key: cord-0057975-r25kaiz0
authors: Francis, John G.; Francis, Leslie P.
title: Surveillance for the “New” Public Health
date: 2021-03-18
journal: Sustaining Surveillance: The Importance of Information for Public Health
DOI: 10.1007/978-3-030-63928-0_6
sha: 99ddb529617e476c450292acf6e04a173db66688
doc_id: 57975
cord_uid: r25kaiz0

Traditional forms of surveillance addressed detecting outbreaks of dangerous contagious disease, case finding and contact tracing, and identifying environmental hazards. But public health today goes far beyond interpersonal contagion and environmental dangers to address patterns of ill health among the population. Increased rates of conditions such as diabetes or substance use disorders are called “epidemics” to be addressed by public health interventions including education, environmental design, economic incentives, regulation, and even prohibition. This chapter explores the ethics of this so-called “new” public health, arguing that paternalistic interference with behavior is justified only to prevent serious harm. Much information collection, however, aims not to protect individuals but to understand health or health disparities at the population level and thus is not properly regarded as paternalistic. The chapter also argues that interferences such as the requirement to wear masks in a pandemic are mistakenly perceived as protecting individuals against their own choices and are better viewed as protecting others from the possibility of harm.

In the last half of the twentieth century, public health increasingly saw itself as having the broader mission of maximizing human health and well-being, a mission aligned with the World Health Organization's definition of health as "a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity" (WHO 2019). Perhaps not accidentally, the time between the end of the Second World War and the advent of the HIV/AIDS epidemic around 1980 was an era of great optimism about the conquest of infectious disease through antimicrobial therapy (Wiley 2012, p. 220) . It was also an era of increasing recognition of the role played by social factors in influencing health-that is, the social determinants of health. In the U.S., optimism about the ability of government to promote overall welfare was at its height during the administration of President Lyndon Johnson, who initiated pursuit of the "Great Society" in 1964.

To further the mission of improving population health broadly conceived, public health directed attention to addressing rising rates of chronic or noncommunicable diseases (NCDs) such as type 2 diabetes, heart disease, arthritis, chronic obstructive lung disease, and many cancers. Policy priorities were increased governmental support for social safety nets, disease prevention, and primary care. The Declaration of Alma-Ata issued by the 1978 International Conference on Primary Health Care invoked the WHO definition of health to urge governments to formulate policies and plans of action to sustain primary health care as part of a national health care system (WHO 2019a) . Importantly, the Declaration asserted the responsibility of government for the general health and well-being of society Varivikova 2010, 2001) These governmental efforts were seen as interconnected with efforts to address individual behaviors judged significantly deleterious to health, such as smoking, poor diet, alcohol consumption, and lack of exercise. Improvement of population health overall thus was judged to require combined efforts by governments and individuals.

This recasting of the mission of public health to improving overall population well-being was dubbed "the new public health." The WHO's Ninth General Programme of Work (1996 Work ( -2001 recognized the new public health as mobilizing communities, public health, and political leaders around concerted efforts to improve health and well-being overall (Ncayiyana 1995) . The Programme also realized, however, that these efforts were encountering resistance to both the language and the scope of the new public health. Resistance took political form, as many countries drew back from supporting or funding increased roles for their governments. Perhaps as a rhetorical strategy, or perhaps as a matter of fundamental changes in the understanding of the etiology of poor health, the new public health did not leave images of contagion behind. Obesity or diabetes were described as "epidemics" for which alarms were to be sounded (WHO 2003) . Today, opioid use and the deaths that have resulted are officially labeled an "epidemic" and a "public health emergency" by the US government (DHHS 2019).

The extent to which the new public health has become the norm is also illustrated by the United Nations sustainable development goals for 2030, which lists "good health and well-being" as Goal 3. Among Goal 3 targets are not only ending the epidemics of AIDS and tropical diseases and reducing deaths and ill-health from hazardous pollution, but also strengthening prevention and treatment of substance abuse and halving the number of deaths and injuries from road traffic accidents. The target of supporting development of vaccines and medicines for conditions primarily affecting developing countries includes both communicable and noncommunicable diseases alike (UN 2015) . The WHO Global Action Plan in support of these sustainable development goals accelerates efforts to address the social determinants of health (WHO 2019b) . While the WHO does not have enforcement authority in the sense that it can act as a health police for the new public health, it does establish priorities for observation and funding that are in turn meant to be influential on the priorities of states parties. In the United States, Healthy People 2020 Leading Health Indicators include nutrition, physical activity, and obesity; substance abuse; and tobacco use (ODPHP 2020) . Healthy People 2020 also devotes a separate topic to the social determinants of health, with a leading indicator of students entering 9th grade graduating from high school within 4 years. Assessing progress on new public health efforts requires surveillance of a wide range of factors from weight to education, far beyond the communicable diseases and environmental subjects of earlier public health.

Primary surveillance goals of the new public health include the incidence and prevalence of NCDs such as diabetes; the frequency and distribution of contributing behaviors such as smoking, diet, and exercise; and the relative distribution of social factors such as education, housing, employment, or economic inequality. Put most generally, the new public health widens the scope of surveillance beyond the incidence and prevalence of contagious and toxic disease to a broad range of social factors that may affect health and well-being. For obesity, these factors might reach beyond weight to include food deserts, safe playgrounds, and cultural norms. For opioids, they might include chronic pain, disability, limited education, unemployment, and other factors implicated in rising rates of "deaths of despair" (Case and Deaton 2015) .

At the international level, WHO structures its data collection in light of the UN sustainable development goals. It lists seventeen different types of health and healthrelated target indicators, including as major categories NCDs, substance abuse, road traffic injuries, and tobacco control (WHO 2019c) . WHO (2019d) clusters major non-communicable diseases into four broad areas: cardiovascular disease, cancers, diabetes, and chronic respiratory disease. According to WHO, in 2016 these four areas accounted for 71% of all deaths worldwide. In low-and middle-income countries, 48% of deaths from these conditions are premature in the sense that they occur before age 70. WHO contends that these conditions are linked to modifiable risk factors such as tobacco, harmful use of alcohol, unhealthy diet, insufficient physical activity, overweight/obesity, high blood pressure, raised blood sugar, and high cholesterol. It estimates that 80% of these risk factors are modifiable through public policy changes. To take examples of how these risks might be modifiable, fewer than half (46%) of WHO member states have policies controlling alcohol use by age and licensing requirements, only 10% of people live in countries with tax rates on cigarettes that WHO regards as sufficient, and 39% of both men and women in the world are obese or overweight. Dementia is another type of NCD drawing attention from WHO; WHO launched a global monitoring system in 2017 to track progress on national policy, risk reduction, and dementia care and treatment (WHO 2019e).

In the US, surveillance for NCDs, related behavioral factors, and the social determinants of health is a complex mix of efforts by the federal government, state governments, and other sources. The US CDC is the primary agency responsible for surveillance at the federal level; it relies largely on information supplied by agreement with the states and on surveys that it conducts, including the Behavioral Risk Factor Surveillance System (BRFSS), the Youth Risk Behavior Surveillance System (YRBS), the National Health Interview Study (NHIS), and the National Health and Nutrition Examination Survey (NHANES). In comparison to the WHO figures, infectious diseases account for a very small percentage of U.S. mortality, only just over 5% according to the latest available data (Hansen et al. 2016) , although this may change with COVID-19. Influenza and pneumonia accounted for over 75% of these deaths from infection. Death rates from HIV/AIDS rose and then fell over the period between 1980 and 2014, largely due to the availability of more successful anti-retroviral treatment. Mortality rates varied significantly by county, however, with respiratory disease death rates higher in the northeast and HIV/AIDS rates higher in the southeast (el Bcheraoui et al. 2018) . CDC groups its data statistics by topics; groupings include alcohol use, cancer, diabetes, heart disease, overweight and obesity, physical activity, and tobacco use. NHIS, BRFSS, YRBS, NHANES, and cancer registries are the primary sources for these data.

