Introduction
How should governments respond to the threat individuals pose to each other when a dangerous and highly contagious disease is circulating in the population? In 1813, the U.S. Congress authorized the maintenance and subsidized the nationwide distribution of smallpox (variola major) vaccine. Smallpox kills 30 percent of those it infects who are unvaccinated and painfully disfigures and/or blinds many more.Footnote 1 President James Madison appointed physician James Smith of Baltimore, Maryland to be the first—and fatefully, the last—National Vaccine Agent. The program mailed cowpox scabs to applicants, with instructions on how to use the material to inoculate patients against smallpox. In 1822, Agent Smith accidentally mailed live-virus smallpox scabs in place of cowpox scabs. The error caused fifty-two smallpox cases and ten smallpox deaths in Tarboro, North Carolina. In response, President James Monroe dismissed Agent Smith and the U.S. Congress repealed the law authorizing the program.Footnote 2
In 1906, typhoid fever (Salmonella enterica serovar Typhi) struck approximately 3,467 New York City residents and killed 639, an infection fatality rate of 18 percent.Footnote 3 Absent treatment, typhoid fever has a death rate between 12 percent and 30 percent.Footnote 4 Officials tied fifty-three of those cases and three deaths to Irish immigrant Mary Mallon, a cook who allegedly infected her victims unwittingly through the food she prepared.
With “force and plenty of it”—but without a warrant—police arrested Mallon. The state imprisoned her for two years without arraignment, charge, or trial in a three-room island cottage “with gas, modern plumbing and electricity … pleasantly situated on the river bank, next to the church”Footnote 5 and “a dog as her only companion.”Footnote 6 Officials released her only after coercing a pledge that she would take “hygienic precautions.”Footnote 7 Upon release, Mallon allegedly resumed spreading the disease. Officials reimprisoned her on the island, again without a warrant, arraignment, charge, or trial, for twenty-three years until her death.Footnote 8
In the 1970s, as part of the World Health Organization’s (WHO) ultimately successful campaign to eradicate smallpox, foreign doctors sometimes resorted to physical coercion to vaccinate resistant Indian and Bangladeshi locals:
Women and children were often pulled out from under beds, from behind doors, from within latrines, etc. People were chased and, when caught, vaccinated … . Almost invariably a chase or forcible vaccination ensued … . When they ran, we chased. When they locked their doors, we broke down their doors and vaccinated them… . I broke the door down and vaccinated—with a struggle—every member of his family, including the man. He was very angry and told me he was going to initiate a case against me … . In the middle of the night an intruder burst through the door of the simple adobe hut. He was a government vaccinator, under orders to break resistance against smallpox vaccination.Footnote 9
In 2010, United Nations (UN) peacekeeping troops from Nepal unwittingly introduced the bacterium Vibrio cholerae, which causes cholera, into Haiti’s water supply.Footnote 10 By 2019, cholera had sickened 800,000 HaitiansFootnote 11 (8 percent of the populationFootnote 12) and killed nearly 10,000Footnote 13 (0.1 percent), an infection fatality rate of 1.3 percent. Three months into the crisis, as the death toll reached 2,500, rioting against UN peacekeepers led to five shootings and two deaths.Footnote 14 Some Haitians blamed voodoo for the disease’s spread.Footnote 15 Angry mobs lynched forty voodoo priests. A government official reported that the victims “were stoned or hacked with machetes before being burned in the streets.”Footnote 16
A novel, dangerous, and highly transmissible coronavirus disease emerged in late 2019. Within months, COVID-19 had spread worldwide. COVID-19 had an initial (unvaccinated) infection fatality rate of 0.47 percent.Footnote 17 Estimates suggest that by March 2023, COVID-19 claimed 7 million lives worldwide.Footnote 18 Some research suggests that excess deaths may be a multiple of that number.Footnote 19 Again by March 2023, COVID-19 appeared as the underlying (87 percent) or a contributing (13 percent) cause of death on 1.1 million death certificates in the United States. Seventy-six percent of those deaths were age sixty-five or older; 4 percent were younger than age forty-five. Estimates associate the disease with 1.3 million excess U.S. deaths.Footnote 20
Governments around the world attempted to contain COVID-19 and protect their populations from it with coercive measures that ranged from spending tax dollars on research, education, and production of a vaccine to stay-at-home orders, mandatory business closures, and masking mandates. Each measure threatened physical violence against noncompliant individuals. On occasion, when individuals did not comply, those threats turned into actual violence and loss of life.Footnote 21 Many coercive measures spurred fierce debate and occasional violent responses from civilians.Footnote 22
According to Lawrence Gostin:
In January of 2020, we saw a city the size of Wuhan lock down 11 million people and then 20 million in the wider Hubei province… . At the time, the world had never seen such a complete shutdown—at least, not since perhaps the 1918 influenza pandemic and even then the records are not clear on this. Certainly, when I helped the CDC draft the Model Emergency Health Powers Act after 9/11 and the anthrax attacks, I envisaged most of the powers that would be used during this pandemic. But I don’t think I could have imagined that a city the size of Wuhan would be locked down. Wuhan was locked down with very intrusive surveillance and punishment. And I remember at the time, I was saying publicly, you know, this could never happen in a western democracy. But of course, that’s exactly what happened. Paris, Milan, London, New York, San Francisco, and even Delhi—even India, the entire country—locked down in the most dramatic deprivation of freedom, which we couldn’t have even imagined.Footnote 23
