A public health framework for COVID-19 business liability Open Access

I. INTRODUCTION

Coronavirus disease 2019 (COVID-19) has visited profound consequences upon virtually every area of American life, including the tort system. Although the ‘tidal wave’¹ of tort claims that some predicted at the outset of the pandemic has not yet arrived,² the novel coronavirus already has generated a storm surge of legislative activity on the subject of tort reform. As of this writing, four states—Iowa, Louisiana, North Carolina, and Utah—have enacted laws providing businesses with sweeping protection against tort liability for coronavirus exposures.³ Several others have passed narrower measures to limit the liability of health care providers or to shield manufacturers and distributors of personal protective equipment from product liability claims.⁴ All in all, more than one hundred bills to expand or restrict tort liability had been introduced in state legislatures by the beginning of August.⁵ Congress, for its part, remains deadlocked⁶ over a Senate Republican proposal to exempt businesses from liability for coronavirus exposures except in cases of gross negligence or willful misconduct.⁷

Tort liability for coronavirus exposures has important implications for both economic vitality and public health. These two values are often characterized as conflicting,⁸ though in a pandemic, they are just as often complementary. Until the coronavirus is contained, it is difficult to imagine the US economy escaping from its most severe crisis since the Great Depression.⁹ And evidence from past recessions suggests that a long-lasting economic downturn could translate into a significant rise in mortality rates.¹⁰

At its best, tort law can advance economic and public-health objectives. This is especially true in environments of asymmetric information, where customers cannot easily verify whether businesses have undertaken appropriate precautions. Consider, for example, a barbershop (or equally, hair salon). A potential patron can easily observe whether a barber is wearing a facemask while cutting hair, but cannot so easily ascertain whether the barber checked her temperature that morning or properly sanitized surfaces and equipment before the customer’s arrival. Liability builds a bridge over that information asymmetry. To be sure, the customer still cannot verify the barber’s precautions directly, but the customer does at least know that the barbershop has a financial incentive to take precautions in order to avoid ultimate payment of damages.¹¹ Not only might this boost consumer confidence (thus contributing to economic recovery), but it also can encourage businesses to undertake the difficult-to-verify precautions that will slow the coronavirus’s spread.

In some cases, however, tort law can be a hindrance to coronavirus containment rather than a help. To continue with the barbershop example, imagine that a barber tests positive for COVID-19 a day after she cut several clients’ hair. Ideally, the barbershop owner would reach out to the positive barber’s recent clients and alert them of a potential coronavirus exposure, thus allowing them to self-isolate and seek testing themselves. With fears of liability looming large though, the owner of the barbershop might balk at this step, worried that those telephone calls and emails to customers could be invitations for lawsuits. After all, if the barber and the barbershop owner stay mum, customers who frequent the shop only once every two months might never learn about the site of their exposure (and thus, might never sue). The course that the barbershop might take in order to shield itself from liability is exactly the opposite of what, from a public health perspective, society should want it to follow.

The example above illustrates a familiar tension in tort law between ‘ex ante’ and ‘ex post’ incentives. Ex ante (ie before an injury), the threat of tort liability generates incentives for potential injurers to take cost-justified precautions. Ex post (ie after injury), the effect of liability may be reversed: fear of lawsuits can deter injurers from taking socially beneficial steps because those steps may increase the risk of personal-injury claims or the expected magnitude of tort judgments and settlements. Scholars have noted this phenomenon in a number of contexts. For instance, James Henderson has examined tort law’s potentially perverse ex post effect on product safety, pointing out that liability may deter manufacturers from conducting routine safety reviews and implementing safety improvements lest those efforts be cited as evidence that the original product was unsafe.¹² Omri Ben-Shahar has considered the effect of product liability on recall decisions, highlighting the concern that manufacturers may hesitate to recall defective products because a recall announcement is often ‘taken as a public “confession” that the product is harmful’, thus attracting the attention of injured users and personal-injury lawyers.¹³ And a number of authors have highlighted a similar dynamic in the medical malpractice context, where liability fears may discourage health care providers from informing patients about medical errors.¹⁴

This article considers the ex ante and ex post effects of coronavirus exposure liability, drawing lessons from other tort law contexts while also accounting for unique features of the COVID-19 challenge. We begin by surveying the existing legal landscape with respect to tort liability for businesses whose operations expose customers and workers to the risk of contracting COVID-19. We then present a framework for evaluating the public health consequences of various liability rules. Our analysis leads us to propose concrete changes to tort law designed to align business incentives with public health imperatives. We also suggest modifications to workers’ compensation systems aimed at encouraging postexposure precautions on the part of employers and employees.

