BU Philosophers Object To University’s Fall 2020 Plans


Two philosophy professors at Boston University (BU), Russell Powell and Daniel Star, have authored a statement objecting to their administration’s plans for how teaching during the Fall 2020 term will proceed, in light of the ongoing COVID-19 pandemic.

While BU administration’s plans for the fall currently provide students with the option to take their courses in-person or online, they do not afford faculty a similar option. Faculty will be required to teach their courses in person. The only way not to would be to take paid sick leave in lieu of teaching, or take unpaid leave.

Professors Powell and Star explain their objections in their “In-Person Teaching Must be Optional for BU Faculty: An Open Letter to the University Community,” the text of which is reprinted below. Faculty at other schools countenancing similar policies may find it useful.

BU’s plan for the fall of 2020 remains very much in flux. At the moment, however, the university appears wedded to the idea that not only should all students who are well and able take up residence on campus, but also that all or most courses be taught in person using a “Learn from Anywhere” approach that would allow students to attend classes either in person or remotely. Students would be permitted to make this decision based on their personal medical condition, family circumstances, travel burdens, and willingness to assume the health risks and potential anxieties that in-person attendance entails. We will not opine here on the practical or moral wisdom of bringing students to campus in the fall, a decision that many peer universities (for good or for ill) have also made. What is unusual about BU’s approach as compared to peer universities, and in our view raises serious moral misgivings, is that the university’s policy as it stands does not carve out a similar sphere of liberty for BU faculty. 
 
The Fall Plan
 
As of early June, BU faculty are being given no choice but to teach in person in the fall, even though this is a deeply personal decision—one that is no less than a matter of life and death—for faculty members and their families. Like the students they teach and mentor, faculty members find themselves in diverse medical, familial, and geographic circumstances and have very different tolerances for risk. A blanket requirement that faculty teach in person without regard to their medical and family situations would be an unconscionable breach of the university’s moral fiduciary duties to faculty members, one that places the overwhelming weight of the health burden of this once-in-a-century pandemic on important and vulnerable stakeholders.
 
The notion that faculty members could choose to take a paid sick leave in lieu of teaching, as suggested to us by some in the university administration, is woefully inadequate for any number of reasons, not the least of which is that a paid sick leave would not (as presently configured) cover situations wherein one’s family members are in high-risk categories, such as if one’s partner is pregnant or if one cares for elderly parents. Furthermore, many faculty members are parents of young children and cannot risk being quarantined from their family for 14 days at a time with each exposure (or worse, if they become infected with COVID-19), leaving them unable to properly care for their children. As it stands, the only recourse for a faculty member who finds themselves in such circumstances would be to take an unpaid leave, which for obvious reasons is not a satisfactory option. There is no doubt that a policy that allows professors, lecturers, and graduate students that fall into certain risk classes to teach online would be considerably better than a blanket policy that admits of no exceptions. If the university chooses to go in this direction, we would strongly urge them to also include faculty who are responsible for the care of young children or elderly family members. 
 
However, we strongly favor a policy that would give every faculty member the option of teaching their classes wholly online. This would allow all faculty members to exercise their autonomy over a fundamental life decision in light of their own personal circumstances and in consultation with their own values and priorities—just as the university has done for students. There are several reasons why this is the morally best policy. 
 
First, and most obviously, it is morally wrong to demand that professors risk their health and that of their family members, given the online teaching alternative. In the absence of a vaccine, the only way to open universities in an even remotely safe manner is to have robust testing, contact-tracing teams, supportive quarantine for those exposed, and full PPE gear for faculty in place by the time the fall semester begins. Given the dire national shortages of these things, to say this is a tall order would be an understatement. Creating a reasonably safe environment must go well beyond “security theatre” (such as temperature taking, hallway segregation, classroom spacing, etc.), which risks creating a false sense of security. There will also be major hurdles to adequately enforcing the required conditions, as some students will (e.g.) refuse to wear masks for any number of reasons, including political ones (and we have heard from the university administration that the campus police will not be in the enforcement business). Even if these conditions could be miraculously met and all protocols abided to, many students, faculty, staff, and people in the surrounding Boston community will inevitably get infected, some will be irreparably harmed, and some will die. It is far from clear that this would be a morally acceptable outcome even if the only alternative were to shut down the university for the fall; but it is patently immoral given the remote teaching alternative. 
 
We fear that despite assurances from the university and the good-faith efforts presently underway to physically reshape parts of the campus, the population density on campus will be too high at certain times to allow for adequate social distancing measures if all courses are taught in person. A mixed approach that allows some courses to be offered online only would considerably lower the population density on campus at critical times.
 
Ethical Tradeoffs
 
In defending the urgent need to bring students back to campus, universities have stressed the value of in-person teaching and the limits of teaching remotely, despite the advent of what only a decade ago would have seemed like miraculous communication technology. It is far from obvious that lecturing while dressed in full PPE gear, including masks, goggles, visors, gloves, and gowns, would be in any way optimal for anyone—as opposed, say, to carrying on these same activities from the safety and psychological comfort of one’s own home. Moreover, it is likely that many classes, if not the entire university, will be forced to switch to fully remote learning mid-semester as outbreaks flare up and students and faculty get exposed to the virus. In any case, it seems rather obvious to us that optimal pedagogy cannot conceivably justify significant risks to the health of faculty members, staff and their families, to say nothing of the wider Boston community. 
 
The only conceivable justification for in-person teaching under pandemic circumstances is that without it many students will choose not to enroll for the fall—and as a result, the economic impact on the university will be so devastating that many faculty members and staff will have to be furloughed or laid off. We are skeptical that this is the forced choice universities are confronted with, especially for institutions with vast real estate holdings and large endowments invested in a stock market that is booming irrespective of national unemployment rates. However, it is incumbent upon BU and any other universities to make this moral case explicitly and transparently, so that its stakeholders can meaningfully evaluate and contribute their voices to decisions that may have a profound effect on their health, their lives, and their livelihoods. 
 
Sending a Moral Message
 
What sort of a message are we sending to students if we encourage them to return to campus because their own health is not dramatically at risk, when they are likely to asymptomatically spread the virus to older and more vulnerable university populations? We are telling them that they should not care (or should not care very much) about taking risks that might seriously harm or kill other people. We are telling them that BU faculty do not deserve the same rights as students. We are saying that faculty are here to serve students at any cost—to provide supposedly optimal teaching environments at the expense of their own lives and the lives of those they love, rather than to work collaboratively with students in their development as responsible citizens of our community and stewards of our planet.
 
Crucially, universities should not take student preferences for how campus life should be conducted in the fall as a fixed point. Instead, they should make the moral case to students that we are all in this together and that we have an obligation to keep one another safe and to support the institution and community we have chosen to be part of. One of us (Powell) co-wrote an opinion piece on the science and ethics of reopening universities that was published in Inside Higher Ed. Since then, the article has been assigned in summer ethics classes, and it turns out that students embrace and appreciate the strength of the arguments therein. Many prospective and current students, who are not themselves at great medical risk, have yet to think through the moral ramifications of attending class in person (an obvious but important one of which is that teachers in the classroom will need to wear masks and screens, whereas online teaching requires no such impediments to teaching well). But they are receptive to reasons. The university must make the case that not only should these students join or continue with BU in the fall, but also that they should do so in ways that do not put others in grave danger. 
 
Are we, or these other ethicists, saying that it’s never permissible for the university to engineer an environment in which its members risk being harmed or harming others? No. We are simply asking that the seriousness of the policy being proposed be acknowledged and the costs and benefits transparently discussed. At the very least, professors and students should, where possible, have the option of teaching and learning online. If, knowing there are serious risks, professors choose to teach in the classroom, and students choose to return to the classroom, that is a decision they should be allowed to make in a way that respects their autonomy, weighing these considerations for themselves.
 
Faculty, not physical spaces, are the life blood of the university. Compelling faculty members, at the pain of their jobs, to risk their lives for putatively preferable pedagogy or unclear economic benefits is to cut off the university’s nose to spite its face.

