key: cord-0050304-3robighd authors: Weinstock, Daniel title: A Harm Reduction Approach to the Ethical Management of the COVID-19 Pandemic date: 2020-08-30 journal: Public Health Ethics DOI: 10.1093/phe/phaa026 sha: c2584bf44a09ecd1ee803b4856353a926ec04604 doc_id: 50304 cord_uid: 3robighd The post-confinement phase of the COVID-19 pandemic will require that governments navigate more complex ethical questions than had occurred in the initial, ‘curve-flattening’ phase, and that will occur when the pandemic is in the past. By looking at the unavoidable harms involved in the confinement and quarantine methods employed during the initial phase of the pandemic, we can develop a harm reduction approach to managing the phase during which society will be gradually reopened in a context of managed risk. The principles that are at the heart of such an approach include a reckoning with all of the harms involved in policy choice, including harms that might be given rise to by policy implementation itself; a focus on the harms to which already vulnerable populations are susceptible; and a strong preference for policies that economize on the use of prohibitions and of coercive state enforcement, and that instead emphasize the agency of citizens in realizing health-promoting behavior change. This framework is applied to a policy proposal that has been discussed in policy circles in a number of countries, that of immunity ‘passports’, and to policies that emphasize the creative use of space and time to achieve physical distancing goals. The initial phase of the COVID-19 pandemic arguably required draconian measures on the part of governments. They needed to act swiftly and decisively in order to avoid having the spread of the virus overwhelm health care systems, and lead to a large number of avoidable deaths. The final phase of the pandemic will come when the virus is no longer a threat to public health. This phase might come about as a result of the discovery and global administration of a safe and effective vaccine, as a result of the achievement of 'herd immunity' that would come about as a result of a sufficient proportion of the planet's population having been exposed, or of the discovery of effective treatments that reduce the mortality and morbidity associated with COVID-19 to an 'acceptable' level, say a level on a par with that associated with the seasonal flu. The intermediate phase that many jurisdictions are entering now, which is one in which, temporarily at least, the transmission of the virus, though not entirely extinguished, does not pose a threat to the carrying capacity of local health systems, poses singular policy challenges. Once more draconian measures have successfully 'flattened the curve', and once we have put in place effective measures of testing, contact tracing and isolation to keep the infection curve from rising again, we will need to put measures in place that will allow us to emerge from confinement-to engage again in economic, cultural, familial, sexual activity and so on-while keeping risk at a manageable level. This is arguably the most complex of the three pandemic phases, in that it will require a flexible mix of policy tools (as opposed to the comparatively simplistic tool of confinement arguably appropriate during the first phase, and as opposed as well to the abandonment of mitigation measures that we will arrive at if and when the public health threat posed by the virus is entirely eliminated). Though some optimistic projections suggest that a vaccine might be ready for mass administration as early as September 2020, it is much more likely that this intermediary phase will last for a few years. The worst-case scenario, in which it is discovered that it is impossible to acquire long-term immunity either through exposure to the virus or through a vaccine, and in which treatments only managed to mitigate its symptoms somewhat, is one in which this new level of risk would come to constitute the 'new normal' (Tidman, 2020) . The policy response appropriate to this phase will have to be responsive to the evolution of our knowledge about the virus and its spread. For example, there is still a dearth of understanding of how the virus affects doi:10.1093/phe/phaa026 V C The Author(s) 2020. Published by Oxford University Press. Available online at www.phe.oxfordjournals.org children. As our knowledge expands, this will obviously have an impact on child-specific policies, to do, for example, with schools (Zhang, 2020) . But policy will also have to be guided by a defensible ethical framework. As we move forward from confinement, we will have to craft policies that both achieve our public health goals with respect to the virus, and that embody-or at the very least that mitigate derogation from-core normative constraints on policy. My contention in this article is that such a policy approach should be inspired by the philosophy of harm reduction. Harm reduction has been thought of by many of its early exponents as a practice more than a theory, a practice rooted in a non-judgmental approach to drug use, one that eschews the goal of abstinence and the accompanying methods of criminalization and of enforcement of prohibitions (Canadian Mental Health Association, 2020). But a certain number of principles can in my view be inferred from this practice, one that might inform a broader range of policy domains (Weinstock, forthcoming) . I will begin (I) by outlining the harms relating to confinement and quarantine, which in many jurisdictions has been the primary policy tool used in order to attempt to 'flatten the pandemic curve', that is to bring the rate of infection down to a level that does not outstrip