To preview: Our proposal would provide businesses with a safe harbor from liability to customers arising out of coronavirus exposure if the business informs those customers via telephone, email, or text message within 24 hours of learning about the potential exposure. As explained below, our proposal would not operate as a get-out-of-liability-free card: it would apply only to instances in which businesses proactively initiated postexposure communication with customers, and it would not apply to claims of gross negligence, recklessness, or willful or intentional exposure. But it would, we suggest, mitigate the potentially perverse ex post consequences of coronavirus exposure liability without entirely sacrificing significant ex ante benefits.

II. CURRENT LEGAL LANDSCAPE

In the USA, the scope of business liability for injuries to customers is defined primarily by the laws of the 50 states and the District of Columbia. In all jurisdictions, absent a specific carveout, customers injured as a result of a business’s operations can bring claims for the tort of negligence. A business generally will be liable for negligence if it breached its duty of ‘reasonable care’ to customers and a customer suffered an injury caused by that breach.¹⁵ That standard, though, is easier to recite than to meet.

Customers bringing negligence claims in the COVID-19 context will face several steep obstacles. The first is to show that the business (or, under the doctrine of respondeat superior, one of its employees acting within the scope of employment) has breached a duty of care. This may be straightforward in some scenarios—for example, when a plaintiff can establish on the basis of eyewitness evidence that restaurant waitstaff members failed to wear facemasks or other nose-and-mouth coverings while serving diners. But in other contexts, it may be substantially more complicated. For instance, does a railroad breach its duty of care if it fails to enforce a face-covering requirement for passengers on an intercity passenger train, or are exceptions (eg for eating and drinking) appropriate? Custom, statute, and regulation are largely uninstructive,¹⁶ and so it generally will be left to individual judges and juries to decide what reasonable care requires in particular fact patterns.¹⁷

Establishing that a business has breached its duty of care is just the beginning. Perhaps the most challenging element of negligence for a plaintiff to prove in a coronavirus exposure lawsuit is factual causation: the plaintiff must prove by a preponderance of the evidence (‘more likely than not’) that her injury was caused by the defendant’s breach of the duty of care. Consider a customer who alleges that she contracted COVID-19 as a result of a restaurant server’s sneeze.¹⁸ The customer would bear the burden of proving, to a judge’s or jury’s satisfaction, that the server’s sneeze—rather than an interaction with a relative, co-worker, acquaintance, employee of another business, or sidewalk passerby—was the cause of her infection. Causation will be especially difficult to show in settings where COVID-19 is endemic and potential transmission pathways are manifold.

To be sure, there will be some circumstances—say, a cluster of cases linked to individuals who dined indoors at the same restaurant on the same evening in an area where prevalence is otherwise low—in which contact tracing will allow a customer to prove by a preponderance of the evidence that transmission at a particular business led to her infection. But we expect there will be many other instances in which infected individuals have no more than an inkling as to where they might have been exposed. A multistate telephone survey of 350 adults who had tested positive for the coronavirus found that 54 percent were not aware that they had been within six feet of another COVID-19 patient in the two weeks preceding illness onset.¹⁹ And of those who could identify a recent contact with another person whom they learned was COVID-19-positive, it is far from clear that they would have been able to establish that fact (or the fact of transmission) to a court’s satisfaction.