Subscribe
Notify of
guest

100 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
John
John
3 years ago

The authors certainly raise important points–not least that the requirement that faculty teach in person will be moot if there is a second outbreak come fall and everything shuts down regardless. If we all reasonably expect that teaching will be made remote at some point in the fall , then why not give faculty the choice of how and where they teach in the first place? This strikes me as a pretty strong objection to the university’s policy.

But I also wonder whether parts of this letter overreach.

For example, I’m not sure what the authors mean exactly by the metaphor ‘life blood of the university’, but there is at least one perspective from which it is simply false that faculty are the lifeblood. Universities, like any institution or business, have a revenue side and an expense side. Students are largely on the revenue side; faculty are largely on the expense side. And once there is a significant decline in revenue–either because of decreased enrollment, decreased tuition, or what have you–there will be knock-on effects that will pretty severely impair the university. But the converse isn’t true. As universities have mercilessly demonstrated over the last ~25 years, you can substantially cut the expense side–by hiring contract faculty, suppressing unionization efforts, denying basic benefits (like paid sick leave) etc.–without severely impairing day-to-day operations. Students, many of whom are unaware of their adjunct professors’ labour conditions, are often pretty satisfied with the learning experience these precariously employed academics provide–and sometimes students are even more satisfied than they are when taught by tenured faculty who often, but not always, undervalue teaching in comparison to their research. (And students’ equal or greater satisfaction should be no surprise, of course, because the highly competitive job market in academia entails that, by and large, the deciding difference between faculty and adjunct is good fortune rather than ability.) Either way, the pervasiveness of adjuncting etc. shows that faculty are more elastic than students, and this suggests they are not the ‘lifeblood’ of the university, at least not on one straightforward interpretation of this metaphor.

For another, I am not sure I follow the argument against taking paid sick leave. So I might be throwing a ball rather than a strike here, but: paid sick leave is already a benefit no longer available to many workers. For those fortunate workers who enjoy this benefit, is it reasonable to attach to it the expectation that the employer also consider the potentially negative effects of the worker staying at home? If I have paid sick leave benefits and come down with a nasty cold, is it reasonable to expect my employer to offer additional accommodations because staying home might imperil those I live with? Run this argument by your ordinary no-benefits labourer (or no-benefits adjunct), and I think most would say ‘no’ because they would be grateful for having the very option of receiving pay without working in the first place. So, I am wondering whether this line of argument stems from getting a bit too comfortable with the privileges of tenured academic labour.

That all being said: if the authors are right that BU has not transparently engaged with faculty on these issues, then that’s plainly unacceptable and should be remedied immediately. Everyone deserves the respect of being communicated to directly about matters that concern their welfare, there is no doubt.

krell_154
3 years ago

The position taken in this letter, and the presented arguments, are indeed so clear and convincing, that it is hard for me to think of any good objections to it.

Greg
Greg
Reply to  krell_154
3 years ago

Then I will try…
If you were a “philosopher in industry” it is very possible that your business will tell you to report into the office on Monday. They will put safeguards in place to avoid litigation but once done, that is the expectation. If you can’t meet the requirements of the job, the door is on the left.

More enlightened organizations will realize that you can do your job at home, giving you more flexibility. What if the organization thinks you need to be in the office? Then that is a requirement of the job and if you can’t do the job, the door is on the left.

If a University is selling a class room experience and students are willing to show up and pay, they need teachers who will show up. If not, the door is on the left.

This “harshness” may be right or wrong but it is the reality beyond the ivory towers.

Irina M
Irina M
Reply to  Greg
3 years ago

This seems like a good place to leave this article: https://academicmatters.ca/neoliberal-response-to-covid-19/

John
John
Reply to  Irina M
3 years ago

Thanks for posting this. It’s a good article. But until tenured faculty speak up loudly and clearly about the sustained ‘neoliberal’ assault on the university in the form adjunct labour, I think that posting this here just adds to the tone deafness of the authors’ letter.

Skeptical skeptic
Skeptical skeptic
3 years ago

A question: why is it a “deeply personal choice” for faculty to teach in person, but not for medical workers, bus drivers, subway operators, elementary school teachers, grocery store workers, retail workers, and other groups? If we expect those groups to perform their jobs in person, why shouldn’t faculty do so as well?

Meanwhile, philosophers should look at actual risks rather than media hype. So far CDC data lists 17,000 deaths in the US under the age of 65 from COVID and 126 deaths under the age of 24. An undergraduate (typically under 24) is twice as likely to be struck by lightning as to die from COVID. A faculty (typically under 65) is twice as likely to die in a car crash as from CVOID. But we don’t do remote teaching to avoid lightning strikes and car crashes.
https://data.cdc.gov/NCHS/Provisional-COVID-19-Death-Counts-by-Sex-Age-and-S/9bhg-hcku/data

face palm
face palm
Reply to  Skeptical skeptic
3 years ago

Did you make sure to factor in the probability that a student or faculty would get the virus, not die or even get sick, but give it to others, including those who are more vulnerable?

John
John
Reply to  Skeptical skeptic
3 years ago

I think you’re on to something here, but there is at least one important difference that I think the authors do make clear in their letter: that technology today makes remote teaching possible. In each of the positions you mentioned a remote option is not possible. So that’s a big ethical difference maker. However, the letter could have been more frank about the limitations of remote teaching, and their illustration of donning medical garb in the classroom is a bit hyperbolic. Nonetheless, their argument here, as I understand it, is that the security gained by staying at home outweighs the drawbacks of teaching remotely. I’d be stretching if I said that didn’t strike me as plainly true, even though both these authors appear on the younger side, and therefore are likely not seriously at risk.

Nonetheless, you’re right insofar the letter does read very much like it’s written by individuals who have gotten quite used to enjoying privileged labour conditions. Tell your average employer about your ‘deeply personal choice’ to not claim sick pay in favour of dictating your optimal working conditions and you will be told exactly where you can stick that choice. Then they’ll tell you where you can stick your philosophy degree. This doesn’t mean the authors are wrong; but the letter is a bit embarrassingly out of touch, especially coming from workers who already have some of the most secure and comfortable jobs in the modern economy.

One hopes they’re as open and direct about the many other labour injustices in the university that they benefit from. But if they’re like most people who enjoy privileges, that might be a boat that remains conveniently unrocked.

Elizabeth
Elizabeth
Reply to  Skeptical skeptic
3 years ago

Is this comment actually saying “statistically you probably won’t get sick, but if you do, statistically you probably won’t actually die” or am I missing some more compelling part of the argument? Also, saying that it is a personal choice for faculty to teach in person at a time when they risk doing harm to others (including their students, their community, or their household members) as a potential virus vector, and being harmed themselves, does not imply that other workers are also significantly more risk and that they often have less choice (health care workers, delivery workers, grocery workers, mass transit workers, etc.). BUT the way to limit risk to those most vulnerable workers is for the rest of us who are able to limit our contact with the virus, and with them (so, not having to commute on mass transit to work, not going to the grocery store after teaching a classroom full of students who may or may not have spread the disease) is how we look out for those who are most vulnerable. Finally, this also really depends on your other identity markers, and it is incorrect to presume that the COVID statistics apply to all persons and groups equally, given the disparities in how the virus has spread and who is dying from it, that shows disproportionate impact on communities of color.

(As an aside: That we are all more likely to die from car crashes than many other things is good reason, in my mind, for us to be on the roads less, not to do other risky things more.)

skeptical skeptic
skeptical skeptic
Reply to  Elizabeth
3 years ago

Responding to the above:

– to facepalm: obviously that’s a risk but are you imagining that if classes are online then 18-24 year olds and faculty will lock themselves in their rooms all day? Lockdowns are ending. People are interacting. The question is how much _additional_ risk is posed by in person classes.