the carrying capacity of local health care systems. Doing so will allow me (II) to identify the general principles that constitute the core of a harm reduction approach. I will then (III) apply these principles to a policy tool that has been much discussed in recent weeks, that of immunity 'passports' or 'licenses'. I will finally (IV) argue that the principles inherent to a harm reduction approach call for a more creative design of policy tools, ones that make maximal use of space and time in order to achieve policies that are most compatible with harm reduction. I will for the purposes of this article assume that strict confinement and quarantine were the appropriate tools to respond to the initial phase of the pandemic. Though no jurisdiction adopted exactly this mix of policy devices, the main ingredients of the policy mixes that were adopted in many places to respond to the onset of the pandemic included, in various degrees of stringency, the following: 1. Substantial restrictions on freedom of movement. At the limit, this involved in some jurisdictions the prohibition on leaving one's place of residence. In others, residents were prohibited from venturing more than a certain distance from their homes, and/or had to be in possession of a form in which they indicated the specific purposes, among a small range of permissible options (acquiring food, medical need, exercise). 2. Prohibitions on all gatherings of people not cohabiting. 3. Prohibitions on all 'non-essential' commercial activity. 4. Limitations or prohibitions on access to public spaces. 5. Quarantining of persons coming into the jurisdiction from abroad, or even from other administrative regions of the same jurisdiction. 6. Lockdown of residential settings containing persons deemed especially vulnerable, most notably persons above a certain age, often pegged at 70. 7. Enforcement of these prohibitions by means of legal sanction, often substantial monetary fines. Many jurisdictions that employed mixes of the measures listed above saw substantial 'flattening' of their infection curves. Whether less restrictive measures, such as those employed in Sweden, would also have given rise to similar results remains controversial (Angner and Aarhenius, 2020) . This controversy is moreover very difficult to settle, given the difficulty of cross-jurisdictional comparisons of infection rates and of death rates established on the basis of often quite different methodologies, and given disagreements as to the appropriate timeframe to employ in order to arrive at an assessment of the comparative success of different methods in achieving the requisite policy ends of containing the spread of infections (Asahi et al., 2020) . I will assume for the purposes of the present argument that the flattening of the curve of infections was a goal of such overriding importance that it was appropriate to subordinate all other policy goals (even health policy goals) to it, and that confinement and quarantine were uncontroversially the best way of going about it. It is a truism that the adoption of this set of policies was costly. All over the world, unemployment rates soared. Viable businesses large and small were pushed to the brink of bankruptcy in a matter of weeks (Jones et al., 2020) . What is perhaps less of a truism is that the adoption of these policies has led to other costs as well, ones perhaps not as easy to express in monetary terms. Most obviously, perhaps, it has had other health costs (Douglas et al., 2020) . Experts warn of a sharp increase in cases of mental health disorder linked to prolonged confinement and isolation (World Health Organization, 2020a) . What is more, the single-minded focus on COVID-19 has led to other health problems being neglected, because health interventions aimed at remediating them have been deemed unsafe, because health resources have been diverted to the treatment of COVID-19 patients, or because patients themselves have prescinded from making use of health resources for fear of placing themselves at risk of contracting the virus in doing so. Looking further down the causal chain, avoidable mortality, both domestically and abroad, is linked to increases in extreme poverty brought about by the interruption of much economic activity (Galea, 2020) . There are less obvious costs to confinement and quarantine that must also be considered in order to arrive at a precise reckoning of the problems that are associated with this policy approach. First, they make use of prohibition and enforcement. This raises a number of issues that must be disentangled (Stott et al., 2020) . First, in the context of a society that places some normative weight on the liberty of citizens, enforcement-based policies come at a cost, even if it is a cost that is in a wide range of cases balanced out by attendant benefits. Where policy objectives can be achieved without the use of coercive force on the part of law enforcement agencies, there is at least a weak presumption in favor of such measures. Second, there are powerful contingent social forces that have made it the case that increases in the use of coercive power by law enforcement officials have disproportionately targeted already disadvantaged minority groups, particularly racialized minorities (Bain et al., 2020) . Third, the use of prohibitions and sanctions risks having the perverse consequence on the part of many citizens to evade these prohibitions, even when they are not antecedently in disagreement with the aim of the prohibition in question (Pettit, 2003) . Fourth, except in the most extremely authoritarian police states, the capacity and willingness of the state to successfully enforce many kinds of