Readers of this blog know Kyle and I took turns in 2020 dancing with reason on COVID-19.
Recently, Todd Zywicki (George Mason University) and Aaron Kheriaty (University of California-Irvine) filed separate lawsuits against their respective universities for not adequately recognizing naturally acquired COVID-19 immunity from prior infection in their vaccine mandates.
Zywicki had a substantial win last month; Kheriaty has his first court date set for later this month.
One of the philosophical issues raised by their cases and their Twitter feeds (yes, Virginia, there is sometimes sanity on Twitter), might be put as follows:
What are the best ways of understanding the arguments for a given vaccine mandate?
*
Consider four distinct claims:
1. It is morally permissible that you get the vaccine.
2. It is morally required that you get the vaccine.
3. It is morally permissible for me to mandate that you get the vaccine.
4. It is morally required for me to mandate that you get the vaccine.
Three quick observations:
First, these claims are phrased using “you” and “me” rather than A and B, Smith and Jones, state and citizen. If you think this is cheating (perhaps because it places you, dear reader, in the cross-hairs of a mandate from me), I invite you to adjust accordingly in what follows.
Second, these claims have interesting connections: arguably, 2-4 each assume 1, 4 assumes 3, and none of the earlier ones, taken separately or together, entail any of the later ones.
Third, a funny true story illustrates what I mean by a mandate.
During playground recess my first day of first grade, I watched wide-eyed and listened open-mouthed as two sixth grade bullies squared off to shout things I’d never heard before:
Bully 1 (Grog?): “Aw, shut up!”
Bully 2 (Goliath?): “You wanna come over here and make me?!”
So, that afternoon, I cautiously and curiously tried an innocent experiment in imitation:
Mom: “Honey, please bring your jacket over here to hang up.”
Me: “You…wanna…come…over…here…and…make…me?”
The next thing I remember from that day is me laying on my back in my top bunk, staring at my ceiling, thinking “I should never say that again…”
Point: mandates are attempts to “come over here and make me” do something; they may vary by context in countless ways while still being mandates.
So even though “Joe Biden Is Not Our National Dad,” mandates can come from presidents or parents.
And even when a University aspires to be a “family” in some sense, mandates from employers are typically different from mandates from family members.
Or bullies.
*
So, then. Let’s consider two arguments for 3, starting with this one:
5. If something is more likely than not to prevent future harm to you, then it is morally permissible for me to mandate that you do it.
6. You getting the vaccine is more likely than not to prevent future harm to you.
Therefore,
(3) It is morally permissible for me to mandate that you get the vaccine.
5 is a statement of a controversial view called “paternalism.” Some actually seem to endorse it if you listen to them closely when they talk about COVID-19. But most realize 5 is appropriate for parent-child relations but not for adult relations. So a better defense of mandates rejects 5, and thus treats 6 as irrelevant.
A different argument for 3 says this:
7. If something is more likely than not to prevent future harm to others, then it is morally permissible for me to mandate that you do it.
8. You getting the vaccine is more likely than not to prevent future harm to others.
Therefore,
(3) It is morally permissible for me to mandate that you get the vaccine.
8 is a locus of controversy, for the plaintiffs above and others today, because they argue that 8 is often simply false, especially when spoken to the millions who have already acquired and maintain a robust natural immunity to COVID-19 and its variants from prior COVID-19 infection.
Even if 8 were able to be stated more carefully so that it quantified the likelihood of future harm prevention as “between high number H and low number L,” it is not clear that L would be both true and relevant to the argument.
It would of course be cleaner for the argument if our world had just two humans, one C19 vaccine, and one C19 virus. But our world has nearly eight billion humans, far more than the few C19 vaccines with an FDA Emergency Use Authorization (including the one with full FDA approval), and hundreds of C19 variants (e.g. the CDC uses today’s “Delta” as shorthand for what scientists label “B.1.617.2”).
7 is also a locus of controversy, and in ways that I think are worth inspecting. When plaintiffs (Zywicki and Kheriaty) discuss studies showing 8 false for many, they also point out what is true for many:
9. You getting natural immunity from the virus is more likely than not to prevent future harm to others.
And 9 would deliver a very different conclusion than (3) when combined with 7:
7. If something is more likely than not to prevent future harm to others, then it is morally permissible for me to mandate that you do it.