– to John: Good point, it’s true that it’s not possible for some of the professions I mentioned to go online. But it would be possible for grocery stores and retail stores to close to customers and go to delivery only. That would have costs. The costs were judged not be worth it. The same reasoning can apply to colleges. Going online has costs to the students–they lose an important experience, they lose interactions with peers, they get inferior classes. I don’t think the cost is worth it.

– to Elizabeth: I wasn’t really giving an argument just asking a question. I think the authors are exaggerating the risks of in-person teaching (as well as making ridiculous claims about what it would involve–no one is going to show up in goggles and medical gowns). I tried to make that point by pointing out that most of us have no problem doing things that have comparable risks. If I wanted to give an argument, it would go like this: (1) it is unreasonable to insist on online classes in order to avoid risks that are comparable to those incurred by driving, etc.; (2) for people who are under 65, the risks incurred by online teaching are in fact comparable to those incurred by driving, etc.. My claim (2) might be an exaggeration–it depends on how much _extra_ risk you think faculty are incurring by teaching in person rather than online. I’m 48, reasonably healthy (though I have asthma), and not worried. If I end up teaching online, I’ll still go outside, go to grocery stores, eat at restaurants, drop my son off at school, and so on. I’ll incur some risk of covid from those activities. I guess I’d incur a bit more risk by also teaching in person. But my guess is that the additional risk is comparable to the risk I take by driving, etc.. Obviously this would be different if you’re planning to sequester yourself inside for the next 6-18 months and would only venture outside so as to teach in person. But that’s not the situation most of us face.

David Wallace
Reply to  Skeptical skeptic
3 years ago

“So far CDC data lists 17,000 deaths in the US under the age of 65 from COVID and 126 deaths under the age of 24. An undergraduate (typically under 24) is twice as likely to be struck by lightning as to die from COVID.”

I can’t make that statistic work. There are about 50 deaths per year in the US from lightning strikes, i.e. about 20 in the five-month period for which these COVID-19 statistics were collected. That suggests about 7 deaths from lightning strikes in the under-24 age range, as compared to 126 deaths from COVID over that period – so 18x as many COVID-19 deaths. (And of course this is only the deaths that occurred with the lockdown, it’s much lower than the herd-immunity numbers would be.)

Or are we comparing lightning *strikes* to COVID-19 *deaths*? I can’t readily find data there, but I’ve seen estimates that 10%-30% of lightning strikes are fatal. Taking the mid-point, that would suggest 35 lightning strikes on under-24s, as against 126 COVID-19 deaths.

What am I missing? Where do you get that 2x statistic from?

skeptical skeptic
skeptical skeptic
Reply to  David Wallace
3 years ago

David– you’re right. I wrote that quickly and wasn’t thinking carefully about the examples. I was comparing total lightning strikes across the population, irrespective of age, to total deaths in the under 24 range. I had seen figures stating 800-1000 lightning strikes (not fatalities) in the US per year, or about 250 per quarter; I was comparing that number to the 124 deaths in same period.

So my particular example is flawed, but I take it that the general point is clear–we’re talking about very, very small numbers.

more skeptical?
more skeptical?
3 years ago

Building on skeptical skeptic’s point above: if you look at the actual risks to college-age and middle-aged people, this is once again a case in which they’re being asked to make enormous sacrifices for the benefit of septuagenarians who have plundered their wealth, desecrated their environment, and voted in racists and fascists. No thanks.

Caligula's Goat
Caligula's Goat
Reply to  more skeptical?
3 years ago

Some of those septuagenarians are their family, friends, and loved ones. Some of those at risk are invisibly immunocompromised and include faculty, students, administrative, and custodial staff. Unless you’re seriously arguing that in-person teaching produces pedagogical benefits of such value that these risks are worth imposing *on all of us* then what are you commenting for? Seriously. Just for the snark? Are you that disaffected?

Kenny Easwaran
Reply to  more skeptical?
3 years ago

When you say “the actual risks to college-age and middle-aged people”, are you talking about the small risk of dying, or are you including the risk of spending a week of one’s life feeling too weak to walk across the room? Because I know several people that have fallen ill from covid and had that result. Even if you ignore the increased risk of infection to third parties, many people might feel that a few percent chance of this sort of illness is too great a risk to pay just to gain the benefit of one term of socially-distanced in-person classwork over online classwork.

Evan Thompson
3 years ago

Published two days ago in the Proceedings of the National Academy of Sciences USA: https://www.pnas.org/content/117/22/11875
“These observations confirm that there is a substantial probability that normal speaking causes airborne virus transmission in confined environments.”
“Our laser light scattering method not only provides real-time visual evidence for speech droplet emission, but also assesses their airborne lifetime. This direct visualization demonstrates how normal speech generates airborne droplets that can remain suspended for tens of minutes or longer and are eminently capable of transmitting disease in confined spaces.”
In light of these and other data, in-person classroom teaching seems like a very bad idea until there is a vaccine and herd immunity.

Penelope Bitzas
Penelope Bitzas
Reply to  Evan Thompson
3 years ago

In addition.. in the area of singing it is considered highly transmissable. Must be in a large space with excellent ventilation. Virologists who have studied this tecommend no in person teaching if one on one until there is a vaccine

LS
LS
3 years ago

The car fatality comparison is not a good one. The CFR for COVID (not the number of deaths per age group, which is irrelevant as it follows from the number of people exposed, which has been impacted by lockdowns) is apparently between 0.5 and 1.0%. (As I’m sure many of you are aware, .25% of New York City has died from this disease, so the CFR can’t go below that number. But even in New York significantly less than half of residents have been exposed–estimates are around one fifth, I believe).

Let’s take the lower number, which is probably an underestimate. If everyone in the US. were exposed, under this lower estimate, that would mean 1.6 million deaths. Of course we would reach herd immunity before that (at about 60%?), so that’s a ceiling. Now let’s bring in cars: the number of car fatalities per year is 37,000. Let’s focus on one decade, an important one for the risk to college professors: the 50s. About 6,500 car fatalities occurred in that age group in 2018. There are about 42 million people in their fifties in the U.S., give and take. Assuming a 1% CFR for that age group (which is in keeping with what I’ve seen), and herd immunity at 60%, the death rate for those in their fifties would be 250,000. Far above 6,500. The hospitalization rate in the 50s is actually 10%–so 2.6 million of those would be hospitalized. Being hospitalized for a significant illness is a serious, life-changing event.

I easily could have made an error here, but let’s be serious: for those in their 50s–not generally considered the elderly–this is a real risk, the biggest they face in their lifetimes. As I’ve pointed out elsewhere, the risk of dying from wingsuit flying (once, of course) has been calculated at 1 in 500 to 1 in a 1000. The risk of getting this disease in your fifties is perhaps ten times higher than the risk of wingsuit flying. For those of us with pre-existing conditions (the sort of which you used to hear people say, “let’s shelter them so we can open up the economy! Funny when it actually comes to giving them accomodations the tune changes and becomes “get back to work”) the risk is significantly higher.

skeptical skeptic
skeptical skeptic
Reply to  LS
3 years ago

The more recent estimates of the case fatality rate put it around .26%, but let’s say it’s 1%. That rate varies by age–we know that most of the deaths are in the 70+ group. (Also, that’s why you can’t just say that .25% of NYC has died and infer than the fatality rate is at least .25%. It depends how the disease has spread. Consider a thought experiment: suppose hypothetical disease COVID-20 kills 50% of people over 75 and 0% of people under 75. Suppose it starts off in age 75+ nursing homes, achieving 100% infection within a month. If you measure the fatality rate at month 1, it will look like it has a 50% fatality rate. But then suppose in the next month it infects everyone outside of nursing homes. If you measure the fatality rate then, it will look much lower; if 10% of people are over 75, it’s fatality rate will be 5% rather than 50%. This is hypothetical and exaggerated, but there’s some evidence that NYC looks something like this: the majority of the deaths have been in nursing homes, so it’s possible that most of the deaths that are going to occur in those homes have already occurred; so, as the infection spreads in NYC, the total deaths might very well not go up very much from where they are now.)