It is fair to say that political factors also play into decisions about what to surveille. In the United States, while automobile accident death rates are widely known, death rates from guns, although very high, are more difficult to tease out from the available statistics. The CDC maintains a data category for violent deaths which indicated that in 2015 approximately 62,000 persons died by violence. These data are obtained from death certificates, medical examination, law enforcement reports, and secondary sources; 27 states collected statewide data for 2015 but the remainder did not (Jack et al. 2018) . Thus, the data are incomplete, although nonetheless telling as far as they go. According to one recent analysis, injury was the leading cause of death for children; 20% of injuries occurred in motor vehicle accidents and 15% were firearm related. Both of these rates were far higher than the rates in other highincome countries; the U.S. auto accident death rate for children was three times higher than the average in comparable countries and the firearm death rate was an astounding 36.5 times higher (Cunningham et al. 2018) . Rates in rural areas were higher than in suburban or urban areas, particularly for automobile accidents. Rates of firearm related deaths and accidental deaths overall also were higher for males and for blacks. The authors of this study conclude that application of rigorous public health methods has had a major difference in reducing childhood deaths from automobile accidents and should be applied to other categories of injury-related deaths. An editorial accompanying the study judges that deaths from gun trauma are an "underrecognized public health problem" that has been mischaracterized as deaths due to "accidents" in a way that conveys a "sense of helpless inevitability" (Campion 2018) . These statistics are another illustration of the point made in Chapter 4, that ethical issues attend what is not surveilled as much as what is surveilled.

State disease reporting requirements focus primarily on contagious diseases or bioterrorism. To take one example of state notifiable conditions, the California Department of Public Health (2018) lists 80 communicable diseases that includes Ebola, Flu, HIV/AIDS, sexually transmitted disease, Hepatitis B, Zika, tuberculosis, and many other less well-known conditions. These are all conditions that have been judged to have the potential to threaten health if they are not identified in order to interrupt transmission. Reportable NCDs and conditions include disorders characterized by lapses of consciousness, suspected pesticide injuries, and brain tumors. The first of these-lapse of conscience-is important for drivers who might be hazardous to others. The latter two are related to environmental hazards. NCDs such as diabetes or substance use disorder are not among the listed conditions requiring reports. National statistics for NCDs such as these typically rely on the federal surveys conducted by CDC, such as BRFSS, NHANES, and NHIS (e.g. CDC 2020). Fragile funding and response rates for these federal surveys are a continuing problem, however. States also maintain registries for birth defects and cancers in response to funding from the federal government, as described in Chapter 5.

The new public health seeks to improve population health and prevent illness of whatever kind (Nuffield Council 2007, v) . Its primary justification is consequentialism applied at the health policy level: policies should be designed to promote the good of health. To the extent that differences in social determinants of health such as education, employment, or housing affect overall population health, issues of equity and justice also require attention. These considerations apply at the population level: overall well-being and its distribution throughout the population are the critical aims of the new public health (Parmet 2009 ). Not surprisingly, this population-level approach does not sit well with individualist and libertarian political ideologies, especially in the United States (e.g. Wiley 2016; Wiley et al. 2015; Jacobson 2014) . In the US, themes emphasized in health policy include individuals' responsibility for their health, development of increasingly individualized forms of health care treatment such as precision medicine, and improved access of individuals to care. The Affordable Care Act supports incentives for wellness programs and smoking cessation that are aimed to encourage individual health improvements rather than addressing social factors influencing health such as job insecurity.

Opposition to the new public health in the U.S. began to flourish during the administration of President Ronald Reagan in the 1980s. The politics in the U.S. had shifted away from confidence in government. Ronald Reagan was elected president in 1980, and with his presidency came sustained efforts to curtail the role of government. Budget cuts enacted in the first year of the Reagan administration vastly reduced benefits for the working poor and reduced the reach of the Medicaid program to provide health insurance for those without resources. Through waiver programs, states also received more flexibility in designing the structure of their Medicaid programs, enabling them to decide whether to cover home and communitybased services for people with disabilities or to require recipients to receive certain benefits through managed care plans. Social safety nets frayed with these attacks on what was judged to be welfare-created dependency. Efforts were made to introduce market incentives through changes in reimbursement structures in the Medicare program that serves the elderly and people with disabilities (Ethridge 1983) . Another example of the introduction of private sector incentives in the U.S. during the period was the Bayh-Dole Act of 1980, which permits universities and non-profits receiving federal funding for research to maintain ownership of intellectual property rights in discoveries resulting from the research. Support for spending on medical research through the National Institutes of Health continued, however.

A particularly vehement and influential early critic of the new public health was the libertarian law professor Richard Epstein, a member of the faculty at the University of Chicago and well-known for his use of free-market economic tools in analyzing law. A conference at the University of Chicago in 2002 addressed the social determinants of health and resulted in a special issue of the journal Perspectives in Biology & Medicine that included Epstein's critique of the new public health and responses to it. In a speech at a conference on obesity held by the conservativeleaning American Enterprise Institute in 2003, Epstein sparked intense controversy when he argued for limiting the role of public health to contagious diseases and what the law calls environmental nuisances-conditions on the property of one person that spill over to harm others. Public health activities aimed at these concerns, Epstein thought, addressed situations in which governmental intervention was needed to avert the market's failure to yield desired results. Because cleaning up the environment involves the production of public goods, as described in Chapter 4, market incentives aimed at individuals cannot be relied on to do the job. Nor can individual actions, coupled with damage remedies when some unreasonably inflict harm on others, succeed in reducing the spread of infection. On Epstein's libertarian view, the government should refrain from interfering with the market, taking action only when the market fails, as it will with infection and pollution. New public health actions designed to discourage unhealthy behaviors such as soda taxes aimed to reduce soft drink consumption fall outside of the proper scope of government, on this approach.

Epstein's justification for his anti-paternalist positions about the limited role of government (2003) invoked consequentialist reasoning directly at odds with arguments offered in favor of the new public health. His argument rested on empirical assumptions of classical liberalism about market functioning and the success of individuals in directing their own lives. Epstein wrote:

My broad thesis is that the "old" public health is superior to the new, whose broad (and meddlesome) definitions of public health help spur state actions-including the regulation of product and labor markets-that in all likelihood jeopardize the health of the very individuals the new public health seeks to protect. The new public health extends regulation into inappropriate areas, and thus saps the social resources and focus to deal with public health matters more narrowly construed.

Epstein judged that there were at least three ways in which the new public health threatened to sap resources. First, deploying resources to the new public health would divert resources away from areas in which they could be employed more beneficially, especially contagious disease protection and provision of health care desired by individuals. Second, it could create counter-productive incentives. For example, Epstein thought that policies aimed at tobacco might undermine incentives for people to take charge of their own health by stopping smoking, instead perhaps leading them to believe they could recover for any harm they suffered from smoking by bringing damage suits against manufacturers or by receiving ameliorative health care at public expense. Here, Epstein argued that health insurers in the private market should take measures to prevent what is called "moral hazard": the tendency for people to engage in more of a risky activity if they believe they will receive compensation when things turn out badly and the risk falls on them. Governmental intervention, he contended, could instead further moral hazard. New public health regulations might also encourage manufacturers to take measures to avoid any regulatory impact, thus incurring expenditures without any value to consumers. Third, new public health measures could direct interventions against individuals in ways that would not be beneficial to them. Epstein used policies such as increasing taxes on sugary soft drinks to combat obesity as illustrations, arguing that they imposed costs on all consumers, whether or not they are at risk of obesity or find its consequences deleterious. These taxes may have a disparate impact on those who are less able to pay, moreover, actually reducing the abilities of low-income families to purchase nutritional food if they also wish to purchase soft drinks. With respect to the impact of new public health policies on individuals, Epstein invoked the familiar view in classical liberalism that individuals are the best judges of their own interests.

