Vaccination by Race, and Why It's Unconstitutional


The Department of Veterans Affairs will apparently be prioritizing "Black, Hispanic, and Native American" veterans for the coronavirus vaccine:

While we have a limited supply of COVID-19 vaccines, we'll offer vaccines to
Veterans based on their risk. In addition to the risk of getting infection, risk of
passing infection to others, and the risk to society if someone is unable to work, we'll
consider factors that may influence the risk of severe disease, including:
• Age. The risk of severe illness or death from COVID-19 increases with age.
• Race and ethnicity. Data shows that some groups of people have been
disproportionately affected by COVID-19. These include Black, Hispanic, and
Native American communities.
• Existing health problems. People with certain health problems (like cancer,
diabetes, or heart disease) have a higher risk of severe illness or death from
• Other factors that raise risk of severe illness or death from COVID-19, such
as living in a nursing home or other group living facility.

But this would violate equal protection rights. (The Equal Protection Clause by its terms applies only to states, but Supreme Court precedents have held that similar principles apply to the federal government, including in the distribution of government-provided benefits.) First, as Sally Satel (a scholar at AEI, and a visiting professor of psychiatry at Columbia University Vagelos School of Medicine) writes,

[L]et's consider what we know about risks to Blacks and Hispanics. Members of these groups are infected with the virus at three times the rate of Whites and die at least two times as often. Their risks of exposure are increased because they are more likely than Whites to work lower-paying jobs that require interaction with the public and to travel to those jobs by public transportation. Blacks and Hispanics are also more likely to live in homes with many family members sharing close quarters.

The National Academies—non-governmental institutions that offer expert advice on science policy—have proposed an allocation plan giving priority to communities that rate high on the Centers for Disease Control's Social Vulnerability Index, which takes into account poverty, unemployment and health-insurance rates, among other socioeconomic vulnerabilities. The index would be applied to each of several priority phases, the first being healthcare workers, the second being those who are medically at risk due to concurrent illness and age, and so on.

Since certain minorities are more likely to be socially vulnerable to infection with the virus—a status with roots in past discrimination—they will disproportionately receive the vaccine early under that approach, consistent with the public-health goal of maximizing communal benefit.

But a person's race, per se, does not put him or her at greater risk for becoming infected or dying from Covid-19. Race is a correlate but risk is the cause. Therefore, allocating the vaccine based on the moral impulse to correct past injustice would not maximize communal benefit.

And, as Hans Bader (Liberty Unyielding) notes, this means that the government must use such race-neutral alternatives, instead of using race as a proxy for risk, or using vaccination as a sort of group compensation for the disparate impact of the illness on certain racial groups. Just as the government is generally not allowed "to use race as a proxy for gang membership and violence" even in prisons, and just as "[r]ace cannot be a proxy for determining juror bias or competence," so race can't be used as a proxy for vulnerability to illness, especially when the other factors that Dr. Satel identifies can gauge risk better.

NEXT: UCLA Academic Freedom Committee on "Academic Freedom to Quote Offensive Source Material in Class Discussions"

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  1. Race is hard to measure. Crowded housing is not. Race is not a risk factor, but poverty and crowding are.

    1. There are more White people living in poverty than there *are* Black people. Just sayin….

      1. That’s perhaps my most ‘deplorable’ complaint, 13% of the population and 98% of the issues. Pareto’s 80/20.

        We stand on the shoulders of giants as the trolls chip away at their feet of clay.

    2. “But a person’s race, per se, does not put him or her at greater risk for becoming infected or dying from Covid-19.”

      “Race is not a risk factor,”

      Maybe true, maybe not true. There is definitely genetic differences in the races that affect susecptability to various diseases. A common example is sickle cell anemia. Likewise, asians (japanese, korean, etc) have had a much lower infection rate to covid that cant be solely explained by preventive behavior.

      To deny that genetic factors related to race play a role is to deny science. Similar to claiming that the genetic differences between male and female have no differences.

      1. Well, if this was an engineered bioweapon, you’d want to have it affect your people less…

        1. “Dr. Ed 2
          December.11.2020 at 2:37 pm
          Well, if this was an engineered bioweapon, you’d want to have it affect your people less…”

          In this country, “our people” includes every citizen of every race.
          Genetics, including those associated with race have an effect on diseases and treatment of diseases. As such, treatment should be adjusted as needed by racial genetic differences. Its based on hard science, not social science or political correctness.

          1. Please pay attention. If the virus was bioengineered (which Dr Ed 2 is not alleging, but the possibility is on everyone’s minds even if they won’t admit it) it wasn’t done by this country. So who this country’s people are is irrelevant to Dr Ed 2’s point.

            And the VA’s announced preference for non-whites is not based on hard science. The fact that race can play a role, and therefore under some hypothetical circumstance such a policy would be both constitutional and justified, is irrelevant to the current circumstances in which it is neither.

  2. “(The Equal Protection Clause by its terms applies only to states, but Supreme Court precedents have held that similar principles apply to the federal government, including in the distribution of government-provided benefits.)”

    OK, how does originalism handle those precedents?

    I happen to think the 14th Amendment (Sec. 1) should, by a further amendment, be “incorporated” against the federal government, but at present that’s not the case, at least under an originalist conception of things. Am I missing something?

    1. Cal Cetin: 1. I’m inclined to agree that a pure originalist (one who doesn’t care about precedent) should take the view that the federal government isn’t bound by equal protection principles as such — whether it involves discrimination against whites, against blacks, or whatever else. One might argue that certain deprivations of life, liberty, and property based on race and the like violate due process; but it wouldn’t apply to denial of discretionary government benefits.

      2. Very few originalists are pure originalists, though; most are willing to go along with sufficiently settled precedents, in part because it’s hard to run a legal system in which everything is always up for grabs. How to reconcile originalism and precedent is of course a famously complicated question, on which much ink has been spilled.

      3. Just to make clear (not that you necessarily doubted this), my posts about the law usually focus on what the law is under existing rules. They don’t talk about what the law ought to be under originalist principles (unless I expressly say that this is what I’m discussing). I’m a pragmatist on such matters: We have a current legal regime, whether right or wrong, and the first thing to do is to understand what that legal regime currently requires.

      1. Prof Volokh — wouldn’t the “nor be deprived of life, liberty, or property, without due process of law” clause of the 5th Amendment mandate “equal protection” as well?

        In other words, how can you have “due process” without concurrently having “equal protection” because inequality comes from the process not being followed. Or am I missing something here?

      2. “my posts about the law usually focus on what the law is under existing rules.”

        Unless you were recently confirmed as a judge, your posts focus on what the law ought to be on a particular question, given your reading of existing law. Vaccination by race is unconstitutional when a court(s) rules it unconstitutional. C.J. Marshall did not assert that “it is the province and duty of law professors and bloggers to say what the law.”

        1. Or, what the law is, in the educated opinion of a Professor of Law at UCLA.

  3. Maybe you should dig back to that post about a “point system” and re-hash this there?

    Or are we supposed to ignore that it’s precisely this kind of fine-scale prioritization that such a system would encourage, over the more general and broad categories the CDC recommended?

    1. I’m not sure I understand: Michael Abramowicz’s point-system post urged the vaccine to be rationed based on various factors — but that doesn’t resolve what those factors will be. This post notes that race isn’t supposed to be one of those factors.

      1. With the CDC’s broad recommendations, we don’t have cause or reason to look at such small-scale differences: under such a system the risk factors of a white grocery clerk vet are not appreciably different then the risk factors of a black grocery clerk vet.

        With a fine-tuned risk-based model, trying to figure out the demographic and behavior differences that increase (or decrease) risk is entirely appropriate. Seeing as we know that race is one such factor (and a fairly large one), this was entirely predictable. Seeing as we know there are people that are going to try and game any system to get their vaccine sooner, the lawsuit was entirely predictable.

        In short, a fine-tuned risk-based model, rather then a coarse risk-based model, is an invitation to both commit and complain about these kinds of issues. Pretending that you can create a fine-tuned model that doesn’t include such invitations is naive.

        1. My kingdom for an edit button. I don’t know if there is a lawsuit, I initially assumed as such because, well, it’s Volokh. But after double checking the OP I didn’t see clear text to that effect and scrubbed most references, but missed one in paragraph two.

        2. “With the CDC’s broad recommendations, we don’t have cause or reason to look at such small-scale differences: under such a system the risk factors of a white grocery clerk vet are not appreciably different then the risk factors of a black grocery clerk vet.”

          Concur I am okay with setting priorities based on genetics including genetics that are correlated with race.

          Your point is valid when the differences in outcomes is small, the additional costs of will quickly exceed any benefits. especially considering the speed in which we are trying get the majority of population vaccinated. With the aim of getting the majority vaccinated within 12 months, there is very little to be gained by trying to prioritize with the difference in outcomes is small. If covid was as bad as ebola, then it would be different story

        3. under such a system the risk factors of a white grocery clerk vet are not appreciably different then the risk factors of a black grocery clerk vet.

          Nor are they under any other system.

          With a fine-tuned risk-based model, trying to figure out the demographic and behavior differences that increase (or decrease) risk is entirely appropriate. Seeing as we know that race is one such factor (and a fairly large one), this was entirely predictable.

          The whole point of this article is that we know nothing of the sort. We know that black and Hispanic people have been hit disproportionately, but we’re also pretty sure we know why, and race doesn’t enter into it. The black grocery worker vet is at exactly the same risk as the white one. All the factors that have made the virus hit black people more than white ones are included in “grocery worker”.

  4. When I get my lab results for creatinine (for kidney function), one of the items listed is “RACE: Non Black”. Apparently, race matters when measuring kidney function. Is medical science violating the Constitution?

    1. That is not without controversy: See https://jamanetwork.com/journals/jama/article-abstract/2735726

      On the other hand, if you have a statistically significant variance in disease that you can not otherwise explain, what do you do?

      The larger issue is that these are statistical norms in the first place.

    2. According to the National Institute of Health, “Serum creatinine concentrations tend to be higher in black than white individuals and people of other races or ethnicities. These differences have been assumed to be largely related to race-related differences in body composition, especially muscle mass.” Thus race may affect a doctor’s determination whether the patient is suffering from kidney disease, and may thus affect his or her recommendations to the patient. I think it’s clear that Blacks are genetically more predisposed to certain conditions than Whites (e.g. sickle cell). If so, those conditions are unlike COVID-19, where the prevalence of the virus in the Black community is apparently attributable to socio-economic factors. I don’t know whether kidney function is such a condition, and perhaps the doctors themselves are unsure.

  5. You have to figure that the VA is damned if they do and damned if they don’t.
    Say they give out the vaccine by medical need first? Statistically that means that more whites would get it first. Well the media and the “Black and Latino Leadership” would be all over that screaming racism. Now what happens if vaccinated people start dying? The same thing. The media and the “Black and Latino Leadership” would scream racism and that minorities were being used as guinea pigs to test the vaccine on or to try and kill them off. When all you have is a hammer, everything is a nail. The name of the hammer is “racism”

    1. Well, the flip side of this is race-based lockdowns.

      If Blacks & Hispanics are three times more likely to be infected, then can we quarantine them without quarantining Whites.

      Beyond the visceral “you can’t do that”, how is it any different from making a similar distinction on availability of the vaccine?

      1. Hrm… makes me think of the Affirmative Action in Higher Education cases, actually.

        See, the nuance you glided right past is that vaccine availability isn’t based on race, even in the VA’s plan. It’s that race will be one of many factors. So just as the SCOTUS has said schools can’t say “these slots are saved for Black kids”, the SCOTUS is fine with schools saying “these slots will be doled out based on these fifteen criteria, of which race is one”.

        By the same token, while a “Black families need to quarantine, white families go on Spring Fling” policy is obviously going to be bad, a more nuanced “we’re assessing these fifteen risk factors, of which race is admittedly one, to determine which families need to quarantine” would probably be acceptable as long as the reasoning was sound.

        All that said, I think these attempts to more finely tune vaccine prioritization are counter-productive and over-thinking it.

        1. I imagine the Black response would be similar to the issue of racial discrepancy in K-12 suspensions.

          While race is not a criteria for being suspended, a disproportionate percentage of Black students are suspended, much like a disproportionate percentage of Blacks would be quarantined under your proposal. And your criteria isn’t even facially neutral which it is with suspensions (schools really don’t want to suspend anyone, it costs them money).

          You wouldn’t get away with it…

  6. I have heard that some states are intending to give priority to “under served communities” as well.

    I recall seeing a study by that paragon of virtue, the UKs NHS which controlling for wealth, education and a number of factors found POC were still more likely to be affected by Covid.

    I’ve often wondered whether by focusing on race we are disguising other underlying factors which could more equitably and rationally determine the impact of policies or health risks.

    I don’t doubt some of those factors may be social or cultural in nature.

    I’d point out that while a higher proportion of blacks are affected by Covid (or poverty or many other hot topics) more total whites total are affected.

    1. Massachusetts is giving priority to prisoners.
      Not everyone is happy about that…

      1. At least prisoners , like nursing home residents live in close proximity to many other people so there is some objective reason.

        1. More to the point, the state has custodial responsibility for them.

  7. I have trouble squaring your constitutional basis of equal protection used in giving preference on vaccinations, with judges refusal to support churches constitutional basis to worship.
    Seems recent rulings toss out constitutional protections against power hungry govt abuse, because pandemic. Prioritizing because of Race, because of pandemic, falls under the same exemptions.

  8. Suppose someone were to invent a treatment which would prevent the formation of human homozygotes for the sickle-cell hemoglobin gene. (Review: homozygotes have the disease; heterozygotes are carriers; if two carriers get together, each child they conceive has a one-out-of-four (25%) risk of being a homozygote and having the sickle-cell disease. I’m saying, suppose we had a preventive treatment to prevent the formation of homozygotes, so prospective parents could get treated and eliminate the risk of conceiving a sickle-cell patient.)

    Naturally you’d want to prioritize African-Americans, since the overwhelming majority of sickle-cell patients are African-American.

    Anyway, my question is: would a policy which gave AAs priority for getting the anti-sickle-cell shield, be constitutional? I ask because giving the anti-sickle-cell shield to non-AAs would be incredibly stupid and wasteful.

    1. It would be unconstitutional, because there’s a cheap test for the gene, so you don’t need to use race as a proxy.

  9. It’s an emergency, folks. Two points:

    1. Go ahead and find the “real” risk factors, the ones for which race is just a proxy. Measure them. Then invent a fine-grained approach to apply those metrics person-by-person, so medical authorities can have a ranked list of who to vaccinate, in what order. Use APA procedures to get it all done. Sounds like a plan, if not an emergency plan. So until that gets done, use race as a proxy and get going. It’s an emergency.

    2. The notion that advantaging proxy-identified, high-risk individuals unfairly rewards them at the expense of others . . . is questionable. Or maybe ridiculous. This time, we are talking about contagion, not just about individuals and what they get. With contagion, what they get, we all get. Pretty clearly, the most efficient way to protect everyone is to get the most virus exposure out the population, as quickly as possible. If proxies help you do that, and if that helps everyone, it ought to be legit to rely on proxies.

    More generally, treating atomized, individualistic ideology as an all-purpose policy panacea is tiring. Sometimes, it seems like folks who argue for individualism in everything are doing it with an eye to exhausting their rivals, as those strive to accomplish demanded impossibilities. Are we sure the law ought to tolerate so much deliberately-created fatigue, let alone enable and encourage it?

    1. And as to a race-based lockdown?

    2. It’s an emergency, folks. Throw the Japs into camps.

      1. Do you understand why Korematsu was wrong? Hint, it was not because emergency powers are wrong. Even Justice Jackson’s famous dissent did not assert that. Korematsu was wrong because there was not actually an emergency, as previously withheld military documents eventually disclosed.

    3. Prioritizing based on race should be a absolute No
      Prioritizing based in genetic factors is an absolute Yes.
      If there are genetic factors which are correlated with race, then prioritizing based on race is a definite yes.

      However, given the cost of testing for the genetic factor and the short time frame to get everyone vaccinated as quickly as possible, then using race as a proxy for the genetic factor is okay with me. With one caveat, justification for the prioritizing by race as a proxy would need to be based on a large enough delta on the IF or the IFR.

      When the oral polio vaccine was developed in the late 1950’s, distribution did not began until 1962/1963 with mass vaccinations at most towns local community centers, in my case the towns high school.

      1. Except that there is no evidence of a genetic factor here.

  10. “But this would violate equal protection rights.”

    Which doesn’t *automatically* mean it’s unconstitutional, though. If, in the time-exigent rollout of vaccines that will occur in the next few weeks, might it pass strict scrutiny? For how long?

    I agree it’s problematical. I’m not convinced that it wouldn’t pass muster in the short term, if used in the best-available-data context of “who do we prioritize in a national emergency situation?” Especially if the gov’t is simultaneously trying to take steps to get to the same result in a more race-neutral manner.

    Then, of course, there’s the problem that by the time a plaintiff with standing can be found and a court case filed, it will already [probably] be moot.

    1. God willing it would be moot by that point. If, in fact, those on the lower-risk end will not have the chance to be vaccinated till early Summer 2021, then your last point would not apply.

      1. H&HS official being interviewed said June (2021).

  11. “Data shows that some groups of people have been
    disproportionately affected by COVID-19.”

    Does it really? It strikes me that people take state data to determine whether one race or ethnicity is more susceptible but does that really reflect fact? When I looked at NY state data it certainly appears to show that minorities are at greater risk statewide but the cases were largely confined to much smaller areas. When looking at county data the cases quite accurately reflect the demographics of the particular county with no distinguishable bias toward any particular race or ethnicity.

    Yes, the counties in question have demographics that are disproportionately minority compared to the rest of the state. They also have considerably higher population densities. It seems clear that the disease spreads more easily with higher population densities regardless of the demographics of those populations.

  12. The VA is run by Federal Union members, thus the inability to grasp that the factors that may impact most ethnic minorities would not be in place for veterans who are receiving care as a part of the contract they entered into upon enlisting or taking commission. And, thus the ignorant woke philosophy in general. In my experience of several decades, the VA has good care providers; the administrators and management would be happier were there no healthcare -just a government administration to benefit federal union members. I noted this policy yesterday, and was discussing it this morning. With luck it is withdrawn by next week as another stupid racist effort by a largely ignorant and stupid group.

Please to post comments