On January 13, 2020, German scientists validated the first diagnostic test. Various foreign and U.S. laboratories soon followed.Footnote 24 Foreign regulatory bodies and public health agencies soon approved and began using these tests in conjunction with contact tracing, isolation, and individual-level quarantines.Footnote 25
For several months, the U.S. Food and Drug Administration (FDA) blocked the use of all such tests. The ban rendered public health efforts “useless.”Footnote 26
The FDA granted the U.S. Centers for Disease Control and Prevention (CDC) the first (and for a time the only) regulatory approval in the United States of a COVID-19 test. The CDC contaminated its own lab and the test kits it sent around the country with the novel coronavirus. The CDC’s “tragic”Footnote 27 failure rendered its test kits “useless.”Footnote 28 Some scientists began developing and using effective tests without FDA approval, putting themselves in legal jeopardy.Footnote 29
Months after inflicting these “devastating”Footnote 30 blows to public health, the FDA removed some barriers to testing.Footnote 31 In particular, it allowed states to certify tests themselves, but it prohibited additional states from doing so after October 2021.Footnote 32
Public goods rationale: An externality problem
Containing a dangerous contagion presents a classic positive-externalities problem. Individuals and private organizations will voluntarily take steps (for example, social distancing, masking) to reduce the risk of acquiring or transmitting disease. Such activities produce benefits for and impose costs on the individual actor. In general, individuals will keep taking steps to reduce those risks until the cost of the last unit of risk reduction equals the benefits to that individual.
Here arises the problem. Many risk-reduction activities produce benefits for others. The individual actor may not take full account of those benefits when deciding how much risk mitigation is optimal. If masks primarily prevent transmission from the wearer to others, individuals may not wear them as frequently as they would if the mask-wearer were to capture all the benefit herself.
In the language of economics, the decision of whether to attend a public gathering or whether to go to work when feeling ill does not carry a price that conveys the full social cost of that choice. Likewise, the decision of whether to mask does not carry a price that conveys the full social benefits. Without prices that push individuals to choose the socially optimal course of action, voluntary action will not deliver the socially optimal quantities of hundreds of different infection-containment activities. To the extent that containing a dangerous contagion involves positive externalities, voluntary action will leave humans to suffer theoretically preventable morbidity and mortality.
The lure of coercion
In theory, government can boost production of infection-containment activities to socially optimal levels with coercive measures that correct those pricing errors. This essay focuses on the kind of coercion that involves the threat or use of physical violence to change another’s behavior.Footnote 33 As contemporaneous protests against police brutality sought to highlightFootnote 34 and the U.S. government’s response to those protests often demonstrated,Footnote 35 government achieves its ends through threats and the use of physical violence. “Saving lives in this pandemic is job one,” tweeted New York City mayor Bill de Blasio in defense of social distancing mandates. “The NYPD uses summonses and arrests to do it.”Footnote 36 Even taxes that fund otherwise noncoercive public health measures (for example, information collection, contact tracing, subsidizing vaccine development) involve coercion. Willful failure to pay federal taxes can result in fines of $25,000 plus court costs and imprisonment for one year.Footnote 37 In 2014, the evidently lawful violence the New York Police Department (NYPD) used to enforce New York tax laws was enough to kill Eric Garner.Footnote 38
Coercion can change behavior. Its ability to solve the externality problem is less clear. History instructs that the reality of government public health efforts falls short of the theoretical ideal.
There is nothing about coercion that enables those who wield it to know the optimal level of the behavior—or even the optimal behaviors—they seek to promote. The optimal amount of masking or social distancing, like the optimal price or quantity of bread, is essentially unknowable. It is a function of the aggregate values individuals place on the benefits and costs of those measures as well as the benefits and costs of available alternatives. If infection control were to involve no externalities, market prices would constantly push people in the direction of providing the optimal quantityFootnote 39—as if “led by an invisible hand.”Footnote 40 Externalities mean that the information and incentives people receive may not always push voluntary efforts toward the optimal quantity.
The presence of externalities does not imply that government action will push toward optimality. Government faces information and incentive problems of its own. In the absence of a price mechanism, government goes where political imperatives demand. In theory, that destination could be optimality or well beyond it. Even amid externalities, there still exists a point beyond which the costs of infection control exceed the benefits. The CDC has never recommended wearing a mask to bed or closing grocery stores, for example.