This libertarian assault on the new public health continues to be influential not only in U.S. public health policy but also in U.S. health policy more generally. States that are reluctant to adopt the Medicaid expansion of the Affordable Care Act continue to try to implement policies aimed at placing more responsibility on individuals, such as co-payments for care or requirements to work in order to receive benefits. So-called "right to try" laws that permit terminally ill individuals to receive experimental drugs from pharmaceutical companies, bypassing the approval process of the Food and Drug Administration, have swept the country, spurred by support from conservative organizations such as the Goldwater Institute and Cato Institute.

Responses to these attacks on the new public health include criticisms of their approach to legal history and constitutional law. Defenders of the new public health also argue that its libertarian critics confuse questions of population health with questions of individual health and mistakenly infer conclusions about populations from conclusions about individuals. Other defenders of the new public health take on its ethical position directly, contending that there are non-paternalistic reasons for many new public health initiatives and that paternalism is ethically justified in some situations. The remainder of this chapter takes up each of these lines of argument. A theme that runs through the analysis is that surveillance for the new public health may be justified even when direct interventions with individual behavior is not, as it generates information about social conditions and resource needs rather than coercing individuals to change their behavior. Some of the fissures in the U.S. health care system that are being revealed by COVID-19 lend further support to this argument.

A threshold line of argument against Epstein's excoriation of the new public health was the critique of his argument that the new public health involved departures from settled traditions in U.S. constitutional law. U.S. constitutional law protects individual liberty and property rights; Epstein contended that the new public health trampled over these protections. His critics replied that the right to property as recognized in U.S constitutional law had never included rights to use property in ways that could cause serious and unavoidable harm to others. Both new and old public health, these critics said, aimed to reduce diseases across the population (e.g. Gostin and Bloche 2003) . U.S. law had long recognized the importance of the state interest in protecting health. What had changed was the pattern of diseases causing mortality and morbidity within the population. Old and new had the same goal-healthbut achieving it took different forms in different circumstances and required different agendas. Along these lines, Wendy Parmet (2009) developed a thorough and farreaching account of how U.S. law has recognized and furthered efforts to pursue population health. Epstein's libertarian conservatism rests on a very strong view of private property rights on which interference is justified only to protect others from what the law judges to be "nuisances." Nuisances are wrongs imposed on either the public or other owners through the unreasonable use of one's property in a way that substantially and unjustifiably interferes with the public or with private owners' enjoyment of their property. Someone's property may be a nuisance if it emits pollutants and someone's body may be a nuisance if it emits organisms causing infections in others. Nuisances may be private civil wrongs giving rise to tort liability if they harm individual property owners such as neighbors. They may be public nuisances and subject the owner to damage remedies, civil fines, or even criminal charges if they harm the public, as a rat-breeding home might do. Epstein's account of U.S. legal history contends that the courts construed the scope of nuisances narrowly until the twentieth century. His critics identify a wide range of nuisances recognized in legal doctrines from colonial times.

While a full account of the U.S. constitutional law of property rights and nuisance is far beyond the scope of this volume, a brief survey may provide a helpful background for libertarian legal challenges faced by public health in the U.S. Under the 5th and 14th Amendments to the U.S. constitution, the federal government and state governments respectively are prohibited from taking property without due process of law and just compensation. This means that if the government seizes an owner's property rights, it must have an adequate justification and it must pay for the privilege. So, for example, if an owner has a right to keep pigs on his property no matter the odors they might release, but the government steps in to stop the pig breeding because people nearby claim the odors are making them sick, the government must pay the owner compensation for the lost value in using his property as a pig farm. This analysis assumes, however, that the property owner had rights in the first place that the government has impinged. Another longstanding U.S. legal doctrine is that private property rights do not include the right to maintain an unwarranted interference with a public right; there is no property right in maintaining a public nuisance (Mugler 1887). Abating such a nuisance does not require compensation because the owner never had the right in the first place.

The scope of this doctrine of public nuisance is where controversy lies, however. In response to circumstances of the times, U.S. law has adjusted and re-adjusted what constitutes unreasonable uses of property that cause substantial harm to the public. Mugler v. Kansas, the leading Supreme Court case holding in 1887 that abatement of a public nuisance did not require compensation, involved prohibition of the manufacture of alcohol. Today, the public nuisance doctrine is being invoked by governments in the litigation against opioid manufacturers, although it is unclear whether this argument will ultimately succeed. Arguably, a rat-infested house is a clear case of a public nuisance justifying intervention. However, many cases are far less clear, especially when they involve regulations that affect property uses without the government's assumption of ownership. It is also an established doctrine in U.S. constitutional law that governments may not "go too far" in regulation limiting property uses (Pennsylvania Coal 1922) . In working out the balance of what it is for regulation to "go too far" and thus amount to a regulatory taking, U.S. courts have shifted back and forth from stronger and weaker doctrines protecting property owners.

The primary area in which contemporary regulatory taking jurisprudence has been developed is not health but environmental protection, where building restrictions have been imposed to protect shorelines or wetlands. The currently controlling case, Lucas v. South Carolina Coastal Council, was decided in 1992. The owner of two lots on a barrier island was denied a building permit under coastal management legislation adopted after his purchase of the property. In an opinion written by Justice Scalia, the Court reaffirmed that any permanent physical invasion of property requires compensation, no matter how minimal the invasion and how weighty the public purpose. Presumably on this analysis, construction of a permanent station on private land for detecting harmful emission levels would require compensation for the invasion. For regulations "going too far" in affecting beneficial uses of the property, the Court held, the test is whether the proposed prohibited uses are beyond what is included in the owner's rights over the land in the first place. Applying this test requires balancing several factors: the degree of harm to the public or adjacent private property posed by the owner's proposed use, the social value of the owner's use and its suitability to the locality, and the relative ease by which the alleged harms from the use can be avoided (Lucas 1992, 1031). Landowners do not have rights to engage in activities that spill over to harm others or the public generally, so emission control regulations would be permissible. But if a use such as a grain storage facility has been longstanding, grain storage is a socially useful activity, and flea-carrying rats can be trapped as they enter adjoining neighborhoods, prohibitions of grain storage could be considered confiscatory and require compensation unless alternative beneficial uses of the property remain. 6.4 U.S. Constitutional History, the New Public Health, and the Powers of Government Some new public health efforts have encountered objections that they are regulatory takings, but the case law on this question is sparse. Massachusetts' law requiring disclosure of ingredients of tobacco products was held a regulatory taking with the court reasoning that it would have great economic impact on tobacco companies' reasonable investment-backed expectations in the formulas they used to make their products without sufficiently promoting public health (Philip Morris 2002) . The property in question was trade secrets in the products' ingredients. The court's reasoning that the state's justification was insufficient was that there was only speculative evidence that disclosure of ingredients would affect consumer behavior. On the other hand, food safety regulations such as prohibition on egg sales from farms that have tested positive for salmonella have been held not to require compensation of the property owner (Rose Acre Farms 2004) . Nor do health warnings about the possibility that perishable agricultural products may be contaminated by salmonella effect a taking, despite the owner's lost sales (Dimare Fresh 2015).

Cases invoking the public nuisance doctrine have been brought against tobacco companies, lead paint manufacturers, gun manufacturers, energy companies and, most recently, pharmaceutical companies making opioids. In these lawsuits, governmental entities have sought to recover public costs for health care, policing, public education, and other public services resulting from the harms they claimed were caused by the product in question. These are cases brought by the government seeking to recover costs for public health, not cases brought by private individuals seeking damages such as for cancers caused by smoking. The public case against tobacco companies settled (Master Settlement Agreement 1998), encouraging the suits that have followed. Other suits have not fared as well, however. States and municipalities have lost in some cases against paint manufacturers (In re Lead Paint Litigation 2007; State 2008) although California succeeded in claiming that manufacturers had promoted lead paint for indoor use until 1950 despite knowing it was harmful to children. (People 2017) Suits against energy companies claiming that the effects of fossil fuels on the climate are a public nuisance have failed (e.g. Native Village of Kivalina 2012). One case against gun manufacturers filed by the City of Gary, Indiana, in 1999 was still continuing as of early 2020 (City of Gary 2019), despite legislative grants of immunity to firearms companies for lawful sales of guns or harms resulting from third party wrongdoing. Public nuisance claims also are involved in the continuing litigation against opioid manufacturers (In re National Prescription Opiate Litigation 2020).