prohibitions are simply insufficient. In the case of the prohibition of gatherings of persons who are not co-habiting, successful enforcement is inconceivable. It would require the law enforcement officials being able to track all movements, or to have 'eyes' in every public-and private-space. The risk of relying on the coercive enforcement of prohibitions is thus that we end up in the worst of all possible worlds, one of unregulated non-compliance. Finally, the use of confinement and quarantine involved significant derogations from core normative considerations. In particular, the imperative to 'shelter in place' has hugely differential impacts on people depending on where they can shelter (The Radical Housing Journal Editorial Collective, 2020). Some people live in homes that afford them private access to the outdoors, and sufficient space to come together with the people with whom they co-habit, and perhaps more importantly to isolate themselves at least temporarily from those with whom they are confining. Others live in much more straitened circumstances that afford none of these opportunities. Considerations of space notwithstanding, some people are lucky enough to be able to 'shelter in place' with persons with whom they have fulfilling (or at the very least non-toxic) relationships, whereas others have been forced for lack of actionable alternatives to 'shelter in place' in toxic, often violent circumstances. To the extent that 'place' is a resource that 'shelter in place' strategies draw on in order to achieve the desired public health goals, they are massively inegalitarian, given how unequally that resource is distributed. 'Stay at home' policies have also been costly from the point of view of the agency and decision-making capacity of certain segments of the population. Mortality and morbidity rates from around the world make plain that persons over 70 are particularly vulnerable to the most severe manifestations of COVID-19. As a result, messaging in many jurisdictions has been to the effect that persons over 70, regardless of their health status, should observe even more rigorous isolation, so as to reduce their risk of infection to as close to zero as possible (Center for Disease Control and Prevention, 2020). Most significantly, for healthy seniors living on their own, this has meant abstaining from seeing their loved ones for extended periods. Now, while there is clearly a public health justification for the more stringent confinement of persons who are at significantly greater risk of being severely adversely affected by the virus, this has clearly given rise to a problematic level of paternalism vis à vis older adults. It would after all not be unreasonable for an older adult to arrive at the conclusion that her well-being, given her preference ordering, is better served by increasing her risk of infection somewhat in order not to have to wait months or years in order to engage in valued activities such as interacting with family members. The situation of persons over 70s who live in collective living arrangements is obviously different. Restrictions on activities of such persons can be justified on the basis of third-party harms rather than paternalistically. But for the large proportion of persons over 70 who live alone, there is clearly normative cost in a policy response to COVID-19 that involves messaging to the effect that they should avoid all contact. Clearly, confinement and quarantine policies come at significant costs of various kinds. Now under the assumptions that I have made arguendo, these policies are justified despite these costs. (Whether they have been justified even when these assumptions are lifted is a question that will have to await another occasion.) Deciding that we have reached the second phase of the pandemic involves deciding that the balance of costs and benefits no longer warrants such a single-minded focus on the restriction of the spread of the virus, and that we can start factoring other considerations into our policy response to it. In this new phase of the virus, the costs associated with confinement and quarantine-based policies begin to loom large. Examining these costs points to a set of principles which, I contend, should constitute the normative framework within which policy should be framed. In the next section of this article, I want to extract and articulate these principles. Putting them together will reveal a significant kinship with harm reduction approaches in other policy domains, most notably drug policy. In this section, I will infer what I take to be the principles that emanate from the foregoing reflection on the costs of the kinds of policies that were adopted by many jurisdictions during the first phase of the pandemic. A first, overarching principle is that in focusing on the elimination of what is perceived to be a bad (call it the primary bad), policy-makers should attend to the other bads that they also have the responsibility to attend to. In particular, they should be attentive to the way in which the tools that they choose to employ in their attempt to combat the primary bad, may themselves give rise to other bads-call them secondary bads-that must be counted in the ledger, no matter how effective they are in combatting the primary bad. This may be particularly difficult to do in the context of modern administrative states, not just because policy-makers are like all human knowers and decision-makers susceptible to cognitive biases that may lead them to focus disproportionately on a particular, sudden and dramatic bad (such as COVID-19), but also because in the context of such a state these distorting biases may give rise to certain government