after Hurricane Katrina, that American management of emergencies leaves something to be desired. People may fear that public preparedness cannot be relied upon. In that context, even assuming a national stockpile has been constructed that is adequate to meet the needs for influenza treatment and prophylaxis, some parts of the population will prefer to depend on their own initiative and resources. Under these circumstances, it would be unwise to intervene, since the private initiatives, however inefficient, are unlikely to affect public measures.

Where public preparedness is inadequate, however, private hoarding could pose a significant threat to public initiatives to remedy the situation. Still, if it is unclear that public measures can be significantly improved, stopping the private measures may prevent some unfairness but do so at the expense of denying people the only measures open to them to preserve themselves. The lesson from this is that appropriate, timely emergency preparedness measures are the best protection against widespread private hoarding.

The Need for Fair, Deliberative Process (Accountability for Reasonableness) in Emergency Preparedness Decisions

Legitimacy and Fairness Problems

The fact of life about resource allocation decisions, including those in emergency preparedness contexts, such as stockpiling, is that there are winners and losers. With winners and losers come disagreement and conflict. The conflict in these contexts, however, is not only about competing interests. Reasonable people will often disagree about how to weigh the values that generally compete in these contexts. I have illustrated this point in several ways. There are reasonable disagreements about how to address choices made under uncertainty, about how much priority to give to those who are worst off, about how to weigh the importance of aiming at best outcomes versus giving people a fair chance at some benefit, and about how much to weigh saving more lives against professional obligations not to abandon patients who can benefit from further treatment. These are all value questions, not technical ones. They unavoidably push toward deliberation aimed at morally and politically acceptable, or legitimate, solutions.

Where fundamental issues of well-being are the subject of such moral controversy, decision-makers aspiring to legitimacy must wear a mantle of moral authority. Under what conditions that moral authority is properly accepted as legitimate by those who are affected by the decisions will be referred to as the legitimacy problem (Daniels & Sabin 1997, 2002).

The legitimacy problem might seem less difficult, or perhaps no problem at all, if moral authority were easily exercised in the following sense: anyone could check to see if decision-makers make choices that conform with moral principles or reasons on which there is prior consensus. In effect, the public might care less about the conditions establishing legitimacy if the fairness problem had a straightforward, principled solution so that it was clear to all what outcomes counted as fair.

Unfortunately, there is no consensus on such principles and so no simple solution to the fairness problem. Instead, as the examples discussed earlier suggest, there is ongoing controversy surrounding competing values. Without a foreseeable consensus on such principles, EPR planners must find a fair process whose outcomes can be accepted as just or fair. The process must be fair to all who participate in it and who are affected by it; obvious sources of bias or conflict of interest must be removed. Attention must be paid to the voice—the values and interests—of different stakeholders. This is a classic appeal to pure procedural justice (Rawls, 1971), in which we rely on fair process to arrive at a fair outcome in the absence of prior agreement on the criteria or principles governing a fair outcome. This fair process—accountability for reasonableness—provides a way to resolve disputes about allocation that are not addressed by more general principles of justice on which people may agree.

Accountability for Reasonableness

Four general conditions ensure accountability for reasonableness. If met, they should, over time, lead members of the public to respect public agency decision-making for its fairness and legitimacy. Although these conditions were originally developed as a general characterization of fair process in health care resource allocation generally, they will be restated here so they focus on the kinds of decisions that must be made in emergency preparedness contexts, for example, in decision-making about the goals and design of a national medical stockpile. The main features of the process are in accord with requirements of public administrative law, so these conditions should readily be met in public agency decision-making. At the same time, the rationale for them provides further grounds for public communication about these decisions.

  1. 1. Publicity condition: Decisions regarding the goals and means for achieving them, including decisions about priorities in access to resources, as well as the rationales for them must be publicly accessible.

  2. 2. Relevance condition: These rationales must rest on evidence, reasons, and principles that all fair-minded parties affected by the decisions—managers, clinicians, patients and the public in general—can agree are relevant to deciding how to meet the emergency medical needs of a population under necessary resource constraints. (“Fair-minded parties” are considered people who seek grounds for their decisions that they can mutually justify to each other.)

  3. 3. Revisability condition: There is a mechanism for challenge and dispute resolution regarding decisions, including the opportunity for revising decisions in light of further evidence or arguments.

  4. 4. Enforcement condition: There is public regulation of the process to ensure that conditions 1–3 are met.