(9) You getting natural immunity from the virus is more likely than not to prevent future harm to others.
Therefore,
(10) It is morally permissible for me to mandate that you get natural immunity from the virus.
Defendants in these cases might argue “we and plaintiffs agree on 7 being true, but we just disagree on 8.” But this would be mistaken.
Plaintiffs deny 7, which explains why they reject an argument from 7 and 9 to 10. Plaintiffs have no interest in mandating that other people acquire natural immunity from getting infected.
This suggests defendants do not really believe 7 either. They are not about to argue from 7 and 9 to 10 for anyone.
Russell DiSilvestro
Philosophy Department
Sacramento State
Interesting breakdown, Russell. A few quick thoughts.
ReplyDeletea) It seems like we could qualify (7) with something like:
(7*) "If something is more likely than not to prevent future harm to others, then it is morally permissible for me to mandate that you do it ***provided that doing it does not incur a substantial risk of substantial harm to you.***"
Such a qualification would allow compulsory vaccinations to SARS-CoV2, but not compulsory exposure to SARS-CoV2, because the prior has an extremely low risk of substantial harm, while the later has a high risk of substantial harm. (This assumes we can come to agreement on what counts as 'substantial' in both cases, but I'll set that aside.)
This wouldn't help the plaintiffs, but it seems like a reasonable qualification in its own right.
b) Assuming your argument here holds, I think it vitiates the Plaintiffs' case. I don't see how anyone can reasonably reject (7), or at least some close variation thereof.
Jessica Flannigan uses the analogy of shooting guns in the air to celebrate. The odds of any single person's bullet hitting another person is very small, but when enough people are firing their guns in the air together, the likelihood that someone will be hit is high. Ergo prohibiting people from firing guns in the air is justified on public health grounds. (And, she argues, the same logic justifies compulsory vaccinations.)
In order to reject (7) (or some close variation thereof) I think the plaintiffs would have to bite the bullet (pun intended) on Flannigan's analogy; they'd have to say that prohibitions against firing guns in the air are unjustified. That seems grossly implausible on it's face.
Garret,
DeleteThanks. A few quick replies, and then a longer one below to you and Chong together.
Re: your suggestion (a), I do think your suggested (7*) is a step is a reasonable qualification in its own right (indeed, it’s a big step in the right direction). But I am more optimistic than you that such qualifications actually strengthen the plantiff’s complaint, at least if their empirical claim stands—namely, that current COVID vaccines, when given to a person with natural (infection-induced) immunity, provide at best no benefit to the person, at worst some risk to the person, and almost certainly no benefit to others, primarily because COVID vaccines provide only “non-sterilizing immunity” (the sort that does not prevent transmission) instead of “sterilizing immunity” (the sort that prevents transmission).
Re: your suggestion (b), I think both my original (7) and your (7*) still permit too many mandates. Consider: they may each permit mandating not firing guns in the air (thanks for reminding us of Flannigan’s great example), but they also permit mandating a reduction of the speed limit from 65 to 25 on all interstate highways. Furthermore, whether the Flannigan example carries over to COVID vaccines depends again on the empirical details of what those vaccines actually do and do not accomplish for given subsets of the population, in this case individuals with natural infection-induced immunity.
Your comments make me realize I should have formulated both (7) and (8) more precisely—both the sake of representing the reasoning of the defendants more charitably, and for the sake of representing the reasoning of the plaintiffs more charitably. I will try to take another stab at that in my longer reply to you and Chong together.
Garret,
DeleteThanks. A few quick replies, and then a longer one below to you and Chong together.
Re: your suggestion (a), I do think your suggested (7*) is a step is a reasonable qualification in its own right (indeed, it’s a big step in the right direction). But I am more optimistic than you that such qualifications actually strengthen the plantiff’s complaint, at least if their empirical claim stands—namely, that current COVID vaccines, when given to a person with natural (infection-induced) immunity, provide at best no benefit to the person, at worst some risk to the person, and almost certainly no benefit to others, primarily because COVID vaccines provide only “non-sterilizing immunity” (the sort that does not prevent transmission) instead of “sterilizing immunity” (the sort that prevents transmission).