agencies being given a predominant role that at least temporarily makes it difficult for other agencies to gain the attention of decision-makers. The decisionmaking agencies that rise to prominence may understandably be subject to the kind of confirmation bias that will lead them to discount or disregard the kinds of secondary bads that the policy tools they recommend give rise to. Call this the 'all bads principle' (ABP). A second principle, one that distances the approach I am recommending from the kind of straight consequentialism that might be suggested by the first principle sketched above, is that particular attention must be given to the ways in which already disadvantaged and marginalized individuals and communities are affected both by the primary bad that policy aims to eradicate or mitigate, and by the secondary harms that are given rise to by the policy tools that are being considered to effect that eradication or mitigation. This principle reflects the conviction that bads such as the coronavirus emerge into a context of already existing injustice that makes it vanishingly unlikely that its effects will be distributed randomly across the population affected. Poor folk who live in improperly built and maintained houses are far more vulnerable to earthquakes than are the rich, for example. And while it has become something of a mantra to claim that 'the virus does not discriminate', the fact is that it does in multiple ways. First, disadvantaged people are more likely to suffer from some of the conditions that are associated with worse outcomes relating to the virus, and, second, are less likely to receive adequate medical treatment. Measures taken to halt the spread of the virus also have differential effects. Those who are able to work while confining are disproportionately concentrated among the upper-middle class. Socio-economically disadvantaged people are disproportionately represented in the jobs that place them most at risk of contracting the virus-working in grocery stores, for example, or in poorly paid jobs in the health sector. It is unsurprising that, in the jurisdiction in which I live, Quebec, somewhere on the order 65% of confirmed infections are borne by women, and that the main 'hot zones' in the province are in disadvantaged neighborhoods where many of the people who occupy these jobs live (in densely populated conditions which compound the difficulties that disadvantaged groups have in effectively protecting themselves from the virus through social distancing). Call this the 'privileging the disadvantaged principle' (PDP). Third, the inventory of harms associated with confinement and quarantine suggests a pair of associated principles to do with the motivational state on the part of citizens that it is best for policy-makers to attempt to leverage in their pursuit of a policy goal. It is a particularity of some policy goals that they can only be achieved through the modification of the behavior of citizens. The spread of the virus occurs through human interaction. It follows that if it is to be slowed, patterns of human interaction must be altered. Now, one way in which to alter human interaction in order to maximize the likelihood of arresting the spread of the virus is to put an end to all human interaction. It is a nostrum among epidemiologists that if it were possible to freeze everyone in place for 2 weeks, the virus would die out for lack of a path of transmission. The next best thing from a purely epidemiological point of view is to reduce the rate of interaction among humans to the greatest degree possible. Thus, confinement and quarantine. But confinement and quarantine are not viable as a long-term strategy. Above and beyond the normative problems that we have identified with confinement and quarantine in terms of their reinforcement of already existing patterns of disadvantage, such policies are unlikely to meet with the voluntary compliance of citizens over the long, and perhaps even the medium term. It runs against the sociality of human beings, the tendency that at least most people have to want to do things with others, including others with whom they do not reside-work, play, engage in artistic pursuits, have sex not mediated through information and communication technology. (Consider: were 'shelter in place' imperatives to be taken literally, it would mean that for the entire duration of their enforcement it would be against the law for people not presently living together to have embodied sex. Has there been a single period in human history in which there has for all intents and purposes been a law against new sexual relationships?) If this is the case, then confinement can only be effective as a tool to combat the spread of the virus either in the very short term, with respect to which the compliance of citizens if predicated upon their understanding that it is only for the short term, in order to achieve a very clearly identified goal, or on the basis of enforcement through the extensive use of the coercive power of the state. For reasons that have already been sketched, the use of state power to alter behavior has significant downsides, both practical and ethical. The principle that is suggested by these brief remarks is that the measures that are put in place in order to combat the spread of the coronavirus ought to the greatest degree possible to economize on coercive enforcement (ECE). Closely related to this is the principle according to which it is far preferable for the modification of behavior required to achieve the desired policy end to be achieved through informed, willing compliance, rather than through the threat of sanction. It may very well be