These four conditions ensure that what might otherwise be largely behind-the-scenes public agency deliberations become a larger, public, ultimately democratic deliberation about using limited resources to protect fairly the health of a population in emergency conditions. The four conditions set the stage for a process of interactive education among all parties, built on accountability. In this way, they provide a foundation for thinking about public communication and education about emergency preparedness and the various obligations and responsibilities that result.

Because emergency preparedness is a context in which resources are clearly limited and priorities have to be set about their use and the criteria that should govern that use, the value of developing a culture of openness about rationales must be recognized. Such openness must take hold against a background in which the US public—and the politicians and health system managers who respond to that public—has little understanding of the need for setting limits or priorities in standard medical contexts. Perhaps the public would be more open to limits in emergencies, but if it has been educated to resist all limits, emergency situations would be more difficult to manage fairly. To change that culture in the long run requires a concerted effort at education, both outside and inside the institutions that deliver care. That education begins with openness about the reasons for the decisions that public or private health providers and insurers make. Over time, this process enables a more focused public deliberation that involves broader democratic institutions. But whether or not the stage is set in this way, it is crucial that decisions made about emergency preparedness engage the public in an open and transparent way.

The publicity condition provides a public record of the ethical commitments to which the agency officials responsible for emergencies adhere in making these kinds of decisions. In effect, the decisions form a kind of case law record, complete with rationales for why decisions are being made. Arguably, this feature improves fairness in decision-making because it provides a basis for judging the coherence and consistency of decisions about emergency preparedness made over time. It gives those affected by decisions, often when they have no real choice to seek alternatives, a way of knowing why they face the restrictions they do. The publicity condition thus satisfies what many believe is a fundamental requirement of justice: the grounds for decisions that fundamentally affect a society’s well-being must be publicly available to that society.

The relevance condition imposes important constraints on the kinds of reasons that should play a role in rationales for decisions about goals and priorities. Ideally, it narrows the range of disagreement. The basic idea is that parties pooling resources to face an emergency pursue a common goal or common good. They enter into a plan that aims to meet their diverse needs under necessary resource constraints on terms they can justify to each other. A fair-minded person can be considered someone who believes decisions should be made on grounds that people can justify to each other. Because hard choices will have to be made about how to meet those needs fairly, the grounds for those decisions must be ones that fair-minded persons can agree are relevant to that kind of decision.

Involving various stakeholders in deliberation is one way to secure broader agreement about what count as relevant reasons. Such “buy-in” is important in generating a sense of legitimacy and fairness. However, the main mechanism for improving legitimacy through stakeholder involvement is not an appeal to a form of grassroots democracy. Most stakeholders who can be involved in the deliberation, whether through open hearings, participation in the whole process, or public comment on elements of the process, are not elected representatives who can claim democratic credentials. Because they are not representatives who through a selection process are held accountable to their constituents, their participation should not be seen as making the deliberation “more democratic.” Nor do such consumers act as proxy consenters on behalf of other consumers. Instead, they add to the breadth of considerations taken up in the deliberation and, through their effectiveness in representing specific views or considerations, improve the quality of argument. Stakeholder participation is given meaning and direction and a connection to the problem of legitimacy if its goal is to improve accountability for reasonableness. It can do this by enhancing the deliberative process, reassuring various stakeholders their arguments are addressed, broadening perspectives on what counts as relevant reasons, and ensuring the transparency the process requires. At the same time, stakeholders risk distorting deliberation by becoming lobbyists for vested interests. There are both benefits and risks to the broad inclusion of stakeholders in deliberative processes, but the belief is that the benefits significantly outweigh the risks.

Even if it narrows the range of disagreement, the relevance condition obviously does not mean that all parties will agree with the specific decisions made. Parties may agree that reasons are relevant but still give different weight or importance to them. As long as fair-minded parties who make the decision and are affected by it can accept that the grounds for it are relevant, however, then even those who do not like or agree with its specific outcome cannot complain that it is unreasonable. This perceived fairness of the process and its outcomes may act as some barrier against self-protective behavior by those who seek ways around limits set in emergencies, but it would be foolish to think that the barrier makes the result secure against various forms of gaming and cheating. Desperate people will behave in desperate ways, but public perception that people are not being treated unfairly will be key to sustaining support for priorities established in an emergency plan.

Fair-minded people will accept many kinds of evidence and reasons as relevant to emergency decisions. These include scientific evidence about effectiveness and safety. In this regard, as noted earlier, it becomes important to be open about the level of evidence about efficacy that pertains to vaccines and antiviral drugs, both for treatment and prophylaxis. It is also important to be open to new evidence and arguments and to revise components of preparedness, including decisions about stockpiling, as new evidence becomes relevant.