The CDC best estimate of case fatalities in symptomatic individuals aged 50-64 is 0.2%. They also estimate that 35% of infected people are symptomatic, so the infection fatality rate would be about .07% for that age range (https://www.cdc.gov/coronavirus/2019-ncov/hcp/planning-scenarios.html). So your estimate of 1% fatality rate is very high for that age range, by a factor of about 15x. Rather than 250,000 fatalities in the group you mention, according to CDC data you could expect 16,600.

16,600 is still a lot when compared to the 6500 who will die in car crashes. I don’t think having in person classes is worth an extra 16,600 deaths. However: that’s not the relevant comparison. The relevant question is how many _extra_ infections are going to be caused by colleges teaching in person, as compared to a situation in which colleges are taught on line? Not 16,600.

Alastair Norcross
Reply to  skeptical skeptic
3 years ago

“Also, that’s why you can’t just say that .25% of NYC has died and infer than the fatality rate is at least .25%. It depends how the disease has spread.” No it doesn’t. It absolutely doesn’t. If 0.25% of the entire population of NYC has already died from Covid-19, the fatality rate cannot be lower than 0.25%. Patterns of spread are irrelevant to this point. If the entire population of NYC has been infected, and no further Covid deaths occur, the fatality rate is 0.25%. Given that both of those antecedents are almost certainly false, the fatality rate is higher then 0.25%. This isn’t rocket science, or even brain surgery. It’s simple maths.

Skeptical skeptic
Skeptical skeptic
Reply to  Alastair Norcross
3 years ago

That’s actually false, Alastair. It’s a simple point: infection fatality rates are defined by the number of deaths divided by the number of infected individuals. It’s true that the NYC-specific IFR can’t be lower than .25%. But no one is debating about IFR in NYC in particular. We’re talking about IFR in the US as a whole.

Obviously if you focus on region specific IFRs they will vary. If Oldsville has all 75+ citizens, then the Oldsville IFR might be 15%.

Alastair Norcross
Reply to  Skeptical skeptic
3 years ago

So you are shifting the goalposts. But, even so, you need to provide some reason why the population of NYC is radically different from that of the US, if you’re now talking about the IFR for the US. Your example of Oldsville is silly, of course, because there’s no reason to think the population of NYC trends significantly older than the US as a whole. Given that we now know that the IFR in NYC can’t be below 0.25%, and is almost certainly much higher, do you honestly believe that it might be lower than 0.25% for the US as a whole? Of course not.

skeptical skeptic
skeptical skeptic
Reply to  Alastair Norcross
3 years ago

I’m not shifting goalposts–the whole discussion was about US fatality rates.

But on your broader point: I wasn’t trying to explain how the CDC came up with the .26% IFR. I was just pointing out that it’s entirely possible that all three of these things are true: the IFR in the US is 0.26%; only 20% of NYC has been infected; 0.25% of NYC has died. Those claims can all be true.

You and Prof L ask how the CDC can believe the IFR is 0.26% when 0.25% of NYC has died. You’d have to look at the studies if you want a real answer, but here are some possible explanations:

(1) we’re assuming that only 20% of NYC has had covid. That’s probably a bad assumption; the number could be much higher. The antibody tests tell you whether you had the virus 4-6 weeks ago. A test that showed 21% of NYC seropositive in late April (that’s when the studies I’ve seen were done) would mean that 20% had recovered from covid by late March/early April. How many more people got infected since late March/early April? Another 20%? Another 60%? We have no idea. It’s entirely possible that most of the people who are going to die of COVID in NYC have already died.

(2) Most of the NYC deaths are in nursing homes. NYC’s nursing homes are filthy and are full of neglected residents (google it; lots of articles on this in the past few weeks), so the fatality rate has been pushed up in NYC relative to places that don’t have filthy, vermin infested nursing homes full of neglected residents wallowing in unchanged diapers (in other words, it’s not just the age of the population, as L suggests, it’s also the conditions in which that population lives); etc.

I’m sure there are other possible explanations as well.

Kenny Easwaran
Reply to  skeptical skeptic
3 years ago

In case you want actual data, here is the age distribution of New York City:
http://www.censusscope.org/us/m5600/chart_age_graph_1.gif

Here is the age distribution of the United States:
http://www.censusscope.org/us/chart_age_graph_1.gif

As far as I can tell, the biggest difference is that New York has more people in the 20-40 brackets, the next biggest difference is that New York has fewer people in the 5-20 brackets, and the third biggest difference is that New York has fewer people in the over 70 brackets.

Sure, it’s hypothetically true that there could be some city with a radically unusual age pyramid (maybe a retirement community?) where an 0.25% total population fatality rate would be compatible with a national infection fatality rate below 0.5%. But New York isn’t it, and there’s no point in thought-experimenting about a city when we have actual data about it.

Prof L
Prof L
Reply to  Skeptical skeptic
3 years ago

The average age in NYC is two years lower than the average age nationwide—Clearly there are many unknowns here, but it’s not a stretch to say, based on current evidence, that .25% is a lower bound of the IFR generally. If we were getting our data from Sun City, AZ, with an average age of 73, sure, we only identify the lower bound of the IFR *in Sun City* from the actual death rate, and not extrapolate to the nation as a whole. But since the population of NYC is relevantly similar to the population as a whole, there does not seem to be any reason to expect the IFR in NYC to differ from the IFR of the US as a whole.

David Wallace
Reply to  skeptical skeptic
3 years ago

These figures look on the optimistic end – largely for the reasons given – but even in their own terms I’m not sure they support the point being made. The all-cause death rate in the US in the 55-65 age range is about 900 per 100,000 per year. If the COVID-19 rate were indeed 0.07%, i.e., 700 per 100,000, and the pandemic were to take about a year to burn through to herd immunity, you’d roughly be doubling your chance of death by being exposed to it. That’s before considering the risk of hospitalization and permanent damage.

skeptical skeptic
skeptical skeptic
Reply to  David Wallace
3 years ago

David–even if we assume that being exposed to COVID doubles your risk of dying, we’d also need to know how much additional risk of exposure results from teaching in person versus teaching online. It’s not a case of going from 100% chance to 0% chance, obviously. So teaching in person does not double your risk of dying.

How much does it increase your risk of exposure? I have no idea, and neither does anyone else. Let’s say I have a 50% chance of getting COVID if I teach in person and a 25% chance otherwise. (I would be very surprised if the difference was actually this stark.). Then if I’m in the 55-65 group, my chances of dying from COVID would be 0.035% if I teach in person and 0.0175% if I teach on line.

LS claims that there are roughly 42 million people in a comparable age group (50-60) and 6500 car crash fatalities per year in that group. If that’s right, there’s a .015% chance of dying from a car crash. So, if we use my made up numbers about risk increase above, the additional risk you take by teaching in person is basically the same as the additional risk you take by operating a car. And yet we’d think it was crazy to mandate online education in order to avoid car crashes.

David Wallace
David Wallace
Reply to  skeptical skeptic
3 years ago

That’s fair, but it seems to leave us saying “on some estimates, the death risk might be as low as the risk of driving, and the hospitalization risk might not be serious; on others, the death risk is quite a lot higher and the hospitalization risk is severe.”

That would be enough to support the view that maybe it will turn out that staying at home and teaching online was an overreaction even for Age-55-65 faculty. But you seemed to be arguing for the stronger claim that we can be confident right now that it’s an overreaction.

skeptical skeptic
skeptical skeptic
Reply to  David Wallace
3 years ago

In general I agree, but I think the point is a bit stronger. I’d say: _according to what the CDC claims is the best estimate of the COVID IFR_, the additional death risk that a 50-60 old person accrues by teaching in person is probably comparable to the death risk of driving a car; according to other models, the additional death risk might be higher.

But my general point was just that the original letter from the BU professors makes it sounds as if there’s some enormous risk to teaching in person as compared to teaching on line; and I don’t think we have good reason to believe that’s true. We might want to be extremely cautious and accept the costs of online education to avoid the potential risk, but it seems to me to be a very large exaggeration to claim that universities are making some terrible moral error by deciding not to do so.

Laura
Laura
Reply to  skeptical skeptic
3 years ago

I have no need to do anything that puts me in contact with other people right now. None. If my university were forcing me to teach in the classroom this summer, any risk presented by that environment would be my only source of infection risk. If I could do my job in some other way, why wouldn’t I, particularly given the vulnerability of the other people I need to protect?

The fact that people with other jobs are not permitted to work without coming in contact with others is irrelevant. The more of us who are able to avoid contact, the better it is for all those who cannot. If some people have more difficult working conditions, that doesn’t mean people who can obtain better conditions should give them up for some false sense of parity. Rather, we try to make things better for as many as we can, and that means staying home when we can make it happen.

Discussing the risks of this illness as if death is the only relevant one is also shortsighted. We don’t know enough about long-term risks to those who recover; we don’t know enough about newly detected long-term symptoms among children who test positive. What we do know suggests significant risk even from getting sick and then recovering.

Luckily, this situation won’t last forever. The work on vaccines and treatments is proceeding vigorously. Rather than simply throwing in the towel, while pretending it’s not that risky and thus not worth the effort, why not turn our creative intellectual powers toward finding ways to make these circumstances as amenable as possible? What is the best we can do for our students, while still minimizing the risks of spreading viruses to one another? In what courses is at least some in-person teaching necessary or optimal, and how can we preserve that option safely? I fail to see how compelling all faculty to be physically present is necessary for those efforts.

John
John
Reply to  Laura
3 years ago

“If I could do my job in some other way, why wouldn’t I, particularly given the vulnerability of the other people I need to protect?”

This crucially depends on ‘remote learning’ being an effective method of teaching. I think that the end of last semester showed most of us that it’s not at all comparable to in-class teaching. In fact, it’s often an utter mess. So I’m not sure it’s fair to say that we ‘can do our jobs’, tour court, remotely. And if that’s right, then some problems with the letter arise: why should students pay full fare for a Zoom lecture? And, next, why should we be paid full fare for giving one?

I have not heard a single proposal about how to overcome that very obvious problem. And that’s precisely the position the universities are in: we have faculty who want to teach from home while still being paid their full salaries; and we have students, who pay those salaries, who don’t want to fork out 50k for a Zoom lecture. What do we do?

Laura
Laura
Reply to  John
3 years ago

At some universities, students actually pay higher tuition for fully online courses. I have taught online before the pandemic made it a widespread practice, and while I do not think the quality was as good as in-person classes, I think it is likely to be as good as socially-distanced in-person classes. Right now, people in admissions and enrollment are working out the projected enrollments for Fall, based on current registrations and incoming firstyear students. We are seeing some drop in enrollment but it’s not extreme; mileage may vary at different institutions. I would think the solution is not for faculty to be forced to teach in-person, but rather for the university to communicate with students and parents who are considering taking this semester or year off, and finding what they need to make enrollment a viable option. For some, maybe it’s meeting in person at least 1/4 of the time but not having the entire course in person. For others, maybe specific courses are desired in person. But until we know what effect these things are having on enrollment, a blanket policy that forces people to assume higher risks makes little sense.

Daniel Star
3 years ago

I would like to clarify one thing about our letter and our views more generally. We wanted this letter to be highly focused and not overly long for strategic reasons. In an interview with the Boston Globe today (an interview that may well end up just being used for a few quotes for a broader article, although we’d also be interested in publishing an opinion piece), we made a point of emphasizing that our concerns to do with the interests of the faculty are concerns with faculty in the broad sense, including part-time lecturers, teaching fellows, etc. And I made a point of saying that “we very much do not wish to see the university moving the in class teaching burden on to teachers outside of the tenure system; that would be extremely unfair.” We also talked about janitors, cafeteria workers, etc., emphasizing (a) that some jobs very sadly cannot be done from home (unlike teaching), and (b) that lowering the population density of students and faculty on campus by allowing many courses to be taught online would decrease the risks of harms for all university employees that remain on campus.

Laura
Laura
Reply to  Daniel Star
3 years ago

This is a very important point: those of us who are able to do our work without increasing the risk of virus transmission have an ethical responsibility to keep the campus environment safer for those who must do their work in person, or who have less control over their working circumstances. I had assumed that this was common knowledge by now, but one goal is still to prevent the hospital system from being overwhelmed by new cases. Capacity percentages of hospitals near me are reported on daily. They are still able to provide optimal treatment because they aren’t past capacity, and the staff still have adequate rest and equipment even though they are under a great strain. Debating about whether the rates of transmission or mortality are this-or-that is beside the point, if the result is that hospitals cannot handle all the patients who need treatment, and thus people die needless deaths or suffer needlessly before recovery.

skeptical skeptic
skeptical skeptic
Reply to  Daniel Star
3 years ago

The idea that having classes online would make campus safer for custodians and staff is either extremely naive or disingenuous. If many students and faculty choose not to come to campus, then janitors, cafeteria works, maintenance staff, etc. will be laid off. If 1/2 the normal student body is on campus, 1/2 the cafeteria workers will be fired. We’ve already seen this even at the richest university, Harvard. If you want to reduce your own risk by working from home, just say so, and don’t muddy the waters by pretending that you’re protecting staff.

Laura
Laura
Reply to  skeptical skeptic
3 years ago

This is not happening at my university, so I can’t speak for what goes on at Harvard, but I don’t think it’s obvious that so many people are going to be laid off. Some classes are going to be held in person unless the government orders a statewide shutdown, and I can help make that environment safer for those who need to be on campus if I stay at home to teach my own classes. There is nothing about teaching my philosophy courses that works radically better in person right now, given that to teach in person we have to spread the class out far from one another and all of us need to wear masks, and I’m not putting people at risk for some small or questionable benefit.

Anatoli Polkovnikov
Anatoli Polkovnikov
3 years ago

I also work and teach at BU and I think this proposal is very unwise to say the least.. Faculty should be able to take unpaid sick leave if they want/need to. Who is going to pay their salaries? I both teach at BU and pay for my son at BUA full tuition and we are not going to pay 50 k a year for Zoom, I guess most parents will do the same. They can ask colleagues to teach for them and then repay them next year..

There is also a moral issue to this suggestion, Suppose there is indeed a risk, but suppose they got a gas pipe leaking in their house, do they want a plumber to give them a Zoom lesson how to fix it? What about doctors, shall they teach us via Zoom now how to cure ourselves? How about farmers, shall they give us all Zoom lessons how to milk cows? Why people expect others to work for them. If they want to solve a problem, they should offer something constructive, e.g. offer to do individual/small group instruction but 10 hours a week instead of four. If you have a problem then expect to work more, not less! Otherwise it is stealing money. I wonder if in India the discussion is at a similar level.

Irina Mikhalevich
Irina Mikhalevich
Reply to  Anatoli Polkovnikov
3 years ago

Anatoli:

I’m not sure how you arrived at your interpretations of this letter. Where do the authors suggest that someone else should “do the work” for faculty who refuse to risk their own and their families’ lives during a global pandemic? First, the letter is clear that ‘faculty’ includes all who teach at BU (full time TT, part time faculty, lecturers, adjuncts, graduate students, etc.). This means that the teaching burdens would not be shifted to anyone else, as you suggest. (To the contrary, the proposal aims to help everyone, including those most vulnerable to economic exploitation, such as contingent faculty and grad students). Second, the letter isn’t proposing giving a paid vacation to anyone who isn’t comfortable teaching in person. Instead, the authors would like for all faculty to have the option of teaching remotely–something that not only requires more work than in-person instruction, but which many peer institutions have already promised to their faculty as a matter of basic respect and fairness. (Note that this option is also something that BU has promised to its students, but not to its faculty). You ask who would pay for it. The answer is obvious: the university. How shall they pay for it? The same way that they always pay their employees’ wages. The suggestion that at-risk faculty assume the additional financial burden of paying their colleagues’ salaries to compensate for lost tuition-revenue is odd, to say the least.