the case that providing citizens with tools through which they can make their interactions safer, by observing distancing, wearing masks, observing respiratory hygiene principles, and the like, though riskier than confinement, all things equal, ends up achieving the goal of limiting virus spread more effectively when the risk of uncontrolled non-compliance associated with more prohibitionist approaches is factored in. And in any event, providing people with principles through which to manage their interactions safely is more respectful of their agency than treating them like obstacles to the attainment of public policy goals that have to be managed, if necessary through coercion. This is in and of itself a weighty reason to prefer policy options that provide agents with tools with which to lessen the risk associated with certain behavior, rather than attempting to prevent them from engaging in the risky behavior altogether. Call this the agency-respecting principle (ARP). I have abstracted four principles from my discussion of the bad-making characteristics of confinement and quarantine-based policies. I have assumed for the sake of argument that these principles are in the 'curve-flattening' phase of the pandemic outweighed by the overriding importance of limiting the spread of the virus to as great a degree as possible in order to avoid overwhelming the capacity of health care systems to treat those who become critically ill as a result of contracting the coronavirus. What follows from this, however, is that in the second phase of the pandemic, as policy-makers appropriately aim to control the virus while at the same time allowing people to return to a greater level of interaction and contact in order to allow them to trade, to create, to play, to pursue new sexual relationships and so on, these principles will have greater weight in determining what mix of policy tools best allows for the attainment of this more complex set of policy objectives. They will not be systematically outweighed by the single public health objective that limiting the spread of the virus represents. The assumption to the contrary in this phase is that, so long as the spread of the virus does not outstrip the capacity of the health care system, it is justifiable to open up society in ways that will predictably lead to some virus spread in order to allow people to return to work, to play, to creation and, yes, to sex. People must be provided with policy tools that allow them to engage in those activities, albeit perhaps in appropriately altered forms, rather than being told that they cannot. I note in passing that these four principles are plausible interpretations of the working theory that has informed the practice of harm reduction, most notably in the field of drug use, for decades. Like harm reduction as practiced in that field, the set of principles that I am putting forward here do not see abstinence as a desirable policy goal. A fortiori, it does not see the coercive enforcement of abstinence as an attractive policy tool. It eschews coercion more generally, even in the pursuit of some set of regulations that might mitigate the harms that can be associated with drug use. Rather than coercion, it seeks to 'meet people where they are', to provide them with information and tools with which to empower them rather than to manipulate them toward conformity with norms of behavior and policy goals to which they do not consent. Harm reduction is moreover highly attuned to the great harms that are created by coercive attempts at enforcing a prohibitionist agenda (The War on Drugs) and to the way in which those harms are disproportionately visited upon the most vulnerable and disadvantaged members of society. These resemblances in my view establish the appropriateness of establishing a kinship between the approach to the normative regulation of policies aimed at limiting the spread of the novel coronavirus and the approach to the problem of drug use that has coined the term 'harm reduction'. It also speaks to the potential for expanding the reach of the working theory underlying harm reduction to a range of other policy domains. I have begun to discuss how this expansion might occur elsewhere, and will for the purposes of the present article refer the reader to these other writings (Weinstock, forthcoming) . I will also note in passing that the appropriateness of thinking of the policy response to COVID-19 in terms of harm reduction has recently been mooted by population health specialists (Marcus, 2020) . How might the set of principles that I have proposed help to orient us through the thicket of proposals that have been put forward as ingredients of a policy mix that might help us to navigate the complex policy waters of the 'post-confinement' phase of the pandemic? I would like to begin this discussion by considering an example that has been at the center of recent debates, and that is seriously being considered for immediate implementation in at least one country, namely Chile (Bartlett, 2020) . The example is that of 'immunity passports', that would allow those who can through a serological test show that they have been exposed to the virus and thus possess antibodies that would prevent recurrence to access social institutions that would remain off limits to those who had not been exposed. 