One reason government action may not push toward optimality is that coercion introduces additional costs into the calculus. Coercion can itself spread COVID-19. An arrest is not only coercive but also an intimate act that requires police to come in close proximity to the arrestee.Footnote 41 Coercion threatens health in even more direct ways. Every arrest is a violent act that carries a nonzero risk of physical harm. In rare cases, arrests lead to serious injury or even death.Footnote 42 Using police to enforce public health measures can further threaten the public by “steer[ing] officers away from fighting crime.”Footnote 43
Coercion can also threaten social trust and cohesion. A 2004 survey found that 90 percent of African-Americans supported quarantines to contain a serious contagion, unless police would have the power to arrest violators, in which case support fell to 33 percent.Footnote 44 In May 2020, African-Americans accounted for 24 percent of New York City’s populationFootnote 45 but 66 percent of the people the NYPD arrested for violating social distancing measures. In Brooklyn, the figure was 88 percent.Footnote 46 Some confrontations between police and people of color became unnecessarily and even gratuitously violent.Footnote 47 Coercion can erode social cohesion if only because some people believe it is morally wrong for government to coerce people in that way.Footnote 48
Coercive public health measures can even push beyond optimality to the point of being self-defeating and increasing the amount of violence in society. Voters in many states responded to coercive public health measures by stripping powers from their state’s public health officials.Footnote 49 In extreme cases, civilians have threatenedFootnote 50 or exactedFootnote 51 retaliatory violence against public health officials.
Limitations of the externality model
While theoretically elegant, the externality model presents limitations. One is a conceptual and practical challenge that may render the externality model too difficult for policymakers and political processes to manage. The efficiency- or social welfare-maximizing quantity of any given public good is unknown and unknowable. Unlike as happens with private goods, policymakers have no “invisible hand” to guide them toward the optimal quantity of public goods. The push and pull of politics is a poor substitute—one might say no substitute—for the price mechanism because the political system allows participants to exaggerate the benefits of policies they support and the costs of policies they oppose. As Paul Samuelson explains: “One could imagine every person in the community being indoctrinated to behave like a ‘parametric decentralized bureaucrat’ who reveals his preferences by signaling in response to price parameters … to questionnaires, or to other devices. But … by departing from his indoctrinated rules, any one person can hope to snatch some selfish benefit.”Footnote 52 Government officials, moreover, do not pay the full price of their decisions. Without a self-correcting price mechanism to guide them and unable to know the extent to which the information they receive about efficiency or social welfare is the result of strategic behavior by self-interested groups, policymakers attempting to solve the public goods problem are groping in the dark.
Another practical limitation is that those who teach and invoke the externality model tend to ignore the coercion inherent in government public health interventions and the effects of such coercion. This limitation is not inherent. An externalities model can theoretically account for whatever additional (great or small) costs coercion introduces. Government officials who invoke this model, however, face strong incentives to obscure the coercive nature of the interventions they propose and the costs of such coercion. In practice, they behave as though the only people engaging in costly behaviors are the civilians whose behavior they hope to change.
These limitations can systematically lead policymakers to intervene even when doing so is not welfare enhancing. A municipality that candidly announces that police will begin slamming young men to the sidewalk for not masking—and that those young men will overwhelmingly be black—is a municipality that would receive more information about the costs of such coercion and would less often enact a mask mandate.
A violence-minimization model
The remainder of this essay proposes a simpler, more parsimonious, more manageable model to guide policymakers during public health crises. In brief, rather than focus on correcting whatever externalities policymakers imagine may exist across countless behaviors that affect disease transmission, policymakers should adopt the rule of acting only to the extent that doing so would minimize the total amount of violence in society. In addition to its relative simplicity, this model has the benefits of showing greater respect for the dignity and autonomy of all individuals, encouraging greater trust of public health officials (and therefore greater compliance with public health measures), and potentially closer approximations of the optimal quantities of positive-externality behaviors.
Public health concerns situations where one person’s decisions affect or pose a risk to the health of another without the other’s consent. Nonconsensual transmission of a dangerous pathogen is an inherently violent act. A virus or bacterium can be as deadly as a bullet or a blade. To infect another with a deadly pathogen, even if unintentionally, is to put their life at risk. To expose them to that risk without their consent is an act of violence.Footnote 53 Another way to describe a pandemic is a period when millions or billions of people simultaneously do violence to others.Footnote 54
Pathogens are not the only dangers in a pandemic. Humans intuitively recognize that the transmission of a dangerous pathogen is an inherently violent act. When that violence reaches a certain point, humans respond with violence, as in Haiti’s 2010 cholera outbreak. (One should not entertain for a moment the idea that there is anything special about Haiti.) Humans also respond by channeling violence through the state. During the COVID-19 pandemic, government agents slammed young men to the sidewalk for not wearing masks and jailed civilians for entering prohibited areas.
A respect for human equality and autonomy countenances some government public health activities. Government’s role and legitimate scope in a public health emergency are the same as at all other times: to protect individual rights by using its coercive powers to minimize the net amount of violence in society. Just as it is morally permissible for government to use the least amount of coercion necessary to protect us from those who would harm us with guns and knives, it is permissible for government to minimize the amount of violence surrounding the transmission of dangerous pathogens. Government action can both discourage civilians from initiating violence (for example, transmitting a disease) and from responding to such actions with further violence (for example, angry mobs, lynchings) by creating the alternative norms of preventing and responding to such violence through more open, democratic, and less violent processes. The tools government uses to protect us during a pandemic—for example, research, education, contact tracing, and quarantines—may be different from the tools it uses elsewhere. The aim of such public health activities nevertheless is or should be the same: to minimize the net amount of violence humans do to each other.