Even when a customer-plaintiff can establish breach of duty and causation, significant additional hurdles remain. In some cases, courts may honor liability waivers that claim to block customers from bringing claims against businesses, though most states impose limits of various sorts on the enforceability of waivers²⁰ (and two states—Louisiana and Virginia—refuse to enforce them in virtually all circumstances²¹). In other cases, courts may conclude that the plaintiff implicitly ‘assumed the risk’ of contracting COVID-19—and thus forfeited her right to recovery—because she participated in an activity with a known risk of transmission (eg a group fitness class at a gym).²² In still other cases, a judge or jury may find that the plaintiff failed to take reasonable self-protective measures (eg wearing a facemask or regularly washing hands), and thus was party at fault for her own infection. Under the comparative negligence doctrine followed by the overwhelming majority of states, a plaintiff who is partly at fault for her own injury will have her damages award reduced. In four states (Alabama, Maryland, North Carolina, and Virginia) plus the District of Columbia, a plaintiff who is partly at fault for her own injury will be barred from recovery entirely.²³

Tort claims arising out of virus exposures in the workplace face all of the above obstacles plus one more: every state and the District of Columbia follow some version of the workers’ compensation exclusivity doctrine. That doctrine forms one side of the so-called ‘grand bargain’ of workers’ compensation, under which employers must provide no-fault insurance for employees injured on the job and employees generally lose the right to recover in tort.²⁴ The workers’ compensation exclusivity doctrine does not apply to independent contractors or to employees injured by a defendant other than her employer. Nonetheless, the doctrine dramatically limits the set of COVID-19 cases to which tort liability would otherwise apply.

Given the formidable legal obstacles facing customers and workers seeking to sue businesses for COVID-19-related injuries, one might wonder why liability issues suddenly have taken center stage in political debates. One reason may be that COVID-19 liability issues intersect with a long-running, ideologically freighted battle over tort liability more generally. Like other COVID-19-related controversies over abortion clinic access, gun store openings, and religious exemptions from shelter-in-place orders, the tort liability debate can be understood only against the backdrop of the decades-old disputes that preceded it.²⁵

A second answer is that in some sectors, causation may be substantially easier to establish, and assumption-of-risk and comparative defenses may be less robust. This is particularly so for the nursing home industry, which has been uniquely hard-hit by COVID-19. As of the end of July, more than 40 percent of US COVID-19 deaths were linked to nursing homes,²⁶ which are simultaneously suffering from reduced revenue (due to fewer admissions) and increased costs (due to added expenses of personal protective equipment and other safety precautions). The threat of nursing-home bankruptcies has public health implications of its own. When an enterprise is deeply under water, its owners have an incentive to make risky gambles that they would forgo if solvent (eg to cut spending on safety and to cram more patients into close quarters).²⁷ Moreover, the specter of bankruptcy may make it harder for nursing homes to raise capital in the short-term for socially desirable safety investments.²⁸ For all of these reasons, there may be a case for nursing home-specific relief that is broader than the safe harbor we propose for businesses generally. We set aside this issue not because it is unimportant, but—to the contrary—because it is sufficiently consequential and complicated to deserve its own careful treatment.²⁹

But even where legal hurdles facing customers and workers are daunting, some plaintiffs likely will succeed in securing large judgments or settlements. Intense media coverage of occasional outsized verdicts may amplify the behavioral effects of those recoveries. Liability fears may be exacerbated further by the fact that some businessowners’ insurance policies explicitly exclude injuries attributable to viruses.³⁰ For businesses whose insurance coverage excludes communicable disease transmission, the consequence of a coronavirus exposure claim is that the business will be left on its own to litigate and potentially settle or pay damages on the claim. Until a vaccine emerges or a comprehensive testing-and-tracing scheme succeeds in containing the virus, business owners may continue to live under a cloud of concern about potentially crushing liability.

For many businesses—and especially smaller businesses with shallower pockets—anxieties regarding tort liability are not new. But the current COVID-19 pandemic almost certainly makes those anxieties more intense. Many businesses—especially bars, restaurants, and other enterprises in which customers are in close physical contact with employees and fellow patrons—already are suffering devastating economic losses.³¹ These losses are becoming harder to sustain the longer that COVID-19 keeps customers away. Liability insurance generally provides less than complete coverage, even under ordinary circumstances.³² And in the context of a communicable disease that may be excluded from insurance coverage, a lawsuit (even one that is ultimately unsuccessful) can be an existential threat to a business’s survival.