1 I want to leave aside for the sake of the present discussion the scientific doubts that have led the World Health Organization to the conclusion that immunity passports should not at present be adopted as a means to return society to a greater level of functioning than had obtained under a confinement regime (World Health Organization, 2020b). These doubts have to do with whether it is possible to develop long-lasting immunity to the novel coronavirus, be it through a vaccine or through exposure to the virus itself, and whether it is possible at present to administer a test that does not produce false positives, which would lead some people to believe that they are immune to the virus when in fact they are not. Would the adoption of a regime of immunity passports be justified if these (for now well founded) scientific doubts could be allayed? There is a strong prima facie case in favor of such a device (Persad and Emanuel, 2020) . It is estimated that the proportion of the population that has been exposed to the novel coronavirus may be as much as 10 times greater than the proportion of the population that has actually tested positive. If this were the case, in a city in which 1% of the population had tested positive, as many as 10% of the population might actually possess immunity (Cotter, 2020) . To allow them to return to workplaces (as well as to places of sport, culture and leisure) would actually benefit all of the members of society, not simply those possessing immunity. They would be able to occupy key strategic places in the economy, they could, for example, greatly enhance the security of our food supply and the integrity of supply chains. They could also provide much needed oxygen for cultural institutions that have been devastated by the pandemic. Considering the proposal through the lens of the PDP, however, casts a darker light on the proposal. It does so for a variety of reasons, first, as we have already seen, the virus has in many societies struck the socioeconomically disadvantaged the hardest. Putting aside the residents of care facilities for older and more frail citizens, who have born the greatest burden of suffering and death, those who occupy poorly paid positions in such facilities, and those who often without the benefit of appropriate protective equipment have unavoidably come into contact with potentially infected people on a regular basis in hospitals, in grocery stores, and in other sectors of the physical economy that have remained open have been overrepresented among those who have been exposed to the virus. Immunity passports, though presented as an asset, risks in such circumstances creating a permanent underclass of workers continuing to work at menial jobs servicing the needs of white-collar workers able to work from the relative safety of home. Second, we should worry about the potentially perverse consequences that the implementation of a regime of immunity passports might generate. Those people who have been made most economically vulnerable by lockdown of the first phase of the pandemic would find themselves incentivized either to acquire proof of immunity fraudulently, so that they are able to reenter the workforce even in the absence of any exposure to the virus, or to seek to become exposed so as to acquire immunity. Either strategy would be narrowly 'rational', from a purely economic point of view. And it would, as per the prima facie case in favor of immunity passports that I briefly laid out above, also contribute to the general welfare, by adding one more worker to the workforce. However, PDP, which enjoins the democratic public and policy-makers to consider policy proposals from the point of view of their effects on disadvantaged and economically and politically disadvantaged people, would warn us against such a strategy. PDP would require that disadvantaged people be protected against policies that not only create inequalities, but also trade on already existing patterns of inequality and disadvantage (Cf. Baylis and Kofler, 2020; Wilson et al., 2020) . There is another reason to cast doubt on immunity passports as tools that might help us emerge from the confinement and quarantine phase of the pandemic. Note that the flipside of the permission that such passports would grant holders of immunity is the restriction that would still be imposed upon those who were not in possession of such a passport. Far from allowing us to emerge from the logic of prohibition and coercive enforcement that was the chief characteristic, ethically speaking, of the first phase of the pandemic, a regime grounded in immunity passports would require the enforcement of selective confinement. They would thus also involve a derogation from ECE, the principle according to which policy tools the effectiveness of which is not predicated on coercive enforcement by the state are to be favored. Now, were it the case that policies such as immunity passports were the only way of getting the economy started again, we might need to consider their (duly regulated) use. 'The economy' is after all not just some sort of free-floating set of processes that somehow hover above and outside of people's lives and well-being. Rather, it contributes directly to their well-being, and a paralyzed economy is in no one's best interest. So if it were in fact the case that we needed to institute a system of immunity passports in order to get the economy going, perhaps this would be, all things considered, justified, provided that, I accordance with ABP, we identify the secondary bads that an unregulated use of such a system might give rise to, and put in place a system of rules governing the use of immunity passports that might have a good chance at mitigating the bads predictably associated with them. All this is not to say that proofs of immunity, if they were to become scientifically possible, might not become part of a complex set of policy tools allowing us to navigate the complex risk environment that the deconfinement phase would represent. Indeed, there may