The remainder of your comment is similarly perplexing. You seem to suggest that the real moral issue is that professors aren’t assuming their fair share of the risk (whether from COVID or generally). We expect people in other professions (plumber, doctor, farmer) to do their work in person irrespective of physical risk, and professors, you seem to say, should be held to similar standards. Putting aside the fact that doctors now conduct much of their routine work via telehealth and the obvious dis-analogies between lecturing and milking a cow, there are several problems with this line of reasoning. First, it falsely equates foreseeable low-level risks endemic to some professions with the unacceptably high risks of death and permanent disability from a novel virus in a context in which such risks were never expected. The latter is not comparable to the risk of injury from a gas explosion or the risk of being kicked by a cow—risks that are anticipated and voluntarily assumed by plumbers and farmers. Moreover, employers are obliged to take reasonable precautions to safeguard the health and safety of their employees. If a gas company were to send a plumber to inspect a leaky pipe without proper training or equipment, they would rightly be subject to legal liability and moral censure. Why hold universities to a lower standard? Just as doctors and nurses are rightly petitioning their employers and government for better protections, just as people working in meat-processing plants are requesting paid time off/testing/PPE, so too faculty are right to advocate for better protections for their own health. Third, your argument comes close to suggesting, absurdly, that as long as some people must risk their lives, it would be immoral to reduce one’s own risk of injury. But, the fact that essential workers face horrific risks to their health does not mean that others do not have a right to telework. Quite the contrary, it means that we must strive to reduce or eliminate the risk for the essential worker (and the systemic problems that have placed an outsized share of the burden of essential labor onto the shoulders of racial minorities and women, but I digress)—and we may do so at the same time that we advocate for ourselves.

I understand that you may not want to pay 50k for school via Zoom. That’s reasonable, and your choice to make, but it is entirely orthogonal to the argument advanced in the letter.

Anatoli Polkovnikov
Anatoli Polkovnikov
Reply to  Irina Mikhalevich
3 years ago

Dear Irina,

I was probably a bit harsh but I do not think the issue is different than what I wrote. Yes there are people at risk and these people should have an option to work from home probably offering much more instead. We have a close friend who is a doctor and who works now 6 days a week full time. These faculty at risk could ask others to substitute this year and repay next year and in our department many younger people offered such options.

But I am sorry you did not answer the question who is going to pay. If remote teaching is offered then most undergraduates will defer, let us face it. No one or almost no one will pay huge tuition for Zoom, there are many free online lectures. Plus for this money (or less) they c an go anywhere in the world and get normal education. So who is paying? I think then those who do not want to come should give up their salary, half of their salary, try other arrangements. You did not like my example with a plumber, but I did not get your argument. There are many plumbers in the same risk group as faculty, are you personally willing to pay same $160 an hour for a plumber who will give you many excuses for why he/she can not come and will offer Zoom instead? I am sure you will find another plumber and students are exactly like that. Graduate students are in the same category, they can defer for a year and look for another job.

One can have special arrangements on one on one basis. People will be ready to help to those in need but requiring this to be policy without giving a source of revenue or sacrifice is not an option, I am sorry.

John
John
Reply to  Irina Mikhalevich
3 years ago

Hi Irina,

I cannot defend Anatoli’s initial remarks, as I’m generally persuaded by your response to them. But I would like to try to shine some light on where they might be coming from.

You’re right that there is no obvious reason why A’s being more advantaged than B means that A must advocate /only/ for B, if s/he advocates at all. So I’m not sure there is anything morally out of order in the letter, per se. But I think what some of us find off-key is that the letter is written by people who are in positions that (i) already grant them remarkable advantages and (ii) that are directly sustained by exploitative labour practices. And yet they are arguing for additional advantages, all without obviously offering concessions, and (perhaps) staying silent at other times (e.g. not writing letters…) about the fact of (ii).

For example, the authors and others in their position will have fared well economically during this period of intense disruption, even though their (and our, and my) contributions to the economy are negligible compared to, say, those of nurses who are being laid off at for-profit hospitals right now. Yet here we are as /philosophy/ faculty, comparatively insulated from both the economic and medical ravages of this pandemic, insisting that our ‘deeply held’ preferences be heard.

So–and I fully admit there is no clear moral imperative here–sometimes when you have it very, very good, it strikes me that the decent thing to do is show gratitude by recognizing this, and then seeing what concessions–sacrifices, even–you can make to help.

Crucially, this does not necessarily that mean teaching in-person is the only concession available. It means only that perhaps that /some/ concession is called on by decency. And yet the letter, from what I can tell, mentions nothing of the sort. The issue Anatoli is raising about sustaining faculty wages in the face of reduced tuition revenue is just the most obvious example of where this conflict takes root. Though I will admit to rolling my eyes at insisting taking sick pay is out of order, as the university should consider the domestic risks of availing one’s self of that job perk (what?!).

So, is the letter morally wrong? I don’t think so. It’s just tone deaf.

Laura
Laura
Reply to  Anatoli Polkovnikov
3 years ago

Personally, I don’t want my kids to be forced to attend in-person classes, and would advise them to take online options wherever possible. This is not a permanent state of affairs; this is a stopgap measure to get us through a very unusual event. I confess I am struggling to understand the general response to this problem that runs along the lines of “other people have it worse, so everyone should have it worse”. How is that reasonable?

Anatoli Polkovnikov
Anatoli Polkovnikov
Reply to  Laura
3 years ago

Students at BU has an option to choose whether they want to study in class or from home. The truth is that no one will pay huge tuition for Zoom. They will defer. For students this virus poses no risk and it will not be gone, at least for long, so it is not a temporary situation until we get herd immunity at least. It is destroyed by sun and goes down in summer like flu.

For older people including some professors there is a risk of course but they want to get full salary for which no one will pay. They cannot expect to do part of their duty for whatever reason and expect that students will pay you in full whether your are a teacher or a plumber. I think this is obvious. Harvard is planning to be remote and most students already deferred. BU does not have such a huge endowment to afford loosing revenue without deep salary cuts. Anyway my point was that the constructive approach is that we, faculty, have to offer something back not just wane for salary from administration offering nothing in return.

Best, Anatoli

Roberta Clarke
3 years ago

I would suggest that the skeptical skeptic and others trying to quantify the risks of placing faculty in the classroom have, in their ignorance of the wider implications of the disease, failed to consider outcomes of the virus other than death. They are missing the mark by not considering the virus’ sequalae. The virus is so new that the time to identify long-term sequelae has been too short to allow what will later become a full exploration. However, for example, Medpage reports 57.4% of 841 hospitalized patients in a study developed neurological symptoms such as acute ischemic stroke (i.e., long term disability), Guillain-Barré syndrome. and inflammatory demyelinating polyradiculoneuropathy which can last for months and can return throughout life. NIH is reporting significant neuropsychiatric effects and post-infection autoimmunity diseases with lifelong implications. Clearly we will discover more sequelae once we have controlled the spread of the virus and have a longer time frame over which to see the long term effects of this vicious virus. You are considering only dead faculty members, not those who, due to the decision to place them in the classroom, may face long term disability and poor functioning.

Roberta Clarke
3 years ago

You might further want to read
“COVID-19 Can Last for Several Months: The disease’s “long-haulers” have endured relentless waves of debilitating symptoms—and disbelief from doctors and friends.” For those of you willing to write off your older colleagues as mere necessary sacrifices to the cause of returning to normal (whatever that may be), note the article states: “Most have never been admitted to an ICU or gone on a ventilator, so their cases technically count as ‘mild.’ But their lives have nonetheless been flattened by relentless and rolling waves of symptoms that make it hard to concentrate, exercise, or perform simple physical tasks. Most are young. Most were previously fit and healthy.” The young are not as protected as many believe. We have yet to learn what damage this virus can do to everyone, due to the short time frame in which we have been able to collect data. See the link below to the Atlantic article.

https://www.theatlantic.com/health/archive/2020/06/covid-19-coronavirus-longterm-symptoms-months/612679/?utm_source=STAT+Newsletters&utm_campaign=87e3d28c70-MR_COPY_01&utm_medium=email&utm_term=0_8cab1d7961-87e3d28c70-149635341

Irina M
Irina M
Reply to  Roberta Clarke
3 years ago

Absolutely right. The fact that COVID may lead to permanent injury and disability in those who do recover is so often ignored in conversations about risk. As a doctor whose name escapes me at the moment (I’ll try to track down the reference) recently wrote, in many ways COVID is more like HIV and other debilitating diseases than like a transient flu.