At the same time, a case of COVID-19 can cause financial devastation for the individuals and families on the plaintiff side of exposure-related lawsuits, especially where an infection results in the death or long-term disability of a breadwinner. These predicaments ought not to dissuade us from addressing liability from a public health perspective. Rather, they underscore the importance of analyzing the public health consequences with due attention to the unique economic considerations at play.

III. EX ANTE VS. EX POST

Debates over business liability for coronavirus exposures so far have been framed largely in now-familiar terms of restarting the economy versus protecting public health.³³ As noted at the outset, those two values are not as opposed to each other as that frequent framing might suggest. Nonetheless, designing (or redesigning) a liability regime for coronavirus exposures entails difficult tradeoffs. We focus here on the tradeoff between ex ante and ex post incentives, which—we argue—is central to the challenge that policymakers face.

Ex ante, tort liability encourages businesses to take additional precautions against the spread of the virus. On the margin, it raises the cost to a business of coronavirus transmission, and therefore increases the incentive for the business to take measures to control the virus’s spread. Some of these measures—facemask requirements for employees and customers, easy availability of hand sanitizer, stricter occupancy limits, spacing parties at least six feet apart—cut across contexts. Others are sector-specific. Dine-in restaurants, for example, might shift to curb-side pickup and delivery or move tables outdoors.³⁴ Fitness centers might erect shields between exercise machines or set up individual ‘pods’ that allow members to participate in group workout classes while separated by translucent plastic barriers.³⁵ Insofar as liability affects business behavior ex ante, the consequences are likely to be salutary from a public health perspective.³⁶

There is, to be sure, some risk that customers and workers will take fewer precautions on their own if they believe they can recover damages in the event of infection, but concerns regarding ‘victim moral hazard’ are attenuated for four reasons. First, comparative negligence principles operate as checks on victim moral hazard—many plaintiffs who fail to take reasonable precautions themselves can expect to see their recoveries reduced or eliminated. Second, in light of the aforementioned legal obstacles to recovery, risk-averse customers and workers may accord little weight in their own behavioral calculations to the relatively low probability of a tort award. Third, most people do not want to contract COVID-19 or to spread it to others and will take precautions against infection and contagion regardless of the potential for tort recovery.³⁷ Fourth and finally, some relaxation of customer and worker precautions is likely optimal (and thus not, technically, a ‘moral hazard’³⁸). By encouraging businesses to create safer environments, liability can induce customers and victims to forgo the ultimate precaution: staying home.

The ex post effects of business liability are less straightforward. Consider a restaurant owner who learns that a server has tested positive for COVID-19. The specter of liability may motivate the owner to alert customers who came into close contact with the server in the last several days (eg by matching receipts to telephone numbers and email addresses in the restaurant’s reservation system). The owner may anticipate that if she fails to alert customers of a potential exposure and they subsequently find out, then they may cite the owner’s inaction as evidence of negligence. But it is likewise possible that liability will have an opposite effect: the owner may choose not to alert customers because she fears that customers—if they knew the source of their infection—would sue. Customers then may fail to take precautions such as isolating themselves or obtaining a COVID-19 test because they do not know they have been exposed. This lack of knowledge may exacerbate the virus’s spread.

The failure of the federal government and most states to implement comprehensive contact-tracing systems strengthens our concern about the ex post consequences of tort liability for coronavirus exposure. As of early August, according to an NPR analysis, only three states and the District of Columbia had hired a sufficient number of contact tracers to meet estimated need.³⁹ If postexposure outreach by businesses were supplementary or superfluous to robust state-run contact tracing efforts, the potentially perverse ex post consequences of liability might be of reduced importance.⁴⁰ When business outreach is the primary mechanism through which customers and employees will learn about exposures, though, the risk that liability will undermine incentives for outreach becomes more worrisome.