It is fair to say that legal doctrines about the scope of public health in the U.S. are continuing to evolve. Whether governments will be able to recover damages on the theory that health-harming products are public nuisances, or commercial entities will be able to assert property rights in their activities, will continue to be adjusted. The views of libertarian conservatives such as Epstein represent only one pole in the debates, but one that is nonetheless powerful, especially with the current make-up of the U.S. courts.

Other critics attack the libertarian position as confusing questions of individual health with questions of population health. These criticisms begin with the argument made by epidemiologist Geoffrey Rose (1985) that sick individuals and sick populations are fundamentally different problems requiring fundamentally different approaches. What explains the variance in individuals' susceptibility to disease-their genes, exposures, behaviors, or choices-may be quite different from what explains the variance in populations' susceptibility. Population-level approaches such as tobacco control, in Rose's view, would be far more effective in reducing consumption than efforts to modify individual behavior. This analysis would imply that while libertarians may be correct that regulation may affect particular individuals in ways that would not be beneficial to them, they are not correct that the same regulations would not be beneficial for public health overall. The Nuffield Council on Bioethics, in developing a stewardship framework for public health ethics, put the point thus:

It takes only a moment's thought to recognise that many of the 'choices' that individuals make about their lifestyle are heavily constrained as a result of policies established by central and local government, by various industries as well as by various kinds of inequality in society. People's choice about what to eat, whether or not they allow their children to walk to school, or the kinds of products that are marketed to them, are often, in reality, limited. This means that the notion of individual choice determining health is too simplistic (2007, p. v).

Libertarians and their critics make quite different assumptions about the efficacy of population-level interventions and the role and capability of individuals, however. For libertarians, individuals are better judges of their own interests and have considerable ability to control their own lives. Population-level interferences may be wrong-headed on a grand scale. An example would be the assumption that health is a paramount value for the population when, for many members of the population, health is not at the top of their list. Many may judge health to be less valuable than other goods in their lives such as pleasures of consumption. For the libertarians' critics, by contrast, individuals may sometimes be wrong, or sometimes have information that is incomplete. Moreover, many features of individuals' health are beyond their control, no matter how hard they try or how carefully they choose. These features will require interventions at the population level. They go far beyond infectious disease or environmental toxins. No matter how much an individual might want fresh vegetables, she cannot will them into the local grocery store. Nor can she exercise successfully if the streets are unsafe or the air is polluted.

These points about the difference between population-level and individual-level questions hold for information, too. Some important information can only be gleaned at the population level just as effecting some changes in behavior require changes at the population level. Understanding the role of a wide variety of factors in causing disease, unearthing rare side effects or drug-drug interactions, or figuring out which gene variants are deleterious, cannot be accomplished with information about single individuals. In this respect, gathering information is like altering the built environment: just as individuals cannot will the presence of healthy food in their local stores, they also cannot will the information about rare events from a single case or even a significant group of individual cases. For this reason, it may be important for public health to have access to data about wide ranges of individuals.

The opioid problem illustrates how important population-level data may be. It took a considerable time after deaths from opioid overdoses started to rise for observers of health trends to recognize the extent of the problem. Evidence that rates of addiction to prescription opioids and opioid related deaths were rising began to appear over the first decade of the twenty-first century (Okie 2010) . Opioid overdose deaths began rising sharply around 2000, when they surpassed rates of death from cocaine. A pub med search indicates that the earliest characterization of the situation as an "opioid crisis" or "epidemic" appeared about 2011 (e.g. Dhalla et al. 2011; Knoppert 2011; Manchikanti et al. 2012) . In 2010, the FDA proposed a Risk Evaluation and Mitigation Strategy (REMS) to require physician and patient education about opioid risks. REMS was met with controversy because it was voluntary only. By that point in time, CDC data indicated that the highest rates of addiction and death were occurring in predominantly rural areas, although the data were incomplete because of variations in state surveillance systems. Had the data been better, the extent and shape of the problem might have been recognized earlier and addressed with greater success.

At the core of ethical objections to the new public health is paternalism. Described most generally, paternalism is interfering with someone's liberty for their own good when they do not agree to the interference (e.g. Dworkin 1972) . So, it would be paternalistic to impose fines on willing smokers to get them to stop smoking and to prohibit people who are obese from purchasing sugary soft drinks. The object of this Essay is to assert one very simple principle, as entitled to govern absolutely the dealings of society with the individual in the way of compulsion and control, whether the means used be physical force in the form of legal penalties, or the moral coercion of public opinion. That principle is, that the sole end for which mankind are warranted, individually or collectively, in interfering with the liberty of action of any of their number, is self-protection. That the only purpose for which power can be rightfully exercised over any member of a civilised community, against his will, is to prevent harm to others. His own good, either physical or moral, is not a sufficient warrant. He cannot rightfully be compelled to do or forbear because it will be better for him to do so, because it will make him happier, because, in the opinions of others, to do so would be wise, or even right. These are good reasons for remonstrating with him, or reasoning with him, or persuading him, or entreating him, but not for compelling him, or visiting him with any evil in case he do otherwise. (Mill 1859, pp. 17-18 ).

On Mill's view, paternalism unjustifiably thwarts liberty and does so in ways that are likely to prove counterproductive. Importantly, Mill applied his claims only to adults; whether paternalism may be justified in the case of children or "the race in its nonage" he regarded as separate questions.

Over the century and a half since Mill wrote, philosophers have debated whether his position rested fundamentally in his commitment to liberty or in his overall utilitarianism. The difference is this: if liberty is the foundation for anti-paternalism, then in the final analysis paternalism must be unjustified even if interference would have the best consequences overall. On the other hand, if utilitarianism is the foundation, then anti-paternalism must have the best consequences overall in order to be justified. If paternalistic interference would have better consequences overall, a utilitarian would have to admit that it is justified, so utilitarian anti-paternalists must contend as an empirical matter that in the long run interference would not work out for the best. And they have done so, claiming that interferers will misjudge what is good for others, that interference will be counterproductive, or that interference will be far too costly to outweigh any benefits. A typical way for utilitarian antipaternalists to structure these empirical arguments is to apply them to rules: rules that prohibit paternalism will have the best consequences overall, they contend, even if their application to a particular case would appear counter-productive. So even if taxing sodas would prevent dire health consequences for a few, a policy of taxing sodas would burden others more or potentially backfire by causing resentment. Along similar lines, some might contend that restricting access to opioids, while it might do good in some cases, will result in more untreated pain overall.

Writers following Mill have also debated whether his commitment to liberty was to political liberty-the absence of state or social coercion-or to autonomy in a deeper sense of some kind of freedom of the will. In addition, what counts as coercion has been viewed very differently by commentators. Threats of injury or punishment are clearly coercive, but what about offers or changes in the structure of choices that are designed to change behavior? Behavioral economists call these changes in "choice architectures." Some argue that such an offer may coerce if it is so generous that a person has no choice but to accept it. An example would be paying very poor people substantial sums of money to lose weight. Some also argue that an offer may coerce if it would provide someone with a good that he or she has no right to receive. Paying kickbacks to clinics to allow patients to receive treatments for which they would not otherwise qualify would be an example.

Others, however, contend that offers do not coerce if they do not threaten to violate rights or obligations and people may choose to forego them, even if the choice might be to give up a benefit that is very attractive under the circumstances (e.g. Wertheimer and Miller 2008) . So on this view, "wellness" discounts for health insurance offered to people who lose weight or stop smoking are not coercive, even if the discounts are very attractive, as long as people may choose not to accept them. 6.6 Paternalist and Non-paternalist Ethical Objections to the New Public Health Some also argue that drug company payments to clinics for writing prescriptions would not coerce those receiving the prescriptions, although it might be objectionable on other grounds such as driving up the costs of health care.