An important corollary is that—contra de Blasio—“job one” in a pandemic should not be to save lives per se but to minimize the incidence of violence. The idea of elevating autonomy (equivalently, the minimization of coercion) over saving lives is foreign neither to law nor medical ethics. Such is the purpose of the doctrine of informed consent. To prioritize life or health over autonomy would sanction government forcing patients—including not only COVID-19 vaccine resisters but also the elderly, Jehovah’s Witnesses, Christian Scientists, and anyone with a living will or “do not resuscitate” (DNR) order—to take any medicines and to submit to any medical procedures that would lengthen their lives or improve their health. It would allow forcible vaccination of individuals against their wills for any modicum of reduced risk, no matter how small, toward others.
In the language of the externalities model, a violence-minimization rule takes nothing away from the economic good of greater health. Rather, it assigns overriding weight to the economic good of autonomy and overriding negative weight to the economic “bad” of coercion. Prioritizing autonomy and violence-minimization can improve government public health efforts. To the extent those efforts disregard autonomy, they can engender distrust and even resistance among civilians who see their own government as denying their equality by robbing them of autonomy. Prioritizing saving lives over autonomy can thus cause—and has caused—government public health efforts to backfire.
The existence of a dangerous pathogen does not give government officials carte blanche to take whatever steps they think might combat disease any more than murders justify all actions by police that might catch a murderer. To be a legitimate use of government power, public health activities must maximize autonomy by minimizing the sum total of violence, including the transmission of dangerous pathogens as well as civilian and government responses to that danger.
Ethicists typically argue that government public health activities must be effective, proportionate, essential, and the least restrictive means of achieving a given goal.Footnote 55 The imperative to reduce the amount of violence in a public health crisis requires government activities to meet criteria that are similar to but different from these in important respects.
Reduce net coercion
First and foremost, any government public health activity must prevent more violence than it introduces. If the government collects taxes to finance an education campaign about how to reduce transmission of a pathogen, the campaign must produce a greater reduction in coercion (in terms of transmission of the pathogen) than it creates (by threatening individuals subject to those taxes with fines, arrest, and prison). The more dangerous the pathogen and the more effective the public education campaign, the greater the reduction in violence—and the higher a tax the campaign can justify. The greater the cost of the tax (see below) and the more coercion government uses to collect it, the less likely the education campaign will result in a net reduction in violence.
Minimize total coercion
A net reduction in coercion is necessary to justify government action. But it is not sufficient. To be a legitimate use of government power, public health activities must also minimize the total amount of violence in society.
Suppose that absent government action, a given pathogen would produce fifty units of violence. A government public health activity that generated forty units of violence yet eradicated the pathogen would produce a net reduction in violence. So far, so good. The forty-unit activity would not be morally permissible, however, if a ten-unit activity could also eradicate the pathogen. The availability of the ten-unit strategy means the forty-unit activity would introduce thirty unnecessary units of violence.
The preceding criteria encapsulate (or at least approximate) the ethicist’s directives that government action be effective, proportionate, essential, and minimally restrictive. A government activity would have to be effective to satisfy a reduction-in-violence criterion because to do the latter requires reducing transmissions or the severity of illness. The requirement that the activity minimize total coercion satisfies the proportionality and least restrictive means requirements. Finally, an activity that minimizes net violence satisfies the requirement that the activity be essential.
Public health activities vary dramatically in the amount of coercion they employ. The following are categories of public health activities in ascending order of coercion:
-
• Do nothing. When government officials do nothing or “simply monitor the current situation,”Footnote 56 they introduce no coercion beyond the taxes that finance such monitoring.
-
• Information/coordination. The least coercive interventions involve government gathering and transmitting information to the public about how to reduce the risk for themselves and others. They involve no coercion beyond a small tax burden because they leave civilians free to choose how to use that information. In 2022, the entire CDC budget of $16 billion accounted for 0.3 percent of federal outlaysFootnote 57 and 0.06 percent of U.S. GDP.Footnote 58
-
• Recommendations. Government goes a step further when it recommends that civilians change their behavior. Such recommendations can become coercive even if government officials do not initially attach penalties for noncompliance and even if they assure the public that compliance is voluntary. If noncompliance with official recommendations leads to harm, courts may adopt the recommendations as the standard of care in tort cases, thereby rendering coercive what was officially voluntary.
-
• Subsidizing behavior. Next, government can subsidize activities it wishes to effect, such as working from home or developing and taking vaccines. By December 2022, “[t]he federal government ha[d] spent more than $30 billion on COVID-19 vaccines … incentivizing their development, guaranteeing a market, and ensuring that these vaccines would be provided free of charge to the U.S. population.”Footnote 59 As with education campaigns, the only coercion these subsidies introduce comes from the taxes that fund them. When government imposes new conditions on existing subsidies, such as when Medicare threatened to withhold subsidies from hospitals where workers did not vaccinate,Footnote 60 it introduces no additional coercion.Footnote 61
-
• Financial penalties. More coercive than subsidies to encourage socially beneficial behavior are financial penalties for socially harmful behaviors. Public health mandates typically give individuals the option of either complying with the mandate or paying a fine. Such penalties employ the same type of coercion as tax collection. Such penalties also include the implicit penalties that befall the noncompliant when government offers those who engage in certain behaviors a deduction or credit against existing taxes. Because financial penalties target specific individuals and seek to coerce certain behaviors rather than merely raise tax revenue, civilians are likely to see them as more coercive and a greater threat to autonomy than are the broad-based taxes that finance education campaigns, official recommendations, or even direct subsidies.