Note that the tradeoff between ex ante benefits and ex post costs of liability is not new to the COVID-19 context. The tension between incentivizing precautions before injury and potentially discouraging mitigation efforts after injury is a familiar from other contexts. As noted above, other scholars have highlighted similar concerns with respect to product liability, where tort law may deter manufacturers from implementing safety improvements or initiating recalls, and with respect to medical malpractice, where liability concerns may deter health care providers from alerting patients and their families to errors.⁴¹ Anup Malani and Ramanan Laxminarayan have noted an analogous dynamic at the international level: countries are reluctant to report outbreaks inside their borders because they fear the trade and tourism consequences. Malani and Laxminarayan suggest that international organizations such as the International Monetary Fund and World Bank should provide aid to countries that report outbreaks so as to ‘reduce the burden or pain of sanctions after disclosure’.⁴²

We consider a similar approach in the context of COVID-19—though at the level of firms rather than nation-states. ‘Disclosure’ (here, informing customers that they may have been exposed to the coronavirus) potentially triggers ‘sanctions’ in the form of tort claims. In the next part, we explore strategies for reducing the sanctions that follow from promptly alerting customers to potential exposures.

IV. LIABILITY TO CUSTOMERS: A PROPOSAL

How should policymakers balance the likely beneficial ex ante public health effects of liability and the potentially perverse ex post consequences? We begin by considering the question of liability to customers before turning to the distinct problem of liability to employees.

One possible approach is to impose a duty to warn on businesses once they learn that a customer has been exposed to the virus through their operations. Such a duty almost certainly exists under common law: as summarized by the Restatement (Third) of Torts, when an actor’s prior conduct ‘creates a continuing risk of physical harm’, the actor ‘has a duty to exercise reasonable care to prevent or minimize the harm’.⁴³ In theory, the postexposure duty to warn should encourage businesses to reach out to customers because if they fail to—and a customer subsequently finds out—the customer will sue the business for breach of the duty. In practice, however, we expect that common-law duties will be insufficient to resolve the problem of ex post incentives—for three reasons.

First, the problem we have identified is not solely a problem of businesses learning about potential exposures but staying mum. It is also a problem of businesses failing to make investments that would facilitate postexposure outreach. Consider a restaurant or barbershop that accepts walk-in customers without a reservation or appointment. Optimally, the restaurant or barbershop would collect cell phone numbers from its walk-ins so that it would be capable of contacting them in the event of an exposure. But if maintaining that data simply exposes the business to liability for violating a postexposure duty to warn, then the business has much less of an incentive to collect the information in the first place. Whether a business is capable of contacting customers ex post is thus likely to depend upon difficult-to-verify investments that the business makes ex ante.⁴⁴

Second, and notwithstanding the existence of a postexposure duty to warn, some businesses may calculate that they are better off saying nothing (and risk being sued for violating two duties) rather than reaching out (and face a greater risk of being sued for violating one). This is especially true where state-run contact-tracing operations are threadbare. If infected customers are unlikely to learn about the site of their exposure absent affirmative outreach, then the threat of liability for violating the postexposure duty to warn is largely empty.

Third, even when an infected customer learns that a business has violated its postexposure duty to warn, it is far from clear that the customer will be able to recover for the duty-to-warn failure under general tort law principles. Typically, plaintiffs can recover only for damages that they have suffered as a consequence of the defendant’s breach. But of course, a defendant’s failure to alert a plaintiff of an exposure after the fact cannot be the cause of the plaintiff’s infection. In this respect, the application of the post-injury duty-to-warn doctrine to the COVID-19 context differs from other warning contexts, such as a physician’s duty to warn a patient of ongoing health risks from a flawed medical procedure⁴⁵ and a manufacturer’s duty to warn consumers of defects that come to the manufacturer’s attention after the time of sale.⁴⁶ In those other contexts, the person to whom the duty to warn is owed generally will be the same as the person who will be injured by the failure to warn. In the customer coronavirus exposure context, by contrast, the person most likely injured by the failure to warn is a contact of the exposed customer rather than the customer herself.