Re: your suggestion (b), I think both my original (7) and your (7*) still permit too many mandates. Consider: they may each permit mandating not firing guns in the air (thanks for reminding us of Flannigan’s great example), but they also permit mandating a reduction of the speed limit from 65 to 25 on all interstate highways. Furthermore, whether the Flannigan example carries over to COVID vaccines depends again on the empirical details of what those vaccines actually do and do not accomplish for given subsets of the population, in this case individuals with natural infection-induced immunity.
Your comments make me realize I should have formulated both (7) and (8) more precisely—both the sake of representing the reasoning of the defendants more charitably, and for the sake of representing the reasoning of the plaintiffs more charitably. I will try to take another stab at that in my longer reply to you and Chong together.
Russell, maybe your title should be “arguing against vaccine mandates” by raising problems with arguments for vaccine mandates. I wanted to raise one factor that is relevant to both arguments for vaccine mandates and then say a bit more about each argument.
ReplyDeleteOne missing consideration in your arguments is the gravity of the harm involved. This makes me think of Jurist Learned Hand and his formula for determining when a defendant should have taken precautions to avoid harm. According to the formula, if the burden of taking precautions is less than the probability multiplied by the gravity of the resulting injury (B < P X I) then the defendant must absorb the burden and take the necessary precautions. The formula reminds us to consider not just the likelihood of harm, but also the gravity of harm. If not taking precautions may result in the loss of $1, then no big deal. But, if not taking precautions may result in death or bodily injury, even if the likelihood is small but not insignificant, then we may need to rethink our hesitancy in taking precautions.
During the current pandemic, we’ve learned that the chance of hospitalization is about .021 and the chance of death is about .016. These are not big numbers, but they are also not insignificant. 687,876 people have died of COVID-19 in the United States alone. For those 60 years and older, the hospitalization rate is greater, about 9.2%.
With this in mind, we could revise your arguments. We could revise 5 to read: 5*. If something is more likely than not to prevent death or bodily injury to you, then it is morally permissible for me to mandate that you do it. We could revise (7) to read (based on a duty not to harm others): 7*. If something is more likely than not to prevent death or bodily injury to others, then it is morally required for me to mandate that you do it.
Now, a bit more about each of your arguments. You said, “[b]ut most realize 5 is appropriate for parent-child relations but not for adult relations.” I don’t know about “most.” I realize that some liberal and libertarian thinkers would not accept restrictions on activities that may cause harm to self, but many, if not most, don’t agree with this, particularly when the activity may cause death or great bodily injury. Most would agree that we shouldn’t allow individuals to donate their second kidneys.
Personally, I don’t care as much about people putting themselves in danger, but I do care about certain vulnerable populations, and the elderly would qualify as such during the current pandemic.
In your second argument, 8 is indeed the locus of the controversy. It too can be revised: 8*. You getting the vaccine is more likely than not to prevent death and bodily injury to others. 8 and 8* raise empirical questions that are being debated, though I’m convinced that two doses of a vaccine (and two exposures as opposed to one) would provide more effective and lasting immunity (this probably satisfies the prevailing general acceptance within the relevant scientific community standard).
Whether we can substitute 9 for 8 will depend on the variations in the likelihood of harm and whether the probabilities are sufficiently similar. I am doubtful, but again these are empirical questions. Also, with Garret, I also don’t see how one can reasonably reject 7.
We may continue to have disagreement on the second argument, but I think the factual issues will be resolved in its favor. I think the state or institutions should require vaccines or something similar to prevent harm to others, even if the risk of harm is low but not insignificant because the gravity of harm to a vulnerable population within our society is so great.
Chong,
DeleteThanks. A few quick replies, and then a longer one below to you and Garret together.
Re: your reference to the formula for Jurist Learned Hand, I welcome it. (And it both reminds me of, and somewhat answers, the funny quote from the recent Gerald Dworkin book on philosophical humor, where David Hills says this: “Philosophy—The ungainly attempt to tackle questions that come naturally to children, using methods that come naturally to lawyers.” https://dailynous.com/2021/08/13/bits-laughing-matter-guest-post/ ). It’s a big step in the right direction. I just think we should calculate the Burden using a multiplication of Probability and Injury as well, which I will try to do in my longer reply to you and Garret below.