-
• Invasive coercion. In extreme cases, government may physically coerce individuals to perform the behavior government desires, such as forcible stay-at-home orders, curfews, quarantining, or vaccination.Footnote 62
To minimize the amount of violence in society, government officials must begin with the least coercive category and work their way down this list. Officials may move to the next category only after exhausting every option in previous categories. Crucially, officials must stop at the first intervention that would not clearly reduce the net amount of coercion.
Measure at the margin
To ascertain whether an intervention would reduce the net amount of violence in society, officials must consider the effects of that intervention at the margin. Even when government does nothing, the private sector takes steps to minimize violent transmission of pathogens. In early 2020, many people stopped traveling, many businesses sent their workers home or shut down, and many individuals began wearing masks and practicing social distancing before government told them to do so. Policymakers must evaluate only the additional benefits and costs that an intervention would produce, given what civilians and other government interventions are already doing.
Marginal analysis requires steps that may seem alien to policymakers. Even when government intends to reduce coercion at the margin, for example, it introduces coercion inframarginally. A vaccine mandate does not get credit for the reduction in transmissions that comes from those who would have vaccinated anyway. However, a vaccine mandate does coerce those who would have vaccinated anyway; it deprives them of the freedom to make that choice themselves. Moreover, the marginal benefits of a government activity fall to the extent that it crowds out private violence-reducing activities. Suppose a vaccine mandate prods 30 percent of the population to vaccinate but leads 2 percent of the population who otherwise would have vaccinated not to do so. Perhaps those 2 percent wish to protest the mandate or they believe they no longer need to vaccinate because the mandate will produce herd immunity. In that case, because the mandate crowded out noncoercive vaccinations, it deserves credit only for boosting vaccination rates by 28 percentage points. Finally, many policies become more coercive at the margin. The twelfth week of lockdown is more coercive than the first week; raising the income-tax rate from 40 percent to 50 percent is more coercive than raising it from 10 percent to 20 percent; and so on.
Some coercive COVID-19 policies reduced the consumption of things that extend life (for example, medical care) or improve quality of life (for example, the arts, romance, socializing, sports, watching children play, mobility, worship) to the point where they triggered sometimes violent civilian resistance. Those responses suggest that policymakers may have pushed some of those interventions to the point where they increased violence at the margin.
Measure all margins
Policymakers must also take account of the effects of public health interventions at all margins, including the government’s ability to conduct successful public health interventions in the future. One crucial margin is social cohesion and trust in government. Public trust is essential for funding and executing successful public health interventions. Agent Smith wrote President Madison in 1809 that he believed a federal agency could eradicate smallpox, but only “if the Confidence of the people can be preserved.”Footnote 63 Government officials did not directly add to the amount of violence in society when they flipped from discouraging masking to encouraging it. However, issuing dramatically conflicting recommendations within a short span of time reduced the public’s confidence in government, which reduced government’s ability to use public health tools to reduce violence in the then-current crisis and future crises. Politicization of public health matters can cause half of a polity not to trust public health officials.Footnote 64
A consideration that should weigh heavily on public health officials is the impact of their actions on minorities. Government invariably enforces public health measures more strictly against minority and immigrant communities that lack political power, which leads to distrust and noncompliance among entire communities. According to Mary Beth Keane, “quarantines in our nation’s history have always meant taking a group of people, usually in the lowest income bracket or of the same minority group, and placing them apart. Looking at a history of quarantine means looking at a history of discrimination.”Footnote 65 In 1901, New York City police raided immigrant homes in the middle of the night, “administered smallpox vaccinations by force,”Footnote 66 and tore symptomatic children from their parents.Footnote 67 The practice led immigrant parents not to cooperate with police and public health officials, just as violence against Indian and Bangladeshi villagers led other villagers not to cooperate with (often white, English-speaking) WHO doctors.
Government officials (and civilians) must consider the long-term effects of political and legal precedents that public health activities set. Will a given public health intervention lead to mission creep that distracts public health agencies from their mission? The eradication of smallpox may have been a great success, but how might less enlightened government officials use those powers in the future? Might government officials hesitate to declare that a public health emergency has ended because they are reluctant to lay down the additional powers they wield while emergency declarations are in effect?Footnote 68
Subsidiarity
A violence-minimization framework is easier to articulate than to implement. The considerable uncertainty about the foregoing questions and trade-offs is a function of both the uncertainties inherent in health and medicine and in trying to maximize the “good” of autonomy without benefit of a functioning price mechanism.
It is because of these uncertainties that a key criterion of a violence-minimization framework is subsidiarity. That is, government responses to public health crises should always come from the most local level possible.
Nationwide or even multinational public health interventions may sometimes be beneficial because a contagious disease may spill over from one locality or one state to another. Smallpox eradication is an example of a sensible global public health strategy (see below). When case rates are dramatically lower in some states or in some localities than in others, however, it makes little sense to have a national policy on masking or closures of schools and businesses.