We note three important qualifications to the arguments in the previous paragraphs. First, and perhaps most speculatively, if an antiviral treatment becomes available that can stop the progression of the infection when administered before symptoms emerge, then an exposed customer may be able to argue that she herself was injured by a business’s failure to warn because—but for the failure—she would have received the antiviral treatment. There is, moreover, some evidence (primarily from rhesus monkey studies) suggesting that the drug remdesivir—if administered shortly after coronavirus exposure—can significantly reduce disease severity.⁴⁷ As of this writing, however, priority access to the limited supply of remdesivir is allocated to hospitalized patients who require supplemental oxygenation—not to individuals recently exposed.⁴⁸ Thus, a theory of injury along the lines of ‘if you had alerted me sooner, I would have taken remdesivir before the disease progressed’ remains highly implausible.⁴⁹

Second, and somewhat more plausibly, family members and other close contacts of exposed customers (second-degree exposures) who contract COVID-19 may bring failure-to-warn claims against businesses that do not reach out to customers (first-degree exposures), on the theory that—but for the business’s failure to warn—the exposed customer would have self-isolated and thus would not have exposed the close contact. Lawsuits arising out of second-degree exposures, though, will face especially high causation hurdles. The plaintiff will need to prove two links in the transmission pathway, when proving one link is already hard enough. Moreover, the plaintiff will need to show that her exposure occurred because of the failure to warn—in other words, that (a) she was exposed to the virus after the business, if it had fulfilled its duty, would have reached out to the exposed customer, and (b) the exposed customer would have self-isolated if she had been informed in due course.

Third, even if businesses face a low probability of being sued successfully for violating a common-law duty to warn customers postexposure, and even if compensatory damages in those cases are likely to be small relative to social costs, the threat of punitive damages may amplify the deterrent effect of the common-law duty.⁵⁰ The deterrence effect of punitive damages in this context, though, is limited by two factors. One is the ‘judgment-proof problem’.⁵¹ The amount of punitive damages that can be collected from any given defendant is effectively capped by the defendant’s total assets (and often by less than that amount, due to structures that confer limited liability). Another is US Supreme Court case law which strongly discourages punitive damages awards that are 10 times compensatory damages or more,⁵² and which prohibits punitive damages awards based on harms to third parties.⁵³

None of this is to suggest that the common-law duty to mitigate damage after injury should be eliminated or relaxed in the COVID-19 context. The existence of the duty may, at the very least, have expressive value even if it is unlikely to result in many successful claims.⁵⁴ But the postexposure duty to warn is, we believe, insufficient to resolve the problem of ex post incentives.

A potentially more promising approach is to create a liability safe harbor for businesses that contact customers within a certain window after the business learns of an exposure. We recommend a window of 24 hours from the time the business learned or reasonably should have learned of the potential exposure.⁵⁵ For example, once a server informs a restaurant that she has tested positive for COVID-19, the restaurant would have 24 hours to inform customers who came into contact with the server about the potential exposure. If the business spoke with the customer, left a voicemail message, or sent an email or text message during that timeframe, the business would fall within the scope of the safe harbor.⁵⁶ The safe harbor would be available only for 14 days from the time of exposure—a timeframe chosen to match the upper bound of the incubation period for COVID-19.⁵⁷ That is, informing a customer about an exposure more than 14 days after it occurred will not trigger the safe harbor (even if the alert comes 24 hours after the business learned of the weeks-ago exposure). Any customer who is not informed during the relevant timeframe would retain access to preexisting tort law remedies. We also suggest that the safe harbor be limited to claims of ordinary negligence, and thus would not apply to claims of gross negligence, recklessness, willful misconduct, or intentional exposure.

We do not expect that every business will make use of the safe harbor with respect to every potential exposure. Identifying and contacting potentially exposed customers will be relatively easy for some businesses (eg a restaurant that collects email addresses and telephone numbers via an app such as OpenTable). It will be nearly impossible for other businesses (eg a stall at a farmers’ market that takes payments in cash). The purpose of the safe harbor is to ensure that businesses that can identify and contact customers to alert them of potential exposures are not deterred by tort liability from doing so. If the business cannot identify and contact customers, then there is less risk that tort liability will lead to deleterious ex post consequences.

Such a safe harbor could, concededly, weaken the ex ante public health benefit of liability: businesses may take fewer precautions if they know that promptly alerting customers and workers afterwards will shield them from suit. But this concern is outweighed by three considerations. First, the availability of the safe harbor would be far from assured. If the business does not learn about the potential exposure within the 14-day window, it cannot exculpate itself from liability. Second, few reputation-conscious businesses will want to find themselves in a situation where they must inform customers about a potential exposure on their premises. Third, the safe harbor’s existence may have the ex ante (pre-exposure) benefit of motivating businesses to maintain more detailed records of their customers’ identities and contact information, thus facilitating postexposure outreach. A restaurant or barbershop that accepts walk-in customers, for example, could begin asking for those customers’ names and phone numbers (which, in turn, would allow the business to reduce its own liability risk in the event of a coronavirus exposure). Thus, while the existence of the safe harbor may marginally reduce the incentive of businesses to take some ex ante precautions, it may increase the incentive of businesses to take other ex ante precautions.