What may be hardest for stakeholders to grasp are budget restrictions that are themselves not justified. Here, there is a cultural conflict between the behavior of managers in public agencies, where traditionally budget decisions are not defended with public rationales, and the strong form of openness demanded by accountability for reasonableness.

Accountability for reasonableness adopts a middle path between the poles of “make everything explicit ahead of time” and “let experts muddle through behind the scenes.” The middle path takes the best features from both positions. Like those who advocate explicitness as a condition of fairness, accountability for reasonableness adopts a strong publicity condition. But what is made public need not be a set of principles agreed upon ahead of time; rather, it may be the result of deliberation by experts and other stakeholders about the strongest reasons and arguments supporting a conclusion. Like those who advocate “muddling through” implicitly, accountability for reasonableness draws on the insights that come from examining situations carefully in light of all the evidence.


Emergency preparedness raises many of the same questions of distributive fairness that are raised in other contexts in which resource allocation decisions for health are made. In many of these questions, reasonable people will disagree about how to reconcile concerns about maximizing some measure of aggregate health—including numbers of lives saved in emergency contexts—with concerns about distributing that health fairly. Because consensus is lacking on principles that can resolve those disagreements, EPR planners need to engage in a fair deliberative process to achieve outcomes that are perceived as fair and legitimate. This chapter describes four central conditions that are necessary, if not sufficient, for achieving fairness in such a process. These conditions are compatible with requirements of administrative law and set the stage for appropriate forms of communication with the public about emergency preparedness.


Arras, J. D. (2005). Rationing vaccine during an avian influenza pandemic: Why it won’t be easy. Yale Journal of Biology and Medicine, 78, 287–300.Find this resource:

Brock, D. (1988). Ethical issues in recipient selection for organ transplantation. In Matheiu D. (Ed.), Organ substitution technology: Ethical, legal and public policy issues. Boulder, CO: Westview Press, 86–99.Find this resource:

Daniels, N. (1985). Just health care. New York: Cambridge University Press.Find this resource:

Daniels, N. (1986). Why saying no to patients in the United States is so hard: Cost containment, justice, and provider autonomy. New England Journal of Medicine, 314, 1381–1383.Find this resource:

Daniels, N. (1993). Rationing fairly: Programmatic considerations. Bioethics, 7, 224–233.Find this resource:

Daniels, N. (2008). Just health: Meeting health needs fairly. New York: Cambridge University Press.Find this resource:

Daniels, N., Kennedy, B., & Kawachi, I. (1999). Why justice is good for our health: The social determinants of health inequalities. Daedalus, 128, 215–251.Find this resource:

Daniels, N., Kennedy, B., & Kawachi, I. (2000). Is inequality bad for our health? Boston: Beacon Press.Find this resource:

Daniels, N., & Sabin, J. E. (1997). Limits to health care: Fair procedures, democratic deliberation, and the legitimacy problem for insurers. Philosophy and Public Affairs, 26, 303–350.Find this resource:

Daniels, N., & Sabin, J. E. (2002). Setting limits fairly: Can we learn to share medical resources? New York: Oxford University Press.Find this resource:

Kamm, F. (1993). Morality, mortality: Death and whom to save from it, Volume 1. Oxford: Oxford University Press.Find this resource:

Levine, C. (2009). The Seattle God committee: a cautionary tale. Health Affairs Blog, November 30. Available at

Murray, C. J. L., & Lopez, A. D. (1996). The global burden of disease. Geneva: World Health Organization.Find this resource:

New York State Workgroup on Ventilator Allocation in an Influenza Pandemic. (2007). Allocation of ventilators in an influenza pandemic: Planning document. New York State Department of Health/New York State Task Force on Life and the Law. Available at

Rawls, J. (1971). A theory of justice. Cambridge, MA: Harvard University Press.Find this resource:

Sen, A., & Williams, B. (1982). Utilitarianism and beyond. Cambridge: Cambridge University Press.Find this resource:

Shunemann, H. J., Hill, S. R., & Kakad, M. (2007). WHO rapid advice guidelines for pharmacological management of sporadic human infection with avian influenza A (H5N1) virus. Lancet Infectious Diseases, 7, 21–31.Find this resource:

Sidgwick, H. (1907/1966). The methods of ethics (7th edition). New York: Dover.Find this resource:

World Health Organization Europe. (2002). The European health report 2002. Copenhagen: WHO Regional Office for Europe. WHO Regional Publications, European Series, No. 97. Available at

World Health Organization Regional Office for Europe. (1999). Health 21: The health for all policy framework for the WHO European Region. Copenhagen: WHO Regional Office for Europe.Find this resource: