Three Rationales for Vaccine Mandates
by Michael C. Dorf
Yesterday SCOTUS heard expedited challenges to the Biden administration's OSHA vaccine rules for workplaces (transcript here) and to healthcare worker vaccine rules for Medicaid/Medicare recipients (transcript here). I don't want to say there is nothing to the challengers' arguments. Surely they're right that it's a bit odd to use number of employees as a proxy for COVID spread risk. Hundreds of truckers each in their own long-haul cabs but working for the same company face substantially lower risk than 99 workers on a crowded factory floor; yet the OSHA emergency rule applies to the former but not the latter.
Still, I don't think that the lack of precise tailoring renders the rule arbitrary and capricious, given the deference ordinarily accorded administrative agencies. Nor are the challengers' other arguments at all good.
OSHA is empowered to protect workers' health (the word is right there in the name Occupational Safety and Health Administration), which does not mean it can't protect workers from risks that also exist "out there in the world" (as Chief Justice Roberts put it) if the workplace heightens them (or maybe even if it doesn't).
In any event, my main takeaway from the argument was rare sympathy for Justice Breyer's rambling. Although he has lately taken to rambling in every oral argument, he was understandably tongue-tied when, early in the OSHA argument he asked incredulously whether, in light of the Omicron surge's impact on cases and hospital capacity, the challengers were really asking for emergency relief at this moment. It's one thing when Tucker Carlson or Senator Ron Johnson talks as though a vaccine mandate, rather than COVID itself, is the emergency. It's quite another to hear that argument taken seriously by the Supreme Court.
And yet, I agree with Amy Howe (on SCOTUSblog) and other observers who think the Court is likely to invalidate at least the OSHA rule. I might have more to say about the issues of statutory construction and administrative law after the decisions are issued, but for today, I want to say a little bit about three different kinds of rationales for a vaccine mandate.
I hasten to add that the Biden administration's rules aren't even vaccine mandates. They're conditions and choices (although the challengers in the OSHA case say that the testing alternative isn't real right now because of the scarcity of tests). In any event, I want to analyze a genuine mandate of the form everyone without a very good medical excuse (like a documented vaccine allergy) must be vaccinated. I do so to clarify the government interests that underwrite vaccine mandates. I'll use categories drawn from constitutional law because they are familiar and helpful, but my main interest is how to balance the underlying values as a matter of policy (which, as Prof Segall would likely remind us, is also what's going on in the constitutional analysis).
(1) Paternalism. In a less libertarian society than ours, paternalism would be a sufficient basis for a vaccine mandate, just as it is a sufficient basis for laws that require car drivers and passengers to wear seat belts, workers to use safety goggles when operating certain machinery, and motorcycle riders to wear helmets. To be sure, many jurisdictions eschew some such obligations on libertarian grounds. Thus, only 15 states and the District of Columbia require all motorcycle riders to wear helmets; most other states mandate helmets only for minors. Still, paternalism is a generally acceptable justification for regulation. Or as Holmes famously put the point in his Lochner dissent, "[t]he liberty of the citizen to do as he likes so long as he does not interfere with the liberty of others to do the same [is] a shibboleth . . . ."
Yet while paternalism may be a morally and constitutionally adequate justification for most laws, it isn't an adequate justification for all laws. In particular, even in cases denying relief to the particular plaintiffs in right-to-die cases, the Supreme Court has recognized a constitutional dimension to the common law right to refuse medical treatment. If--as the Court assumed and five Justices specifically averred in the Cruzan case--a competent adult can refuse lifesaving medical treatment in the form of a feeding tube and intravenous hydration, it seems to follow that such a person can also refuse potentially lifesaving preventative medical interventions like a vaccine.
But Cruzan didn't purport to overrule Jacobson v. Massachusetts, the 1905 case that upheld mandatory smallpox vaccination--although some libertarians suggest that Jacobson may no longer be good law. Are those libertarians right?
It's possible that the current Supreme Court would say they are, but that's not a necessary conclusion. Jacobson can be readily reconciled with the modern case law with the following proposition: Government generally may not invade an otherwise competent person's bodily integrity in the name of that person's own wellbeing but may do so to further some very substantial (or compelling) other-regarding interest. And indeed, the Jacobson Court discussed the state's interest as one in preventing the spread of smallpox, i.e., it focused on the danger to others from failure to vaccinate, not on the danger to the unvaccinated individuals themselves. Modern case law pretty clearly accepts that line. For example, Breithaupt v. Abram and Schmerber v. California allow the state to insert a needle into an unwilling person's arm to draw blood to test for drunk driving.
(2) Herd immunity.
When the FDA and CDC initially approved COVID vaccines for emergency use, there were data to support the conclusion that they were safe and effective at preventing severe disease but it was not known whether they would prevent transmission. Once the vaccines were in wider use it became clear that with respect to the initial strains and the Delta variant, vaccination not only reduces the likelihood that a person will become seriously ill from COVID but also reduces the likelihood that one will spread it. So long as that was true, vaccine mandates could be justified on roughly the same grounds as the smallpox vaccine mandate in Jacobson: going unvaccinated puts others at risk, not just oneself.
Protecting others was (and remains) important because some people either cannot be vaccinated for medical reasons or do not develop a robust immune response. Accordingly, they depend on herd immunity to protect them.