The safe-harbor approach has the virtue of administrative simplicity. No legislature or court would need to specify in advance which individuals a business must contact in the event of an exposure. Rather, the very fact that the business had contacted the customer within the 24-hour window after a coronavirus exposure would be a complete defense to a claim of ordinary negligence arising out of that exposure. This feature of our proposal means that the scope of the safe harbor would be robust to changes in public health guidance. For example, as of this writing, the CDC recommends that contact tracers reach out to any person who was within six feet of a COVID-19 patient for at least 15 minutes within the 48-hour period before that patient first experienced symptoms. (For COVID-19 patients who test positive but do not have any symptoms, the relevant 48-hour window is the 48-hour period before the test specimen was collected.)⁵⁸ This ‘6-15-48’ rule may well change over the course of the pandemic. The scope of the safe harbor, however, would remain the same. Hypothetically, if ‘6-15-48’ becomes ‘6-15-72’, and the business reaches out to customers who came into prolonged proximity of an infected individual in the three days before symptom onset or specimen collection, then the safe harbor would apply to any claim brought by those contacted customers. If the business reaches out to a narrower range of individuals than what the Centers for Disease Control and Prevention recommends, the scope of the safe harbor would be correspondingly narrower too.

The safe-harbor approach also has the virtue of political plausibility. With Republican lawmakers in Washington generally arguing for significant limits on liability and Democrats defending the status quo,⁵⁹ the safe-harbor proposal strikes a compromise between those polar positions. To be clear, this is not the primary purpose of the proposal: we seek to strike a balance between ex ante and ex post incentives, not a balance between the political left and right. Framed differently, our proposal does not compromise on public health, but instead aims to adjust the tort liability system so as to better promote public health. The safe harbor’s compatibility with political exigencies is a benefit, but not a motivation.

The mention of national politics raises the question of whether such a safe harbor, if adopted, should be adopted at the federal level or the state level. Historically, tort law has fallen within the domain of the states, and state-level experimentation can assist the development of innovative, evidence-based solutions. That said, federal preemption of state common-law tort law claims is not unusual,⁶⁰ and a congressional statute establishing the proposed safe harbor would likely lie well within Congress’s Commerce Clause powers given the significant cross-border effects of COVID-19.⁶¹ While we favor the state-by-state approach, we recognize that congressional enactment is no doubt the quicker route to the implementation of our proposal nationwide.

V. LIABILITY TO EMPLOYEES: TWO PROPOSALS

The proposal in the previous part addressed liability to customers. Liability to employees raises a similar but not identical set of challenges. In particular, any change affecting employer liability to employees for coronavirus exposure must account for the institutional details of state workers’ compensation regimes.

As noted, workers’ compensation arrangements are ‘no fault’: the employee’s ability to recover does not depend upon proving that the employer failed to exercise reasonable care. Workers’ compensation still can generate incentives for employers to take ex ante precautions against workplace injuries: employers that self-insure will want to reduce the frequency of injuries and resulting payouts, while employers that pay experience-rated premiums for insurance will want to reduce claims by their employees in order to keep premiums down.⁶² The ex ante effects of workers’ compensation on employee precautions potentially cut in the opposite direction, though again, moral hazard may be mitigated by self-preservation instincts.