Many less drastic changes in choice architecture alter available options or the ways in which these options appear to people in order to take advantage of insights from behavioral economics about how people make decisions. Changes in presentation of the items in a buffet line is the classic example: people will take more desserts if the desserts are prominently displayed than if they are on lower shelves or last in line. Wellness programs often use these devices, offering small rewards for behavioral changes such as better diet management or taking walks during lunchtime breaks. Whether such "nudges" are paternalistic and, if so, questionable interferences with liberty, is highly controversial in many areas of the new public health (e.g. Thaler and Sunstein 2008) .

Ethical arguments for the new public health answer the charge of unjustified paternalism in a variety of ways. First, pursuing the goals of the new public health may not be paternalistic at all, if it occurs with consent. While it is paternalistic to make people do things for their own good, and perhaps even to force information on them against their wishes, it is not paternalistic to gather information that they actually consent to share. Whether it is paternalistic to gather information or to interfere with behavior without actual consent, but in ways people would have consented to, is more controversial, however. Second, there are non-paternalistic reasons for surveillance of NCDs, especially reasons rooted in justice; for example, data about the interplay between education and health may suggest needs for reform in primary and secondary education in the U.S. This is an argument from justice, not an argument from what would be best for individuals or even groups. Some contend, however, that as long as non-paternalistic reasons are coupled with paternalistic ones, problematic paternalism remains.

Finally, the paternalistic objection itself many be questioned. There are at least three lines of reply to the claim that new public health surveillance wrongly coerces people to do things for their own good. First, gathering information through surveillance does not entail intervening with behavior, although the information may lead to interventions. People may consent to learning about whether obesity and type 2 diabetes are connected without consenting to attempts to alter their weight; indeed, this information may be necessary for decisions to be well informed. Second, a distinction between impositions that override individuals' choices and impositions that result in what people would have chosen with adequate knowledge can be helpful. Finally, there are circumstances in which paternalism may be justifiable to avert significant harms, although these circumstances must be carefully delineated if surveillance is to be sustained.

It is not paternalistic to interfere with someone who has voluntarily agreed to the interference. Of course, the devil is in the details about what counts as voluntary agreement. Nor it is paternalistic to interfere for reasons other than the individual's own good, such as for the protection of harm or injustice to others.

Obtaining information when people consent to acquiring and receiving the information in order to make their own choices is not paternalistic. Information gathering is only paternalistic if it is designed to obtain information that people do not consent to provide or to receive, and to do so for their own good. Cholesterol testing is a good example. If a patient is required to have his cholesterol tested for his own good, and does not consent, this would be paternalistic. If he consents to the testing because he wants to know his own health status or wants to help other family members know whether there is a family history of high cholesterol, undergoing the test would not be paternalistic. It would, however, be paternalistic to require him to receive his results if he does not wish to have the information but is being tested only for the benefit of his family. Surveillance for new public health objectives is not paternalistic if it obtains information that people actually consent to have collected and does not force them to receive information for their own good when they do not consent to do so. Much information obtained and used by the new public health would not violate these constraints. Individuals may willingly provide and receive health-related information. Problems occur when either some do not want to give information or some do not want to hear it.

An important complication here is the point made above about the difference between population-level and individual-level questions. The new public health uses population-level data, but this data must be gathered from individuals, even if they are not identified in how the information is gathered. Some would object that it is unjustified to collect information from individuals for the good of the population unless the individuals would agree to collection of the information for their own good. Flanigan (2013) for example argues that unless individuals consent to collection of information from them, it is paternalistic to coerce them to provide the information for the overall public good. Instead, she contends, they must be permitted to opt out from this kind of surveillance in order to respect their liberty. Cholesterol again may illustrate. Suppose that at the population level it is valuable to be able to answer questions such as: what are the correlations between cholesterol levels and health consequences? What characteristics of people or their behavior might affect these correlations-for example, do people with high cholesterol who exercise tend to have fewer adverse cardiovascular events than people who do not? What are the benefits and side effects of various anti-cholesterol medications? What characteristics of people or their behavior might be related to the frequency of these benefits or side effects? Meaningful answers to these questions will require data that are sufficiently representative of the population, especially to detect low-frequency events; allowing people to opt out may compromise the analysis. The question is whether collecting population level data for the good of the population is paternalistic unless the collection is modeled on individual consent, as Flanigan argues that it is.

Arguably, collecting the data to answer these questions is not paternalistic because it is not done for any individual's good against his own choice. Instead, it is done at the population level for the benefit of the population. To be sure, there may be individuals who would prefer not to participate in the gathering of populationlevel data for many different reasons and would not consent to have data drawn from them used in the effort to answer any of these questions. Whether it is justified to use data from them, however, is not a question of paternalism; it is a question of the ethics of using data drawn from individuals who do not and would not consent, when the data are needed for the benefit of others. Whether data are needed depends on whether any opt out policies would undermine representativeness-itself difficult to know without having at least some data about the nature of the population. Privacy scholar Mark Rothstein (2010) has argued that de-identification is insufficient to protect privacy because of risks of stigmatization and data uses individuals would find objectionable. He notes the possibility of selection bias in data but contends that this speculative risk must be balanced against other potential harms. Others have noted that selection bias may be particularly problematic for underrepresented groups (e.g. Hoffman 2010). These groups may be small in number in the first place and may also for historical, religious, or cultural reasons be more likely to opt out. Such opting out will potentially compromise the information available to group members who do not wish to opt out, so the argument is that the data collection is not paternalistic but for the benefit of these others. Justification must, then, rest on other grounds such as the equity considerations explored in Chapter 4.

Groups present complicated issues in this regard, however. Some members of a group may disagree with the choices of others about the good of the group and argue that it is unjustified paternalism to make them share information for the good of the group. The population-level reply given in the preceding paragraph would be that this is not paternalism because the individuals are being required to share information for the good of others. However, it is possible for groups or even nations to be the subject of paternalism. Michael Barnett (2015) argues that forms of global governance such as those encouraged by the WHO are paternalistic if they impose humanitarian goals without the agreement of those to whom they are applied.

Suppose, for example, that members of a religious or ethnic group have immigrated and settled together in a new area. They are sufficiently cohesive to be described as a group; although they do not have a governing structure, some worship together, many join together for celebrations at a community center, and there are interlocking structures of friendship and familial relationships among members of the group. They retain dietary preferences from their homeland that have been associated with weight gain and high rates of diabetes among similar immigrant groups. Public health officials in the jurisdiction in which many of the group members reside note that there are high rates of obesity in the schools and of admissions for cardiovascular disease in local hospitals. Death certificates also indicate higher than average rates of premature death. Public health might therefore wish to gain information about the group and its dietary practices in order to determine whether or not further intervention might be warranted for the good of the group's health. Many group members, however, might object to the collection of information about their health, such as information from their medical records, grocery store purchases, and the like. This surveillance might be characterized as paternalism aimed at the group-information to enable analysis about health interventions that might be needed to protect the group against itself. (Surveillance likely would also violate equal protection if the information was collected from group members only because of their religion or ethnicity.). Even if recognizable group leaders consented to the surveillance on behalf of the group, it still might be characterized as paternalism against individual group members who object; we take up these questions about the implications of group consent in Chapter 7. Analogous points might be made about whether democratically made decisions to paternalize reflect the "consent of the governed" and therefore are not paternalistic to those who voted against them (Nys 2008) , arguments that we also consider in Chapter 7. In the remainder of this chapter, we consider whether non-paternalistic arguments such as equity may be convincing, and whether, if not, there may be circumstances in which paternalism may be justified, whether it is directed against individuals or against groups.

Paternalism means imposing choices on people for their own good. Many apparently paternalistic public policies also have non-paternalistic justifications.