Subsidiarity reduces the likelihood of government officials introducing coercion where it is not necessary. State and local officials generally have better information about their constituents’ needs and fewer forces diverting them from those needs than do officials of the national government. They also have greater incentive than national governments to tailor public health policies to minimize violence in their state. By contrast, nationwide policies can introduce unnecessary coercion in some states while simultaneously failing to address preventable private coercion in other states. A clear example of subsidiarity and its benefits is the FDA’s (perhaps reluctant) decision to allow states to certify COVID-19 diagnostic tests themselves.
Subsidiarity can also lead to better public health interventions. As states and localities experiment with different interventions, they can learn from each other’s successes and failures. Nationwide interventions do not create the same opportunities for competition and learning.
The rule of law
Another key criterion is that government officials must act within democratic constraints, in particular, law. Democratic and legal institutions constrain government officials and are an essential check against excessively coercive public health measures.Footnote 69
Public health officials frequently exceed and even disregard their legal authority. New York officials imprisoned Mary Mallon “without due process of law, without legal representation, indeed, without even a trial.”Footnote 70 The principal investigator said that “she was held without being given a hearing; she was apparently under life sentence; it was contrary to the Constitution of the United States to hold her under the circumstances.”Footnote 71
When government officials coerce civilians without legal authority, their actions are morally comparable to the Haitian mobs that rioted against voodoo priests. Even if those extralegal actions reduce net violence in the short term, they violate the principle of egalitarianism by asserting arbitrary power over others. They can also be self-defeating, and increase violence over the long term, by reducing trust in government generally and in public health officials.
Humility
A final component of a violence-minimization framework is humility. Government officials should err on the side of not introducing coercion. Even when pathogen transmission is a violent act, government’s attempts to reduce such violence can produce unintentional consequences, including introducing violence where there was none. State laws criminalizing the transmission of human immunodeficiency virus, for example, sometimes punish the victim of transmission rather than the perpetrator.Footnote 72
Indeed, government is as likely to undermine public health and trust in government as it is to promote either. Prior to COVID-19, decades of state regulation suppressed the supply and mobility of health professionals, guaranteed that otherwise feasible telehealth systems would not be in place when the pandemic struck,Footnote 73 suppressed hospital capacity and innovation, and prevented rapid market entry once COVID-19 created shortage conditions.Footnote 74
Federal officials performed no better. They diverted resources from public health to less effective medical interventions,Footnote 75 including repeatedly siphoning money from the federal Prevention and Public Health Fund to subsidize health insuranceFootnote 76; suppressed contactless access to essential medicines (internet purchasing, vending machines, interactive kiosks)Footnote 77; caused more than 7.3 million workers to lose coverage unnecessarily during the pandemic by tying health insurance to employmentFootnote 78; put the most vulnerable workers at higher risk for contracting COVID-19 by encouraging them to return to work for fear of losing their coverageFootnote 79; paid hospitals more when patients acquire preventable infections and less when they avoid infectionsFootnote 80; crowded vulnerable seniors in housing where COVID-19 spread rapidly by subsidizing nursing homes over other forms of long-term careFootnote 81; suspended requirements that hospitals report infection rates once COVID-19 struckFootnote 82; and (again) prohibited effective COVID-19 diagnostic tests when the need was greatest.
Government routinely undermines public trust by deciding countless divisive questions that individuals would otherwise decide for themselves, including how to educate children; whether and what type of health insurance to purchase; whether to enter the country; whether to fund contraception; whether people can possess, exchange, and ingest recreational drugs; whether churches should be subject to stricter rules than bike shops; and so forth. When government arrogates to itself the power to decide such questions, it overrides the values of one tribe or another, alienates large segments of the public, and sows resentment and mistrust.
The institution that will implement the most elegant, coercive public health strategy that scholars can devise will be the same institution that has repeatedly shown such lack of foresight and regard for public health that it left the United States “not only ill-prepared and poorly positioned to deal with COVID-19, but also uniquely susceptible to the spread of this illness.”Footnote 83 Policymakers should therefore exercise humility to the point of erring on the side of intervening too little. Government officials should not assume they have better information and incentives than civilians do. Federal (state) officials should not assume they have better information and incentives than state (local) officials do. They should introduce coercion only when there is clear and convincing evidence as opposed to mere likelihood that an intervention would minimize the level of violence in society.
Applying the violence-minimization principle
Applying the violence-minimization principle to a given public health intervention requires policymakers to answer a series of questions:
-
1. Do government officials have clear legal authority for the intervention?
-
2. How much violence would the intervention prevent, given existing civilian and government efforts to reduce such violence? To what extent would the intervention affect the amount of violence individuals inflict on each other? Would it reduce pathogen transmission? Would it make transmission less dangerous (for example, vaccines or medical treatments)? Would it avert violent reactions or retributions to the threat of transmission (for example, Haiti)?
-
3. How much additional violence or coercion would the intervention introduce? How severe would it be? How many civilians would it affect?
-
4. Would the intervention itself provoke violent responses? Are the behaviors the intervention seeks to produce or the penalties for noncompliance so costly or uneven that civilians would resist, as they have in Brussels and elsewhere?Footnote 84 Would government officials, intentionally or not, inflict disproportionate coercion on certain groups?