Re: your revised (5) and (7), I also think those are big steps in the right direction, but I worry (to recycle an example I used in replying to Garret) that they permit mandating the reduction in speed limit from 65 to 25 on all interstates.
Re: the empirical disagreement, it’s worth noting that some vaccine mandates used by universities like Sac State already explicitly recognize natural infection-induced immunity as the equivalent of vaccine-induced immunity. But do see my longer reply to you and Garret below.
Chong,
DeleteThanks. A few quick replies, and then a longer one below to you and Garret together.
Re: your reference to the formula for Jurist Learned Hand, I welcome it. (And it both reminds me of, and somewhat answers, the funny quote from the recent Gerald Dworkin book on philosophical humor, where David Hills says this: “Philosophy—The ungainly attempt to tackle questions that come naturally to children, using methods that come naturally to lawyers.” https://dailynous.com/2021/08/13/bits-laughing-matter-guest-post/ ). It’s a big step in the right direction. I just think we should calculate the Burden using a multiplication of Probability and Injury as well, which I will try to do in my longer reply to you and Garret below.
Re: your revised (5) and (7), I also think those are big steps in the right direction, but I worry (to recycle an example I used in replying to Garret) that they permit mandating the reduction in speed limit from 65 to 25 on all interstates.
Re: the empirical disagreement, it’s worth noting that some vaccine mandates used by universities like Sac State already explicitly recognize natural infection-induced immunity as the equivalent of vaccine-induced immunity. But do see my longer reply to you and Garret below.
Chong and Garret,
ReplyDeleteThanks again for your initial replies. As someone who has been following the back-and-forth regarding vaccine mandates in general, and Kheriaty’s ongoing case in particular (his first day in court is…today, September 27), I want to continue this thread with what I hope are re-castings that build on your insights, while still remaining tethered to what folks are actually arguing about empirically and legally.
So let me re-try re-casting the harm-to-others argument:
(7**) If you doing something prevents some amount “A” more risk of harm to others than it produces for you, then it is morally permissible for me to mandate that you do it.
(8**) You getting the vaccine prevents some amount “A” more risk of harm to others than it produces for you.
Therefore,
(3) It is morally permissible for me to mandate that you get the vaccine.
A few notes on this version:
I am deliberately leaving the precise numerical value of A unstated, but the crucial thing is that it’s the same in (7**) and (8**).
Let me propose how to calculate A.
People who crunch numbers on risk for a living probably have way better formulas for this. But a starting point is to choose whether to use subtraction or division:
Subtraction: A = (risk of harm to others) – (risk of harm to you)
Division: A = (risk of harm to others) / (risk harm to you)
I will use subtraction, though I’m honestly not sure this is the best approach.
So let’s assume A = (risk of harm to others) – (risk of harm to you)
Abbreviated: A = O - Y
How do you calculate O and Y?
Simple. Use a version of Jurist Learned Hand’s formula that Chong pointed out, where the burden of taking precautions is compared to the probability times the gravity of the resulting injuries. Hand asked: B < P x I ? We will ask: Y < O ?
O, the risk of harm to others, can be calculated like this:
The risk of harm to others = (the probability of harm to others) times (the magnitude of harm to others)
Abbreviated: O = Po x Mo
Y, the risk of harm to you, can be calculated like this:
The risk of harm to you = (the probability of harm to you) times (the magnitude of harm to you)
Abbreviated: Y = Py x My
So, then. If this is correct, much now turns on the probability and magnitude of harm to you, and the probability and magnitude of harm to others.
*
Without meaning to comprehensive here, let me just close my long batch of replies this round by quoting from age 9 and 10 (or 6 and 7, depending on how you count) of document Document 31 from Kheriaty’s legal paperwork in his trial with the UC system today (September 27), a copy of which can be found online at this url:
https://07b259f3-7a04-4bc7-8bec-e3a2469917d2.filesusr.com/ugd/b26c37_a440123dde114adbbd8bac5afbba3550.pdf
“Despite submitting 72 pages of written expert testimony and citing to 84 sources, Defendants do not present evidence to contest the following facts Plaintiff established in his moving papers:
1. That the CDC admits that vaccinated individuals can become infected with and spread the SARS-CoV-2 virus (“non-sterilizing immunity”), but naturally immune individuals’ do not spread this virus (“sterilizing immunity”).