be some domains of activity where there are no safe options but to ensure that practitioners have the required immunity, both for their own sake and for that of the persons with whom they interact. Some areas of health care require physical contact, as do some commercial activities such as physical therapy and hairdressing. But to say that there are some areas in which we would be justified in requiring proofs of immunity is not to say that we should, in virtue of the harm reduction framework I am putting forward here, employ them more widely. In the final section of this article, I want to suggest that we should seek out ways of exiting the confinement phase of pandemic response that satisfy the principles that I have briefly articulated above. In my view, this involves making more efficient and creative use of space and time in the way in which we redesign our spaces of work, play and culture. Let us begin by noting something that should be obvious. Confinement is not and end in itself. It is rather the most certain way of ensuring that people will not infect one another. If human interactions are associated with some level of risk, then ban as many such interactions as possible! But the goal is not that of limiting interactions per se, but rather of limiting viral spread. Another way of achieving this goal is not to prescind from interacting, but to interact in ways that minimize the risk of spread. As we now know, this involves such low-tech practices as physical distancing, handwashing, the widespread use of non-surgical masks in public settings, and the like. These tools provide us with the building blocks through which to achieve the policy goal of limiting virus spread not by prohibiting interaction, but by changing the way in which we interact. The question that should be top of mind among policy-makers who want to economize on the coercive enforcement of prohibitions on interaction and maximize the degree to which the limitation of viral spread is compatible with continued manifestations of human agency is how should we use these building blocks to make our practices safe, or at least as safe as possible. What I want to briefly suggest in this final section is that in this pursuit, space and time can be our allies (Weinstock, 2020) . Consider, first, space. One of the reflexes of many public officials in the initial phase of the pandemic was to make spaces unavailable to people, in order to disincentivize them from leaving their homes. But there are myriad reasons why it is reasonable, even in pandemic conditions, for people to feel as if they need to get away from their homes. They need to exercise (for their general health and well-being but also in order to make them more resilient were they to be infected by the virus), they may need to get away temporarily from a psychologically difficult home situation, and so on. Restricting access to parks, to parking lots, to school grounds, means forcing people who will leave their homes regardless unless they are coercively prohibited from doing so into more exiguous spaces, mostly sidewalks, where they are much less likely, try as they might to be able to observe a twometer physical distancing requirement. Gradually, officials in a number of cities around the world have realized that willing compliance on the part of citizens with virusinhibiting practices would be more likely to be forthcoming if they opened up spaces and enjoined citizens to observe social distancing within them, rather than attempting to prohibit access. Thus, Vilnius has turned the entire central part of the city into an open-air terrace allowing people to eat and drink in bars and restaurants while remaining two meters apart from people with whom they do not co-habit (Henley, 2020) . Montreal, one of the hardest-hit cities on the planet, has at time of writing just announced that it would be reserving hundreds of kilometers of roads and streets to pedestrians and cyclists in order to allow people to exercise while observing physical distancing (Remiorz, 2020) . We should be thinking along similar lines as far as our work and study spaces are concerned. Not all jobs can be done remotely, and even those that can may be done better if at least some aspects of them are done in places of work. Now, the challenge of rethinking the use of space in these contexts is that much work and study needs to be done indoors, in spaces that are therefore less expandable than are the outdoor spaces that I just mentioned. Now, there are doubtless creative design solutions that might make some indoor spaces retrofittable in ways that satisfy physical distancing requirements. What is more, any modern city contains huge amounts of unused indoor space. Abandoned factories, sporting facilities that are at best only used a few hours a day (and that in pandemic conditions may be more difficult to use safely for the purposes of team sports), commercial spaces that are deliberately left empty by landlords who under current tax systems may be incentivized to 'write them off'-all these spaces might be mobilized in order increase the amount of space at our disposal to make work, study and certain kinds of indoor play and culture compatible with social distancing norms. Yet even the maximally efficient use of the indoor space at our disposal will at some point hit a physical limit. This is where time comes in. Our capacity to engage in activities that require some level of physical presence and 'in person' interpersonal interaction is constricted not just by our inefficient use of all of the space that we have at our disposal, but also by our uncreative use of all of the time that we have at our disposal to make use of