Laura
Laura
Reply to  Roberta Clarke
3 years ago

Yes. Some students may be dismayed at the shift to online teaching, but others want the security of being online right now, and still others have parents who prefer that security! My kids are asthma-prone and others in my family have serious risk factors. I don’t want my kids getting sick regardless of how much we hear about older people being more at risk, and I don’t want my kids transmitting the virus to their relatives either. Given that most people seem to agree with lockdown and social distancing plans that were pursued this spring, I wonder about the assumption that we need to be able to offer as many in-person classes as possible. For many students, I imagine the far more dismaying prospect involves a lack of close social contact with other students. To solve that, depending on the living arrangements at different types of colleges and universities, perhaps small groups who live in close proximity can decide to quarantine together, or if they are groups of e.g. first year students, perhaps they can even take a gen ed course or two together as a cohort and thus reclaim benefits of in-person classes for at some of their work.

skeptical skeptic
skeptical skeptic
Reply to  Laura
3 years ago

“Given that most people seem to agree with lockdown and social distancing plans that were pursued this spring, I wonder about the assumption that we need to be able to offer as many in-person classes as possible.”

I think this is the assumption that needs to be questioned. There are two ways of thinking about the lockdown:

1. The lockdown was designed to flatten the curve. That is, the lockdown was a temporary measure designed to prevent hospitals from being overwhelmed.

2. The lockdown was designed to keep us safe until there is a vaccine.

The original argument for the lockdown was (1). Because of the images from Italy and China, most people though hospitals really would be overwhelmed all over the country without lockdowns. So they agreed with (1), at least when (1) was presented as something that would last for only a few weeks. But your assumption that I’ve quoted seems to be based on (2). The problem with (2) is obvious: it requires that the lockdown last at least another 6 months and possibly another year or so.

Put another way: if the purpose of the lockdown was to flatten the curve, that’s already happened. Hospitals are empty in most parts of the country. So we should open everything up.

On the other hand, if the purpose of the lockdown was to minimize risk until there’s a vaccine, many people (myself included) think the cost is too high and the risks are too minimal. Even if you disagree with that argument, you should acknowledge that you need a much stronger argument for (2) than for (1), because the costs of (2) are much higher than the costs for (1).

Besides, here’s the reality: most people are not going to stay locked up inside for another 6-12 months. Most people are going to interact with others, send their children to school, go shopping, see friends, and so on. If you’re doing those things, the additional risk that you take by going to an in-person class is probably very small. If I were a student I would absolutely take that risk. My institution seems to be moving to in person teaching and as faculty I have no problem with that.

Anatoli P
Anatoli P
Reply to  skeptical skeptic
3 years ago

I totally agree. Flattening the curve happened in late March. Now what? Do people seriously believe they will defeat a virus like that, once it spread to over 50,000,000 across the world (all statistics shows that official numbers are at least 10 percent lower than real as mild cases are never recorded). Apart from a negative economic effect obviously more older people will die because of lock down as we will get herd immunity later and it is harder to isolated for a long time. In a couple of years with continuing lockdown we will start to see increases in other deaths because many people would loose access to health care with their jobs, would miss cancers and other diseases, which could be cured if detected early during routine visits. There is a huge toll on kids who do not go to schools and do not develop emotionally and socially as the should (apart from education). It is almost clear now that there will be no efficient vaccine, there were many announcements, which most likely were money driven. I honestly do not understand what people want to achieve now. I also read the Stanford article, for an average individual under 65 the risk of dying from Covid is comparable to the risk of dying in a car roundtrip from Boston to Florida. For such individuals without preconditions it is 50 times lower (according to the published statistics approx. 98.3% death are people with preconditions) so we are roughly talking about a roundtrip from Boston to Worcester, an average daily commute for many people. I really do not understand what people want from this lockdown behaving like an ostrich, who escapes from danger by hiding its head in the sand.

Irina Mikhalevich
Irina Mikhalevich
Reply to  skeptical skeptic
3 years ago

“Most people are going to interact with others, send their children to school, go shopping, see friends, and so on.”

And that is how we end up back at (1): needing to flatten the curve so as to avoid overwhelming hospitals.

“If you’re doing those things, the additional risk that you take by going to an in-person class is probably very small.”

It may be small risk *relative* to the huge risks you are already taking, but it is not small in absolute terms. Driving drunk may pose a *comparatively* small additional risk for someone who (e.g.) skydives for a living, but we wouldn’t say that the skydiver therefore takes a small risk when she decides to drive home drunk.

Anatoli P
Anatoli P
Reply to  Irina Mikhalevich
3 years ago

The risks are very small on everyday scale for most people. Just check the numbers. Suppose we take the official figure of 0.3% death rate, divide it by a factor of 5 for people under 75 (most people who work belong to this category). This number is already small, now divide it by 50 to get the risk for people without preconditions. It is tiny. Of course those at risk should be maximally isolated and protected, but it is a small fraction of population. Now risks of prolonged isolation are very high and accumulate fast too. At the end people will realize there are far more deaths from undiagnosed conditions but it will be too late. Next do you have an estimate for increased risks for those who loose health insurance? And finally I do not understand why people do not realize that longer isolation simply spreads the number of deaths over time, does not decrease it, likely increases it as long isolation for vulnerable is more difficult. Many companies are now making money on spearing scare and likely fuel this. If you ask any person will he/she risk loosing their house, access to schools for their kids, access to entertainment over doing one round trip to Florida what the answer will be? We already have 3 months of isolation, just extrapolate to another 3 months, then another 3 months and so on. What will change? Vaccine? We have flu and no effective vaccine, this will likely be a similar situation. Plus a properly untested vaccine will pose much higher risks to kids and young adults than COVID. For 50 million people we heard about a dozen cases (say even 50) when kids died. It is 1 per million, it is far far less dangerous than most other things they endure including ordinary flu. Do not believe me, read this paper or at least conclusions there: https://www.medrxiv.org/content/10.1101/2020.04.05.20054361v1.full.pdf. There are many studies like this now, just somehow mass media/politicians do not talk about this. In early March no one understood what the virus is about. Now they are either afraid of admitting own mistakes or are paid very well by companies making huge profits now. My only hope now that Europe/Southern States, where most restrictions are being lifted and where nothing bad will certainly happen, will show that long isolation is a loosing strategy from any standpoint.

David Wallace
Reply to  Anatoli P
3 years ago

Can you provide sources for those ‘5’ and ’50’ numbers, and an evidence base for the claim that a vaccine will pose ‘much higher risks’ than COVID for children and young adults, and that flu vaccine has no significant effect on the epidemiology of flu? I can’t find them in the paper you cite. I also note that that paper includes hypertension as a precondition and notes that 38% of the 40-59 population have it.

I’m potentially sympathetic to the ‘risks are being overstated for the non-elderly’ argument, but there is a danger of hyperbole in the other direction.

Anatoli P
Anatoli P
Reply to  David Wallace
3 years ago

I agree that going to another direction is dangerous but by now there is enough statistics. I will mention some of the sources I found, I personally cross-checked them against data in various countries/states and found that say Massachusetts is consistent with Italy and Illinois.