Arguably, these policies are therefore not paternalistic, although some would still argue that to the extent that the paternalistic justification for these policies carries weight, they remain problematic. The requirement to wear seat belts in automobiles or to wear motorcycle helmets may be justified to reduce the costs of accidents to others. Like wellness programs, however, some of these programs may primarily be efforts to protect people from themselves; if so, they are paternalistic. Some justifications for new public health surveillance, however, are clearly not paternalistic but are rooted in justifications such as equity or justice. An equity-based justification does not make people do things for their own good; it intervenes to eliminate unfair disparities that adversely affect marginalized groups, as we explained in Chapter 4. Many of the efforts of the new public health are aimed at improving health equity. Take, as an example, evidence about education as a social determinant of health. Longstanding assumptions in U.S. political ideology are the role of individual choices in success and the need for education to give people opportunities to make these choices. One of the earliest lines of argument about the social determinants of health was that improved education would yield improved health. The first of the "Healthy People" assessments of the health of the U.S. population and how it might be improved, issued in 1979, attributed at least half of mortality in the U.S. to "unhealthy behavior or lifestyle" and only 10% to inadequate health care (DHHS 1979, 1-9) . The report opined that "the health of this Nation's citizens can be significantly improved through actions individuals can take themselves, and through actions decision makers in the public and private sector can take to promote a safer and healthier environment for all Americans at home, at work, and at play" (DHHS 1979, 1-12) . To enhance children's health, the report recommended increased support for preschool programs such as Head Start; it also emphasized the "special importance of the school" and how "[o]ur children could benefit greatly from a basic understanding of the human body and its functioning, needs, and potentialand from an understanding of what really is involved in health and disease" (DHHS 1979, 4-16) . Education was judged core to reducing risks of teen pregnancy, smoking, poor diet, and drug and alcohol abuse. Studies published during the decade following in the American Journal of Public Health addressed topics such as the need to write smoking cessation manuals at a 5th grade reading level (Meade et al. 1989) , correlations between smoking rates and education (Novotny et al. 1988 ), relationships between a wife's education level and her husband's susceptibility to coronary disease (Strogatz et al. 1988) , and correlations between education and teen pregnancy (Joyce 1988) . Editorials also weighed in, on topics such as the role of education in disparate mortality rates from cardiovascular disease between whites and blacks (Wing 1988) or poor nutrition (Editorial 1988) . Improving education, it seemed, might be the very best way to improve population health.

The relationship between education and health is far more complex than any simple correlation between increasing education and improving health, however. Developing understanding of the interrelationship among social determinants of health and policy interventions that may prove successful in addressing health inequities requires ongoing data analysis. Here are several examples. In 2002, a review of socioeconomic status and health judged that environmental risk exposures were a major determinant of health and included school facilities among environmental factors. This study noted that low income schools were significantly more likely to have problems with building repair and infrastructure such as heating (10% vs. 30%) and to be overcrowded (6% verses 12%) and significantly less likely to have teachers with degrees in math or science teaching these subjects (27% versus 43%) (Evans and Kantrowitz 2002) .

To take a more recent example, in a report prepared for the U.S. Agency for Healthcare Research and Quality, researchers reviewed the extensive evidence of the impact of education on individual, community, and population health (Zimmerman et al. 2015) and developed an ecological framework for analysis. The reported differences are significant: a 9-year difference in life expectancy between adults age 25 with a high school diploma and those with a college degree. Understanding what is at work behind these data about education and health is critical to designing policies to address them. For this, a far more complex evidentiary picture is needed. Zimmerman and colleagues sketch out what some of this evidence might be. There are many direct and indirect effects of education. Health literacy is one direct effect: people with higher levels of education are better able to understand their health needs, communicate with providers, understand care plans, and navigate the health care system. Adults with more education also can access other benefits: employment, income, and social networks and support. Education also correlates with reduced levels of risky behavior such as smoking, alcohol use, poor diet, and inactivity. Adults with more education are also at lower risk of longterm stress. Evidence also reveals neighborhood-level effects from factors such as the availability of resources, social capital, and social organization. Lower-income neighborhoods are more likely to be unsafe, have poor access to nutritious food and green space, and have poorer schools. The evidence suggests that the relationship between education, social contexts, and education is complex and interactive, however, and may be influenced by external factors such as legislative funding decisions and social safety net buffers. Nor can educational inequalities be separated from inequalities by gender, race, ethnicity, sexual orientation and disability, say Zimmerman and coauthors. Causality may go in reverse as well, if ill health adversely affects educational success because of increased absenteeism or cognitive difficulties. Education may also be a proxy for other factors, such as the lack of family stability. The authors recommend further research to better understand the connections between education and these other social factors, including research following the model of community-based research in which community members help to shape research questions.

An even more recent example questions the role of education, arguing that wealth is far more significant than education in affecting later educational outcomes and SES. According to this study, education and SES in turn will correlate with health, as just described. This study examined correlations between early measures of academic ability and educational success and found that half of the children from low socioeconomic status families who had high test scores in kindergarten were behind by eighth grade; for Black students from low SES families, the rate was even higher at 60% (Carnivale et al. 2019 ) and for Asian students it was the lowest, at 20%. The authors of this study concluded that in addition to good schools backed by sound educational policy, success requires providing students with the environmental supports they need for success.

These studies, and the need for the data to conduct them, reflect the strong equity concerns surfaced in Chapter 4. They involve children and likely opportunities. The disadvantages they involve are likely corrosive, interacting to yield even worse outcomes for the futures of those they affect. Surveillance to examine relationships among factors affecting opportunity is justified as a matter of social justice, not as a matter of protecting people for their own good.

On the other hand, some examples of surveillance for the new public health are paternalistic to either individuals or groups, even if they are described in other terms. Consider wellness programs that require individuals to "know their numbers" about blood pressure and cholesterol. These programs may be justified on non-paternalistic grounds such as reducing overall health care costs or workforce absenteeism, or perhaps selecting healthier employees or encouraging them to join or stay in the workforce (Jones et al. 2018) . But these reasons may be thinly veiled paternalistic efforts to prod people to take better care of themselves and to take charge of their health. Indeed, some data suggest that while the programs do affect employee behavior they have no significant impact on absenteeism or health care costs (Song and Baicker 2019) . If these programs require health-related information to be obtained and presented to individuals for their own good, whether or not the individuals want the information, they are paternalistic.

Public "Bads."

Economists and others who rely on economic reasoning such as Epstein point to market failures as justifying governmental intervention. Markets fail in the case of public goods because individuals can receive their benefits without incurring any of the costs of producing them, so have no incentive to pay the costs. Markets may also fail when private behavior spills over to impose costs on others as with pollution; in such cases, if the transaction costs of organizing individuals to oppose the pollution are high, there is a utilitarian economic case for intervention to force the polluter to internalize the costs of these externalities. Public health law professor Lindsay Wiley (2012, p. 213 ) and others argue that there are also public "bads"-negative externalities inflicted on the public without their consent-that should be the subject of governmental attention. Examples are environmental factors that encourage unhealthy eating or unsafe parks that make outdoor exercise very difficult. To defend the role of the state in addressing these public bads, Wiley points to the idea of a "public nuisance" described above: an unreasonable interference with a right common to the general public, such as a stream (2012, p. 235) . In public nuisance suits, the required proof is of harm to the public in general, not to individual members of the public. The harms are to the collective rather than to individuals considered separately from one another, even a reasonably large number of them. Wiley's argument is that the science of epidemiology is yielding information about how the social environment affects population health, information such as that summarized in the discussion of education above. To refuse to expand the scope of public health to these public bads, Wiley argues, is to detach practice from science. On this view, epidemiological harms such as the food or transportation environment should be addressed by public health, just as individual harms such as salmonella in a bunch of spinach consumed by a particular purchaser should be. For many public policies there will be both paternalistic and non-paternalistic reasons that can support them; these might be called "mixed motive" paternalism. Sometimes, the non-paternalistic reasons offered for a policy are thinly veiled excuses for paternalism. Moreover, critics might argue (e.g. Flanigan 2013), the objections to paternalism are so strong that the non-paternalist reasons are overridden by them. So, the question of paternalistic justifications for the new public health cannot be avoided. We now take on the question of paternalism directly.