-
5. How would the intervention affect less obvious margins? What would be the long-term effects? Would the intervention affect incentives? Or trust in government? How would future government officials misuse the same power?
-
6. Would the intervention result in a net reduction in violence? Is the evidence that the intervention would produce a net reduction in violence clear and convincing? Or does the evidence merely suggest that a net reduction is more likely than not?
-
7. Are there other interventions that would achieve a greater net reduction in violence? For example, intervention at a more local level of government?
Requiring satisfactory answers to these questions would result in fewer yet more effective public health interventions.
The historical examples of smallpox eradication, New York’s imprisonment of Mary Mallon, and select episodes from the COVID-19 pandemic illustrate, at least in hindsight, how to operationalize the violence-minimization framework.
Smallpox
The tremendous violence humans do to each other by transmitting smallpox and the comparatively minuscule cost of violence to eradicate it make this case relatively easy to evaluate. Overall, the WHO smallpox-eradication campaign passes a violence-minimization test with flying colors. Even so, individual acts of coercion during that campaign may fail the test due to marginal effects and long-term consequences.
Smallpox is no joke. When Europeans brought the disease to the Americas, it killed up to 95 percent of some indigenous tribes.Footnote 85 In the twentieth century, it still had an infection fatality rate of 30 percent. Up to another 9 percent suffered eye disease, including blindness.Footnote 86 Many more victims suffered painful pustules and disfiguration. Sadly, “[d]uring the 20th century alone, an estimated 300 million people died of the disease—more than twice the death toll of all the military wars of that century.”Footnote 87
The WHO launched its eradication campaign in 1967. By 1980, smallpox was gone. Particularly when we account for nonfatal injuries, that campaign ended a disease that caused twice as much human-on-human violence as World War I, the Russian Revolution, the Chinese Revolution, World War II (including Hiroshima and Nagasaki), the Korean War, the Vietnam War, the First Iraq War, the Rwandan Genocide, and all other conflicts from 1900 to 1999 combined.
Smallpox eradication relied on government coercion. Taxes funded the deployment of public health professionals to outbreak areas as well as the contact tracing, testing, and ring-vaccination strategies they performed. Ann Marie Nelson reports:
The total cost of smallpox eradication was $315 million between 1967–1980; $30 million was from the United States and the rest from developing countries… . Since the eradication of smallpox in 1977, the U.S. Public Health Service estimates a monthly savings of more than $30 million in material and human resources (cost of production, administration of the vaccine, treatment of adverse reactions, record-keeping and surveillance, immigration control, quarantine, and education).Footnote 88
Imagine a government intervention that employed as little coercion as smallpox eradication did but instead permanently eradicated war. Such a tax would indeed purchase civilization.Footnote 89
Smallpox eradication may be the rare example of a government program that truly and clearly paid for itself. It prevented more violence than it introduced. It reduced government outlays more than it increased them. It even reduced the incidence of government coercion (immigration controls and quarantines).
Even so, the success of smallpox eradication does not justify every coercive act that contributed to it. The eradication effort itself did violence to U.K. residents Janet Parker, who contracted smallpox via laboratory exposure and died, and her mother Hilda Witcomb, who survived.Footnote 90 The coercion and violence that some foreign WHO doctors employed in India and Bangladesh violated egalitarianism and may not have been strictly necessary to achieve eradication; at the margin it may even have led to a net increase in violence by breeding resentment that hindered both smallpox eradication and subsequent public health efforts:
[C]oercion can leave behind a residue of resentment that sours public attitudes toward the next vaccination campaign … . It is also worth considering whether some of the resistance that vaccinators encountered in the villages of India and Bangladesh in 1975 might not itself have been the result of prior half-completed but unsuccessful immunization campaigns in which coercion had played a role.Footnote 91
Mary Mallon
New York officials’ treatment of Mary Mallon did not pass the violence-minimization test. Imprisoning Mallon on North Brother Island for twenty-six years likely reduced the amount of violence in society. Evidence suggests that Mallon spread typhoid to dozens of victims. It further suggests that, had the state not imprisoned her, she likely would have spread it to dozens more. Her imprisonment likely saved lives.
New York officials did not minimize the amount of violence in Mallon’s case, however. It is possible the state could have achieved the same reduction in violence without coercing Mallon to the extent they did. The state imprisoned her long after it released other asymptomatic typhoid carriers. Officials might have stopped further transmissions through some combination of educating Mallon about the disease and how to avoid transmission; teaching her a different, safer, and more lucrative trade; paying her not to work; giving her a job that posed little risk of transmission; and having her check in periodically with public health officials. There is no evidence that officials tried any of these less coercive strategies. They ultimately employed her in a lab on the island—a captive worker. Mallon wrote her attorney: “There is a visiting doctor who came here in October. He did take quite an interest in me. He really thought I liked it here, that I did not care for my freedom.”Footnote 92 Officials also believed that removing Mallon’s gallbladder would have rendered her noncontagious, but she would not consent to the procedure. They were right not to coerce her to submit to that procedure.