those spaces (Lee et al., 2020) . It may come to be seen by future historians as an oddity of the pre-pandemic phase in human history that we insisted on concentrating as much of the commercial and professional activity that we have into the same, roughly 8-hour period of the day. We have introduced some flexibility into the '9-5' day, allowing workers to adopt slightly different work schedules in order to accommodate family obligations and commuting challenges, but we are far from making efficient use of the full 168 hours that is at least in theory available to us each week. A space that we use from 9 to 5 is actually three spaces if we make use of it for 24 hours. Thus another way of stretching out the indoor spaces in which we interact so as to make them compatible with physical distancing requirements is to, as it were, stretch out time as well. The creative use of space and time to redesign our cities would seem to be called for by ARP, which requires that in formulating policy governments emphasize the willing, informed compliance of citizens in ways that enhance rather than curtail their agency. At the same time, PDP enjoins us to be careful that this 'stretching out' not be carried out in ways that reinforce patterns of disadvantage. We want in particular to avoid that 'graveyard shifts' fall by default to the politically and economically disadvantaged. We need to think of egalitarian ways in which to apportion time. One can imagine a system of lotteries, that would assign work and play times at random, or we could imagine a system in which people express their preferences so as to make this assignment as close as possible to what people actually want. These seem to me to be design problems, rather than problems of principle. The general point that I want to suggest through these short remarks is that we ought in emerging from confinement to privilege policy tools that allow people to do things safely, rather than prohibiting them from doing things for fear that they might be risky. We should do this in ways that do not whether by design or by inattention reinforce patterns of advantage and disadvantage. We should also avoid regimes the effectiveness of which is predicated on the coercive enforcement of prohibitions by the state. The creative use of space and time through which to make our work, play and creative activity safe in the post-confinement risk environment is, for the harm reduction approach that I privilege here, the best way in which to start looking for ways in which to make the management of risk and the enabling of human activity and interaction compatible. I have argued that though governments around the planet that have made use of confinement and quarantine may have been right to do so, given the exigencies of the first phase of the COVID-19 pandemic, these measures came at a cost. Getting clear on what the full range of those costs were allows us to identify principles that we should give weight to as we move into the complex phase of managed risk. I have tried to show that these principles echo the working philosophy that has underpinned the practice of harm reduction, and that these principles have quite a bit of power in assessing the various policy tools that already have been, and will be, at the disposal of policy-makers as they attempt to negotiate the complex policy waters that the post-confinement phase represents. Notes 1. Some of the arguments in the next few paragraphs have been developed in greater depth in Ravitsky and Weinstock (2020a) and (2020b). The Swedish Exception? Behavior Public Policy Blog Benchmarking the COVID-19 Pandemic across Countries and States in the U.S.A. Under Heterogeneous Testing. Mrdrxiv: The Preprint Service for Health Sciences Policing, and the Absence of Health Data. The Conversation Chile's Immunity Passport Will Allow Recovered Coronavirus Patients to Break Free from Lockdown, Get Back to Work. The Washingron Post /20/8daef326-826d-11ea-81a3-9690c9881111_story. html COVID-19 Immunity Testing: A Passport to Inequity Harm Reduction COVID-19 Guidance for Older Adults Coronavirus in Boston: 10% of Bostonians Have COVID-19 Antibodies. The Boston Herald Mitigating the Wider Health Effects of COVID-19 Pandemic Response The Poor and Marginalized Will Be the Hardest Hit by Coronavirus Vilnius to be Turned into Vast Open-Air Café. The Guardian Coronavirus: A Visual Guide to the Economic Impact. BBC News Quarantine Fatigue is a Thing. The Atlantic Want to Make Social Distancing Even More Effective? It's About Time. The Conversation The Ethics of COVID-19 Immunity-Based "Licenses Rational Choice Regulation: Two Strategies Are Immunity Licenses Just? Should Immunity Licenses Be an Ingredient in Our Policy Response to COVID-19 Montreal Reassigns Streets for Pedestrians, Bicycles During Pandemic. The Globe and Mail A Turning Point, Securitization, and Policing in the Context of COVID-19 COVID-19 Ad Housing Struggles Coronavirus Might Be Here Forever, WHO Official Warns. The Independent A Harm Reduction Approach to Physical Distancing Two Varieties of Harm Reduction. Health Care Analysis The Big Debate: Should Those with Immunity Get a COVID-19 Digital Passport World Health Organization (2020a). Mental Health and Psychosocial Considerations during the COVID-19 Immunity Passports" in the Context of COVID-19 Why the Coronavirus Hits Kids and Adults So Differently. The Atlantic This paper has been presented in a preliminary form at a webinar of the Max Bell School of Public Policy at McGill University. The author thanks participants for helpful questions. This paper has benefited from generous funding from the Canadian Institutes for Health Research. None declared.