Is it ever justifiable to make people do things for their own good? Is paternalism more plausible depending on the importance of the good? If so, is health an especially compelling good? Does the justification of paternalism change depending on whether coercion is used, or whether softer methods can achieve the same ends?

Contemporary debates about paternalism take the form of asking whether paternalism can be justified. They start out with the assumption that it is paternalism-not the opposite-that requires justification. It is worthwhile pointing out, however, that this way of structuring the argument itself makes the assumption that the burden of justification must be borne by the paternalist. That this assumption seems so reasonable stems from presumptions about the importance of liberty: that it is interferences with liberty that require justification, not liberty itself. The normative burden of persuasion is on the defender of interference, not the defender of liberty. Perhaps liberty is so critical a value that it is plausible to place the burden of persuasion on those who would limit it. However, the problem of justifying paternalism might be conceptualized quite differently in views that were not wedded to liberalism by prioritizing liberty in this way.

The presumption of liberty just described functions in an argument against interference with individuals' choices about how to live their lives. It thus applies to questions such as whether to interfere with people's food choices in order to promote healthy weight. Applying these presumptions to surveillance rather than to choices about health-related behavior, however, may be less plausible. Take, for example, surveillance about the effects of soft drink consumption on health. Is the evidentiary burden on public health to show that consumption of soft drinks is associated with weight gain that is in turn associated with diabetes, or on opponents of regulation to bring evidence that the associations do not exist? Is the evidentiary burden on public health to show that there are no unanticipated consequences of soft drink regulation, or is the burden on opponents to show that there are such unanticipated consequences? The logic is similar with opioids: does the proponent of intervention bear the burdens of demonstrating harm or the non-existence of unanticipated effects? The answers may have significant consequences in practice, if interventions are delayed for considerable periods in order to allow time to gather and assess the evidence.

Evidentiary burdens of persuasion have implications for the need for surveillance. If evidence is required to justify paternalism, but cannot be obtained, the result will be inability to justify the paternalism. Strategies for keeping evidence at bay will thus prove successful in keeping intervention without apparent justification-as tobacco companies and firearms manufacturers have realized to their advantage. Thus, there may be reasons to separate surveillance questions from paternalism: having the information may be critical to deciding whether an intervention is paternalistic or, even if so, whether it can be justified. In the words of Chokshi and Stine (2013) , a "savvy" state is different from a "nanny" state.

The term "soft" paternalism was initially coined to refer to forms of interference that are needed to ascertain whether an individual is acting voluntarily (e.g. Dworkin 2017). Other writers have used the term to describe interferences that are less invasive than the use of outright prohibitions or serious threats (e.g. Conly 2013; Nys 2008) . In this section, we consider the justification of "softer" forms of paternalism of these two types. Analogously, political scientists use the term "soft power" to describe the ability of a country to persuade others do to what it wants through shaping long-term attitudes and preferences (Nye 2004) .

One kind of soft paternalism involves providing people with information in order to be sure they are acting voluntarily. Sometimes bystanders may be aware that others are in danger, but not know whether their danger is willingly incurred. Is it paternalistic to intervene in such cases? Mill himself argued that it would be justified to stop someone from crossing an unsafe bridge who did not know of the bridge's condition because this would not interfere with his liberty (1859, pp. 182-183) . But for Mill this justification extends only to providing the information: "Nevertheless, when there is not a certainty, but only a danger of mischief, no one but the person himself can judge of the sufficiency of the motive which may prompt him to incur the risk: in this case, therefore (unless he is a child, or delirious, or in some state of excitement or absorption incompatible with the full use of the reflecting faculty), he ought, I conceive, to be only warned of the danger; not forcibly prevented from exposing himself to it." Surveillance needed to provide information would not be problematic soft paternalism, on this view, but it would be problematic to go further in attempting to change behavior.

Lack of information is not the only way in which behavior might not be fully voluntary. Behavioral economists Richard Thaler and Cass Sunstein have popularized the idea of "nudges" as a form of paternalism which, they argue, should be acceptable to libertarians. Nudges work with the observation that human decisions are beset with cognitive biases: we respond to the last event as most salient, are less willing to incur losses than to give up similar gains, and place disproportionate weight on the present. Nudges are defended as responses to the irrationalities created by these biases. For example, the bias in favor of eating a large piece of chocolate cake in the present might be countered by a mirror behind the food line that shows everyone five pounds heavier and information about the calorie count of the piece of cake.

Even if nudges take advantage of biases rather than countering them, some argue, they are permissible as long as people are made aware of them (Aggarwal et al. 2014; Thaler and Sunstein 2008) . Someone is not really tricked into foregoing the chocolate cake by an image of her with five extra pounds, on this view, as long as she is aware that she is being shown the image in order to discourage her from taking the cake. This line of argument-that the person is not tricked-does not show that the interference is not paternalist, however. The interference is still for the person's own good, even though it is designed to get her to think about the future rather than responding to the immediate present. Rather, it shows at best that the interference may not be particularly burdensome; with awareness that the interference is occurring, people may take the cake anyway.

Some theorists defend interferences that may be more burdensome. Sarah Conly (2013) argues that paternalism may be justified to help people pursue their longterm goals, even though in the short run it prohibits them from having what they want. Indeed, Conly contends that prohibitions may actually be more respectful of people as choosers than "softer" measures such as information or nudges, because they are more effective in helping people to achieve their long-term goals. In response to the objection that paternalism treats people unequally because it assumes that some know better than others what is good for them, she counters that paternalism simply recognizes that rationality is flawed for everyone. This recognition, she argues, does not fail to respect us as persons; rather, it assesses our abilities accurately and values our longer-term choices.

Along similar lines, Thomas Nys (2008) argues that "deeper" autonomy justifies paternalism in public health. He assumes the value pluralism of liberalism-that paternalism cannot be justified on the basis that it is objectively true that some things are goods for everyone. Instead, people may value different things, or value the same things differently; some may value health or even life more than others do. However, people sometimes make choices that go against or jeopardize their fundamental conceptions of their good. On Nys' view, interferences that are minimal are permissible; he uses information gathering through a cheek swab as an illustration of an intervention that does not interfere with broader autonomy about major life choices. Nor does it interfere with deep autonomy in the sense of being reflective about the worth of these major life choices.

These defenses of softer paternalism are liberal in the sense that they assume value pluralism. They accept that people may hold a range of different values and may balance these values differently. Longer life may simply not be a value for some. Or, while people may share the value that death is bad, they may balance it differently against other goods. Some-Kneiss (2015) calls them "food lovers"may value the enjoyment of fatty foods more than the enjoyment of additional days or even years of life. Soft paternalists do not undo this evaluative structure; rather, they intervene to further people on the way of realizing this value structure. The intervention is justified in terms of helping people to realize their goals, not in terms of helping them to have better goals. Often this justification is put in terms of a balance, with the intervention seen as most justified when it is minimal but helps people achieve things that are important to them on their own terms. In the end this line of reasoning must yield when individuals insist that behavior is in accord with their ultimate values. Importantly, however, the line of reasoning has greater purchase when the interference is limited to information that will help individuals to determine whether the behavior really is in accord with their ultimate values. Unless an individual has grounds to argue that acquiring the information will compromise her other ultimate values in some way, or that it is against her ultimate values to have the information she needs to determine what is in her ultimate values, surveillance would not be impermissible on this line of soft paternalist reasoning. Kneiss (2015) argues that paternalistic policies may be justified for reasons of fairness combined with soft paternalism. Some behavior is not entirely self-regarding; second-hand smoke from tobacco is an illustration. Sometimes, behavior is not entirely self-regarding because of the impact it has on how others may behave. Suppose some people are poorly educated about food or particularly prone to cognitive bias. Perhaps they are especially vulnerable to the influences of others. Suppose that regulations about food cannot be easily tailored to protect these people while food lovers who do not care about their health continue to eat as they wish. If so, Kneiss argues, it may be unfair to those who are vulnerable not to intervene paternalistically. He considers the objection that such vulnerability may be due to social conditions such as social determinants of health discussed earlier in this chapter and that, if so, it would be preferable to address the social conditions rather than intervening paternalistically. His reasoning is rooted in non-ideal theory: that in circumstances where social conditions are inequitable, addressing factors related to ill health directly may be part of an effective strategy for addressing unjust social conditions. By reducing ill health that leads to inequality of opportunity, we may address social disparities, rather than the reverse. To hold otherwise, Kneiss concludes, is to privilege liberty over opportunity.