New York officials also failed the violence-minimization test because they acted outside of the law. The state wildly violated Mallon’s legal rights. It arrested her without warrant, never charged her with a crime, and imprisoned her for twenty-six years without trial. When Mallon sued for her release, the courts turned a blind eye to these violations. The precedential implications of Mallon’s case are staggering. Mallon’s attorney warned that the power to “clap someone in jail upon the word of some medical man” is a license to violate the rights of “thousands upon thousands of persons.”Footnote 93
Mallon’s case also illustrates that government public health officials, like private individuals, face incentive problems. Some scholars argue that New York officials continued to imprison Mallon out of stubbornness and because she was useful to them as a “lab rat.”Footnote 94 Other scholars say that Mallon’s immigrant status played a part in her captivity.Footnote 95
COVID-19
COVID-19 provides examples of government policies that met the violence-minimization test, policies that did not, and the fluid nature of violence-minimization calculations.
If conventional estimates of costs and benefits are correct, government investments in COVID-19 vaccines also pass the violence-minimization test with flying colors. According to one estimate, the vaccines saved 20 million lives worldwide between late 2020 and late 2021.Footnote 96 If so, they turned what would otherwise have been 20 million lethal assaults into nonlethal assaults. Another estimate suggests that, in the United States, the vaccines averted 18 million hospitalizations and 3 million deaths.Footnote 97 Equivalently, they prevented 3 million lethal assaults and 18 million serious assaults. Again, the U.S. government spent some $30 billion, or roughly $100 per resident, subsidizing the vaccines’ development and distribution.Footnote 98 Given those figures, it would seem that those government subsidies prevented far more coercion than they introduced—all the more so if those U.S. investments also saved lives overseas and reduced spending on compulsory government programs.
The diagnostic-testing debacle of early 2020 quite clearly fails the violence-minimization test. The FDA literally stopped public health professionals from detecting and stopping violent assaults: “The [FDA] did exactly what it shouldn’t have. It limited the diagnostic capacity of this country. It’s insane.”Footnote 99
State governments’ efforts to strip physicians of their medical licenses for discouraging vaccines or prescribing ineffective treatments would also fail a violence-minimization test.Footnote 100 Such efforts first require the creation and maintenance of a system of government licensing of health professionals, which itself requires considerable coercion. Such systems likely introduce more coercion than government could prevent by revoking licenses—which itself constitutes a coercive restraint on the ability to practice medicine—in cases where clinicians dispense advice or treatments that lead to more violent transmissions of pathogens. Even if using government licensing in this manner could reduce more coercion than licensing itself introduces, other interventions—for example, medical malpractice liability or laws that specifically target fraud—could achieve a greater net reduction in coercion.
Other COVID-19 public health strategies fall somewhere in between. Government recommendations that individuals mask, practice social distancing, or work from home likely passed the test as low-coercion coordination mechanisms and means of signaling the seriousness of the crisis. Closing public spaces such as parks and beaches might have passed the test if COVID-19 were far deadlier and more transmissible or when there was some uncertainty on those points. Arresting those who trespassed in public spaces does not pass the test, because such confrontations increased the risk of transmission and other harms. Likewise, closing businesses and schools might have passed the test early in the pandemic (when there was uncertainty about transmissibility and disease severity) and even later (when there were still no vaccines), but not uniformly. Such measures could pass the test in outbreak areas, but not in areas with low case rates.
When effective vaccines became available, they dramatically altered the violence-minimization calculus for all other public health interventions. Masking recommendations, mandates, and school closures that might have passed the test suddenly failed it. The growing and often angry backlash against those measures suggests civilians judged that those measures posed a greater threat to their autonomy than did a virus whose infection-fatality rate had gone from being slightly higher to significantly lower than that of seasonal influenza.
Again, voters in many states retaliated by stripping powers from state public health officials.Footnote 101 The continued pushes for compliance with those measures and multiple rounds of vaccinations may have done long-lasting damage to public trust in state and federal public health officials and their ability to minimize violence in the future.
Potential concerns with and objections to a violence-minimization criterion include that it might permit killing one innocent to prevent the violent killing of five innocents; it could allow the aggregation of many small exercises of coercion to outweigh the violence of killing a few people, and therefore condone the latter to avoid the former; and it would allow a “heckler’s veto” by giving undue weight to strategic retaliatory violence by civilians. Regarding the last concern, it is an odd social-welfare accounting that does not incorporate retaliatory violence. Public health officials should take the potential for such violence into account when weighing the wisdom of intervention, even if it means abandoning low-yield public health interventions. A powder-keg society is one on which public health officials should tread especially lightly.
Conclusion
Public health policy is an expression of how we value other humans. A violence-minimization approach to public health respects the dignity of all individuals by maximizing autonomy.
How exactly to minimize coercion and violence is not something policymakers can calculate with an abacus. Yet it has virtues and advantages over other models. It provides a framework for thinking about public health that maximizes human dignity. It focuses policymakers’ attention on the fact that they themselves introduce coercion and violence into society. It requires policymakers to weigh carefully the one aspect of government public health activities that most breeds distrust. It holds the promise of making government officials more circumspect about violating autonomy. It is perhaps conceptually and functionally simpler than a maximand of efficiency and more likely to arrive at the socially optimal level of public health interventions. Finally, a violence-minimization approach focuses government officials’ attention on the only legitimate reason for government.
Competing interests
The author declares none.