But now we come to Omicron. Although it is too early to tell for sure, it appears that existing mRNA vaccines, if boosted, are pretty effective at preventing severe disease but perhaps less effective than against previous variants at preventing spread from a vaccinated but infected person. I express that view cautiously, to be sure. During the oral argument in the OSHA case, Justice Sotomayor "quibble[d]" with the challengers' lawyer's characterization of the rule as serving to protect "the unvaccinated worker." She cited the risk of spread to vaccinated persons with pre-existing conditions or compromised immunity. And she may well be right, in which case protecting the "herd" remains a powerful other-regarding ground for a mandate, even in the face of Omicron.
However, for my purposes, I'll now assume for the sake of argument that, in the face of Omicron, the vaccines aren't at all effective at preventing spread. Even if that assumption turns out to be false, that is, even if vaccination somewhat reduces the risk of spread, it might not make a whole lot of difference, because during the current extremely widespread prevalence of Omicron, the marginal contribution to risk of spread from being unvaccinated is likely very small. In any event, whatever the precise numbers turn out to be, I want to assume all of that away in order to explore another possible justification for mandating vaccination.
So, consistent with data from the UK, let's suppose that, other things being equal, if one is unvaccinated, one is three times more likely to end up in the hospital with Omicron than if one is fully vaccinated and boosted, and let's also assume that being vaccinated and boosted does not substantially diminish the likelihood that one will spread Omicron. The question I wish to pose is whether the government's interest in preventing the hospitals from becoming overwhelmed, standing alone, suffices to mandate vaccines. Put differently, is keeping people out of the hospital so that they don't tie up resources an other-regarding purpose that justifies overriding the bodily integrity interest?
(3) Limit resource consumption.
At first blush, the answer seems to be yes, of course. Especially now, with the anti-COVID drugs in short supply, every person who could have avoided hospitalization by getting vaccinated and boosted but didn't do so imposes a cost. Worse, that cost will be felt mostly by people who are vaccinated, because the limited drug supply will be prioritized for the sickest patients, i.e., those who are not vaccinated. So if Joe doesn't get vaccinated, he doesn't much threaten Mary via transmission, but he does threaten her by needing a scarce dose of Paxlovid or monoclonal antibodies that will now be unavailable to her.
Yet if avoiding the need for scarce medical resources counts as an other-regarding reason, how do we distinguish all sorts of other diseases and medical conditions that are the result, at least in part, of choice? If Sally goes sky-diving and risks breaking limbs, she may end up delaying medical treatment for someone else who ends up in the ER at the same time as she does. If Phil drinks too much before becoming sober, he might damage his liver so much that he needs a transplant that bumps someone else down the list. (Most transplant centers will provide livers to persons whose liver damage results from prolonged alcohol abuse if they have abstained from alcohol for at least six months prior to transplant.) If Andrea works non-stop as a lawyer, the stress on her heart could cause illness that uses up resources that could have gone to someone else. Steve's habit of eating Big Macs and fries could contribute to all manner of ill health that adds to the strain on health resources. Etc.
However, in none of the cases would we want to say that the people acted badly towards others by risking their own health. Some may have had no choice or had their choices severely constrained by addiction, genetic predisposition, or circumstances. A society that blames people for ill health or other medical misfortune because they tie up scarce resources is, not to put too fine a point on it, a cruel society.
Is refusing to get vaccinated different from sky diving, excess alcohol consumption, workaholism, and poor diet? Well, yes, obviously, because of the fact that contagious diseases are contagious. Remember, however, that for purposes of this third rationale I'm temporarily assuming away the herd immunity interest. Even so, I think we might distinguish those other behaviors on two grounds.
First, the burden of getting a free vaccine is much lighter than the burden of dieting (which is psychologically and physically difficult for nearly everyone), giving up alcohol (if one is addicted), etc. Whether that difference suffices is not clear, however, in light of our tradition of giving greater weight to libertarianism when it comes to refusing intrusions into the body than with respect to other constraints on liberty. That, after all, was the rhetorical value of the (highly misleading) invocation of the specter of mandatory broccoli consumption in the first Obamacare case. I imagine that carved into the stone above the archway of the Cato Institute is the slogan "If government can force a vaccine into your arm, it can shove broccoli down your throat."
So let's consider a second difference between those other strains on the health care system and going unvaccinated during a pandemic: those other strains are what we might call normal. Now that's not true in all respects. Where there is a shortage--as with organs for transplant--one person's decisions about their own health really do affect others. But for the most part, actions that lead to a need for health care in normal times are non-rivalrous. When Sally shows up at the ER with broken bones from her sky-diving accident, she might delay someone else who is there for 15 minutes, but she and the other sky divers do not lead the hospital to adopt crisis standards of care or ration lifesaving medication.
The capacity of our health care system is not static, but it cannot be dramatically ramped up in all needed respects over the course of days, weeks, months, or even a couple of years. That is why we hear about ICU and hospital capacity. If we lived in a society in which we routinely needed twice as many hospital beds and health care workers than we currently have, some combination of government policy and market forces would eventually increase both. Perhaps in such a society, vaccine hesitancy would not be properly categorized as harming others--at least when we indulge my arguendo assumption that the only risk is using resources, not disease spread.
For now, however, we live in a severely health-resource-constrained world. Quite apart from the risk they pose to family members, co-workers, and strangers to whom they may spread the virus, the unvaccinated impose a grave risk by greatly increasing the odds that they will need to consume scarce health care resources.
* * *