2. That when symptomatic cases occur, the rate among vaccinated individuals (“breakthrough cases”) is multiple fold higher than the rate among naturally immune individuals (“reinfections”)
3. That there has never been a single documented case of a reinfection resulting in further transmission of the virus, but there have been many documented cases of breakthrough infections resulting in subsequent transmission.
These uncontested facts alone reflect that treating vaccinated and naturally immune individuals differently is irrational and does not serve the goal of protecting the UC community. The very fact that naturally immune individuals like Plaintiff have sterilizing immunity, while vaccinated individuals can still silently transmit SARS-CoV-2 should end the inquiry.”
Chong and Garret,
ReplyDeleteThanks again for your initial replies. As someone who has been following the back-and-forth regarding vaccine mandates in general, and Kheriaty’s ongoing case in particular (his first day in court is…today, September 27), I want to continue this thread with what I hope are re-castings that build on your insights, while still remaining tethered to what folks are actually arguing about empirically and legally.
So let me re-try re-casting the harm-to-others argument:
(7**) If you doing something prevents some amount “A” more risk of harm to others than it produces for you, then it is morally permissible for me to mandate that you do it.
(8**) You getting the vaccine prevents some amount “A” more risk of harm to others than it produces for you.
Therefore,
(3) It is morally permissible for me to mandate that you get the vaccine.
A few notes on this version:
I am deliberately leaving the precise numerical value of A unstated, but the crucial thing is that it’s the same in (7**) and (8**).
Let me propose how to calculate A.
People who crunch numbers on risk for a living probably have way better formulas for this. But a starting point is to choose whether to use subtraction or division:
Subtraction: A = (risk of harm to others) – (risk of harm to you)
Division: A = (risk of harm to others) / (risk harm to you)
I will use subtraction, though I’m honestly not sure this is the best approach.
So let’s assume A = (risk of harm to others) – (risk of harm to you)
Abbreviated: A = O - Y
How do you calculate O and Y?
Simple. Use a version of Jurist Learned Hand’s formula that Chong pointed out, where the burden of taking precautions is compared to the probability times the gravity of the resulting injuries. Hand asked: B < P x I ? We will ask: Y < O ?
O, the risk of harm to others, can be calculated like this:
The risk of harm to others = (the probability of harm to others) times (the magnitude of harm to others)
Abbreviated: O = Po x Mo
Y, the risk of harm to you, can be calculated like this:
The risk of harm to you = (the probability of harm to you) times (the magnitude of harm to you)
Abbreviated: Y = Py x My
So, then. If this is correct, much now turns on the probability and magnitude of harm to you, and the probability and magnitude of harm to others.
*
Without meaning to be comprehensive here, let me just close my long batch of replies this round by quoting from page 9 and 10 (or 6 and 7, depending on how you count) of document Document 31 from Kheriaty’s legal paperwork in his trial with the UC system today (September 27), a copy of which can be found online at this url:
https://07b259f3-7a04-4bc7-8bec-e3a2469917d2.filesusr.com/ugd/b26c37_a440123dde114adbbd8bac5afbba3550.pdf
“Despite submitting 72 pages of written expert testimony and citing to 84 sources, Defendants do not present evidence to contest the following facts Plaintiff established in his moving papers:
1. That the CDC admits that vaccinated individuals can become infected with and spread the SARS-CoV-2 virus (“non-sterilizing immunity”), but naturally immune individuals’ do not spread this virus (“sterilizing immunity”).
2. That when symptomatic cases occur, the rate among vaccinated individuals (“breakthrough cases”) is multiple fold higher than the rate among naturally immune individuals (“reinfections”)
3. That there has never been a single documented case of a reinfection resulting in further transmission of the virus, but there have been many documented cases of breakthrough infections resulting in subsequent transmission.
These uncontested facts alone reflect that treating vaccinated and naturally immune individuals differently is irrational and does not serve the goal of protecting the UC community. The very fact that naturally immune individuals like Plaintiff have sterilizing immunity, while vaccinated individuals can still silently transmit SARS-CoV-2 should end the inquiry.”