Insofar as employers internalize the cost of injuries attributable to their workplaces through self-insurance or experience-rated premiums, employers have a financial incentive not to inform employees about potential exposures: if the employee does not know that she contracted COVID-19 as a result of a workplace exposure, then she may be less likely to file a workers’ compensation claim. This is not to suggest that all employers will hide the fact of a potential exposure from their employees. Some will act appropriately for moral reasons or to preserve employee goodwill. Others will act promptly in order to minimize the spread of the virus among their workforce or prevent employees from passing the virus along to customers (who then may file lawsuits of their own). Still, concerns about perverse ex post incentives are not trivial, especially amid allegations that employers in some industries (eg meat processing) were slow to warn workers about COVID-19 risks.⁶³

One approach to the workers’ compensation issue, already adopted by executive order in California, is to establish a presumption that any employee who tests positive for COVID-19 within 14 days of performing services for an employer was exposed to the virus on the job.⁶⁴ The presumption does not apply if the employee worked exclusively from home, and the presumption can be rebutted by the employer. Wyoming has adopted a similar rebuttable presumption via statute.⁶⁵ The Illinois Workers’ Compensation Commission withdrew an analogous measure after a defeat in state court.⁶⁶ Several other states have adopted presumptions of workplace exposure for healthcare and public-safety first responders or—more broadly—for all essential workers. Still more are currently considering presumptive-causation legislation.⁶⁷

Presumptive causation rules have nuanced ex ante and ex post effects. Ex ante, these rules strengthen employers’ precautionary incentives, as they increase the likelihood of liability in the event of a workplace exposure. Absent a presumptive causation rule, employers (and their insurers) may not internalize the full cost of workplace exposures because some exposures that do occur on the job will not result in successful claims. But because these rules also increase the probability that an employee will receive compensation (both for workplace and non-workplace exposures), presumptive causation rules can engender employee moral hazard. The worry is that employees will be less risk-averse in their daily lives—on and off the job—in the view that if they contract COVID-19 and experience severe outcomes, they can receive workers’ compensation benefits. (This concern is, again, mitigated by employees’ self-preservation instincts and their likely desire not to infect others.)

Ex post, the likely public health effects of presumptive causation rules are on the whole positive. These rules encourage employers to alert employees of workplace exposures (or, more precisely, lower the disincentive for disclosure), because now there is less to gain from suppressing such information: the employer is likely on the hook either way once the employee develops symptoms, and the employer’s best bet is to mitigate the risk of spread through its workforce. Perhaps most importantly, presumptive causation rules can encourage employees without paid sick leave to seek out testing and, if they test positive, stay home. Under a law like California’s, an employee who obtains a positive COVID-19 test within 14 days from her last day of service qualifies for the causation presumption, and the sooner she obtains a positive test, the sooner she will likely begin collecting workers’ compensation benefits.

Another approach to the liability-to-employees issue—which can be adopted in tandem with the presumption causation rule—is to avail a tort remedy to employees who contract COVID-19 as a result of an employer’s gross negligence while also allowing employers the same safe harbor that we describe above (ie no liability if the employer informs the employee before the earlier of 24 hours of learning about the exposure or fourteen days of the exposure’s occurrence). Note, though, that in most states, this would mark a significant break from the grand bargain: with rare exceptions, the workers’ compensation exclusivity doctrine applies even to gross negligence claims.⁶⁸

Importantly, both a presumptive-causation rule for workers’ compensation claims and a gross-negligence carveout from the workers’ compensation grand bargain could exist simultaneously. In our view, both have the potential to generate public-health benefits. Here, as is often the case, innovative states have taken the lead. Their early experiences are likely to inform efforts at legal change elsewhere.

VI. CONCLUSION

The analysis here suggests that a liability safe harbor for businesses that promptly alert customers to potential exposures can serve to balance the ex ante benefits of liability with possible ex post costs. Such a safe harbor can be supplemented by a presumptive causation rule in the workers’ compensation context, which would improve ex post incentives for employers and employees without radically disrupting the workers’ compensation ‘grand bargain’. A gross-negligence cause of action for employees exposed to the coronavirus on the job can serve an additional salutary function, though we acknowledge that there, the deviation from the grand bargain is somewhat sharper.

Most importantly, the analysis here underscores that a well-tailored liability regime is an important element of a comprehensive COVID-19 containment framework. Legal scholars and public health professionals have an important role to play in developing a set of rules that can harness the positive incentive effects of tort liability while minimizing the potentially perverse consequences. Investments in such a regime can yield dividends in the immediate effort to contain COVID-19 while also aiding in the future management of other infectious diseases.

Footnotes