"Hard" paternalists believe that it is sometimes justifiable to intervene to make sure that people have what is good for them. Hard paternalism requires a justification for the importance of the good and reasons for thinking that the good is more important than any costs of interference. Many hard paternalists are perfectionists rather than liberals: they believe that there are some goods that are so important for human life that everyone ought to have them, whatever their own views about their good.

Examples might be life itself or physical health. These positions are illiberal in that they fail to recognize individuals as ultimate sources of their good. Resolving fundamental questions of political philosophy about liberalism is beyond the scope of this volume. Suffice it to say at this point of the argument that if the justification for surveillance is ultimately rooted in this form of hard paternalism, it must reject liberal commitments to individuals as ultimate sources of the good for themselves.

Others take a somewhat less perfectionist balancing approach, weighing the importance of the choices to people, the degree of the interference, and the good to be achieved by interference. Wilson (2009) , for example, contends that interests such as not wearing a seat belt are simply not very important to people. Some take this position about providing information. Nys (2008) , for example, claims that providing information is not a particularly serious intervention because it is only temporary and does not interfere with the ultimate exercise of autonomy. The Nuffield Council (2007, xix) , in its analysis of the justification of public health interventions, structures an "intervention ladder"; monitoring the situation and providing information are the lowest rungs on the ladder. They place collecting anonymous information about outbreaks of contagious disease low on the ladder. Case reports are higher because of the potential for intervention, however. If these positions are based in the view that privacy in the form of protection from being required to share information about oneself is sufficiently unimportant for everyone that it may be outweighed by other goods, they, too, must reject liberalism's picture of individuals as ultimate sources of their good.

Wearing masks during the COVID-19 pandemic has become a particularly contentious issue in the U.S. It is an excellent example of these issues about paternalism. The non-paternalist argument for mask-wearing is that it protects others than the mask-wearer from the possibility of infection. But there are paternalistic arguments for mask-wearing, too, of the kinds sketched above. Wearing a mask protects the wearer. The wearer may not be a good judge of whether she is infected or likely to be at risk for others. The person chafing at mask-wearing may irrationally discount the data about the benefits of masks or may erroneously believe that masks deprive the wearer of oxygen. The mask-rejecter may confuse a momentary and minor inconvenience for an important value. People who reject any hint of paternalism as support for mask-wearing may contend that their rejection of paternalism is so important that it outweighs the non-paternalistic argument that masks protect others. In this vein, some who want not to wear masks have stated outright that they believe they have the right to place others at risk. For example, a Utah woman opposed to wearing masks stated, "I don't think it's the government's place to tell me I should or shouldn't wear a mask. And if I want to take that risk or if I choose to put people at risk, that should be up to me" (Walker 2020).

Straight-out forms of hard paternalism must reject liberalism. If hard paternalism is required to defend surveillance, it must ultimately substitute judgments about their good for people's own determinations. However, hard paternalism may not be required to support much surveillance. Non-paternalistic justifications are available, such as consent, prevention of public nuisances, and equity. Softer paternalistic justifications are also available, such as exploring how providing people with information may help them to realize their own values in the longer term. These softer informational considerations may intertwine with non-paternalistic justifications such as fairness to others who need the information to pursue their conceptions of their good. Some who object to hard paternalism have insisted on informed consent for individuals to be required to share information about themselves for purposes of public health surveillance. In the next chapter, we explore how liberal values of respect for individuals as sources of their own good can be reflected in surveillance even without relying on models of individual informed consent.

Nudge" in the Clinical Consultation-An Acceptable Form of Medical Paternalism?

The Problem for Children in America

Born To Win, Schooled to Lose: Why Equally Talented Students Don't Get Equal Chances to Be All They Can Be

Rising Morbidity and Mortality in Midlife Among White Non-Hispanic Americans in the 21st Century

Centers for Disease Control (CDC). 2020. National Diabetes Statistics Report

Reconsidering the Politics of Public Health

Coercive Paternalism in Health Care: Against Freedom of Choice

The Major Causes of Death in Children and Adolescents in the United States

Department of Health and Human Services (DHHS). 1979. Healthy People: The Surgeon General's Report on Health Promotion And Disease Prevention

What is the U.S. Opioid Epidemic?

Reportable Diseases and Conditions

Facing Up to the Prescription Opioid Crisis

2017. Paternalism. The Stanford Encyclopedia of Philosophy

The Nutrition Connection: Why Doesn't the Public Know?

Trends and Patterns of Differences in Infectious Disease Mortality Among US Counties, 1980-2104

Let the Shoemaker Stick to His Last: A Defense of the "Old

Regan, Congress, and Health Spending

Socioeconomic Status and Health: The Potential Role of Environmental Risk Exposure

The Politics of Public Health: A Response to Epstein

Infectious Disease Mortality Trends in the United States

Electronic Health Records and Research: Privacy Versus Scientific Priorities

re National Prescription Opiate Litigation, 2020 WL 582151 (United States Judicial Panel on Multidistrict Litigation

Surveillance for Viodent Deaths-National Violent Death Reporting System, 27 States

Changing the Culture of Health: One Public Health Misstep at a Time

What Do Workplace Wellness Programs Do? Evidence from the Illinois Workplace Wellness Study

The Social and Economic Correlated of Pregnancy Resolution among Adolescents in New York City, by Race and Ethnicity: A Multivariate Analysis

Obesity, Paternalism and Fairness

The Worldwide Opioid Epidemic: Implications for Treatment and Research in Pregnancy and the Newborn

Master Settlement Agreement

Opioid Epidemic in the United States

Improving Patient Comprehension of Literature on Smoking

Available from Project Gutenbert ebook#34901

Native Village of Kivalina v. ExxonMobil Corporation, 696 F.3d 849 (9 th Cir

The New Public Health and WHO's Ninth General Programme of work: A Discussion Paper. Geneva: World Health Organization

Smoking by Blacks and Whites: Socioeconomic and Demographic Differences

Public Health: Ethical Issues

Soft Power: The Means to Success in World Politics

Office of Disease Prevention and Health Promotion (ODPHP). 2020. Healthy People 2020 Leading Health Indicators: Progress Update

A Flood of Opioids, a Rising Tide of Deaths

Populations, Public Health, and the Law

People v. ConAgra Grocery Products Co

Sick Individuals and Sick Populations

Is Deidentification Sufficient to Protect Health Privacy in Research?

Effect of a Workplace Wellness Program on Employee Health and Economic Outcomes: A Randomized Clinical Trial

Lead Industries Association, Inc., 951 A.2d 428 (R.I

Wife's Level of Education and Husband's Risk of Primary Cardiac Arrest

Nudge: Improving Decisions About Health, Wealth, and Happiness

The New Public Health: An Introduction for the 21 st Century

What is the "New Public Health"?

About the Sustainable Development Goals

Here's why Utahns do or don't wear a mask during the pandemic and what physicians have to say about it

Payment for research participation: A coercive offer

Rethinking the New Public Health

The Struggle for the Soul of Public Health

Adventures in Nannydom: Reclaiming Collective Action for the Public's Health

Towards a Normative Framework for Public Health Ethics and Policy

Social Inequalities in the Decline of Coronary Mortality

Dementia: Number of People Affected to Triple in Next 30 Years

Understanding the Relationship Between Education and Health