Here is death age distribution (data from late April)

https://www.boston.com/news/coronavirus/2020/04/27/massachusetts-covid-19-deaths

I made an assumption that about 2/3 in the age group 70-79 are over 75. This fraction is probably even higher, but does not matter. With this assumption in mind I got 77% of deaths are people over 75; 84% rate for those over 70. So we got a factor of 5 roughly for people who work depending on how you estimate average retiring age (retired people can self quarantine for a while with no need for lockdown)

There is a batter link on mass.gov with the same data

https://www.mass.gov/doc/covid-19-dashboard-april-27-2020/download

On page 13 it gives a number 98.1% death rate for people with underlying conditions. People with preconditions of course should get all the resources, if cannot work remotely, like free matching salary from feds or the state instead of unemployment benefits to high school kids or healthy individuals many of whom enjoy free money for doing nothing. . But for the rest of the population we get the risk lower by another factor of 50.

Final piece of information. There were many random anti-body testings in Boston (Chelsea), California (Palo Alto if I remember correctly)., NYC, Northern Italy which all show that actual infections are factor of 10 higher than reported. We personally know around 10 people who almost certainly had it, when some called doctors describing symptoms, doctors said it is likely it and advised to stay home unless they feel worse. These cases were never recorded. I know from my colleague in Biology department that many BU faculty went to Biogen, very few tested later and 100% were positive. So BU only probably has a good fraction of faculty who had it already. This was all before the lockdown so these faculty continued to teach.

I checked with colleagues who study this and they confirmed, at least qualitatively that all risks are grossly exaggerated.

Now combine these facts together and you see that for say an average student driving from Boston to Worcester poses higher risk of dying than dying from Covid.

David Wallace
Reply to  David Wallace
3 years ago

This is not a valid use of statistics.
1) The probability of not having a pre-existing condition conditional on dying of COVID is not the same as the probability of dying of COVID conditional on not having a preexisting condition.
2) More importantly, the probability of having a pre-existing condition is strongly dependent on age, so you can’t treat the data on age and the data on pre-existing conditions as independent.

Over and above these statistical issues, virtually everyone who dies of a not-very-lethal disease turns out to have *some* pre-existing condition (it’s not as if the disease determines who to kill just by rolling dice). But many of those conditions are fairly common. The link you provide doesn’t give its definition of ‘pre-existing condition’ but various of the sources I’ve looked at (e.g., the CDC) include obesity, history of heart disease, hypertension, diabetes, past history of smoking, and past history of cancer. Collectively, quite a high proportion of people in the 40-65 range have a pre-existing condition by that basis.

As for your final example, Google Maps tells me that it is 47 miles from Boston to Worcester. The fatality rate from driving in the US is about 150 per 10 billion passenger-miles. The chance of dying in a drive from Boston to Worcester is then about 1 in 1 million. The fatality rate from COVID-19 among the student-age population is no doubt low but I haven’t seen data (and you don’t give data) to support it being anything like that low.

David Wallace
Reply to  David Wallace
3 years ago

(Actually, just on the driving case: maybe you mean that the risk of a *daily commute* between Boston and Worcester is higher than the COVID-19 risk? That sounds more plausible.)

Anatoli P
Anatoli P
Reply to  David Wallace
3 years ago

I do not know how they got the data about preexisting conditions, true it is not written there, but I assume it has to me meaningful. If they say that almost everyone had such a condition, it is a meaningless figure. I assume that it is a medically recorded condition, which is not common.

I think the number 1 per million is about right for students and likely healthy middle aged people. The fatality rate in the age group around 20 is definitely comparable. In the Mass. data there is one case among roughly 10,000 recorded in the age group 20-29 and zero below 20. Of course this is not enough for full statistics but clearly the numbers are very small. Most likely that person had preconditions. What I heard most common preconditions are severe diabetes or cancer. I do not know the data here, but I assume this should be at least a large fraction of people with preconditions who fall in to this category.

In any way if we would start the discussion here and will try to estimate risks for various age groups and compare with other risks we would have a constructive discussion of what the price (in terms of money, lifestyle and health) we as a society are willing to pay and how to minimize the price. This discussion is not even there. Some models, which people developed to justify lockdown show actually more deaths if we prolong the whole thing. What do we achieve by the lockdown if there is a zero chance to eliminate the virus. I do not understand. Obviously six months from now everyone will be saying how smart Sweden was. Anyway this is a separate issue.

I found another official piece of data:

https://data.cdc.gov/NCHS/Provisional-COVID-19-Death-Counts-by-Sex-Age-and-S/9bhg-hcku

If you look into overall death rate by age group (including those with preconditions) then for ages under 24 less than 1% of deaths in USA are attributed to Covid. So for these people the risk of dying from Covid is not higher than from other causes, rather quite a bit lower. Then this number increases to about 10% for older age groups. Now by all estimates the real number of people who got Covid in USA is at least 5% of the population If I use the factor of 10 found in many places. So we see that in the worst case scenario the chance of dying from Covid is 2 times higher (10% times 20) than from other causes. Now if you exclude people with severe preconditions (cancer, diabetes) you will get probably not 50, but say a factor of 10 back and you get that the risk of dying from Covid, assuming every single person goes through it is 5 times smaller than from other reasons.

What is on the other side. If healthy people take this risk, they become immune at least for a long time and not dangerous to vulnerable people, so almost certainly the overall number of deaths will go down of only select, high risk, fraction of population quarantines itself.

This is of course separate from the original article, but I come back to my point, why students, who have less than 0.01% risk of dying from it will pay for Zoom? If is is a question of safety of faculty then why students should pay with their loans or money got from their parents who were saving for years should pay for it. Why faculty who do not want to take the risk should not sacrifice. I do not get it. Also if they buy good PPE they do not increase their risk by going to BU. I agree with Skeptic that there are so many other activities we tae anyway so that teaching is not a big deal.

Anatoli P
Anatoli P
Reply to  David Wallace
3 years ago

Also I am sorry, I do not understand what is wrong with statistics. If out of every 100 people who died 98% had preconditions we only need one more reasonable assumption that infection rates are the same irrespective of the precondition to conclude that there is a 50 times higher chance to die from preconditions. Statistics could be tricky with conditional probabilities, but here I do not see any issues. As we learn in high school at large numbers probability is total number of outcomes divided by total number of experiments (cases).

David Wallace
Reply to  David Wallace
3 years ago

Let me illustrate. (WARNING: these are purely illustrative numbers. They’re not based on any evidence.)
Firstly, to see why your argument above doesn’t work, suppose that only 1% of the population have a pre-existing condition. Then for 98% of the deaths to be among people with pre-existing conditions, the risk to those people would have to be vastly higher than 50x the risk to people with no condition. Conversely, suppose 98% of the population have a pre-existing condition. Then the 98% figure is what you’d predict even if having a pre-existing condition had no effect on your risk of death. You’re effectively conflating Pr(pre-existing condition|death) (which is 98% in the study you quote) with Pr(death|pre-existing condition).

On the more general statistical point, here’s a model. Suppose we have a population of 100,000 people who are exposed to COVID-19. Of those 100,000, 20% are elderly, and 80% are not. Of the elderly, 95% have a pre-existing condition. Of the non-elderly, 50% do. And suppose the death rates are:

Elderly, pre-existing condition: 4%
Elderly, no condition: 1%
Non-elderly, pre-existing condition: 1%
Non-elderly, no condition: 0.1%

Then the total deaths are

Elderly, pre-existing condition: 760
Elderly, no condition: 10
Non-elderly, pre-existing condition: 400
Non-elderly, no condition: 40

That works out as:
Overall death rate: 1.2%
Death rate among the elderly: 3.9%
Death rate among the non-elderly: 0.55%
Fraction of deaths who are elderly: 64%
Fraction of deaths with a precondition: 96%

On your method, we would

– start with the average death rate of 1.2%
– reduce it by a factor of about 3 (because only about 1/3 of the dead are non-elderly), to get 0.4%
– reduce it by a further factor of 25 for no pre-existing conditions (because only 1/25 of the dead have no pre-existing conditions), to get 0.016%.

But in both cases that gets the numbers wrong – wrong by about 20% for the non-elderly (their risk is 0.55%, not 0.4%) and wrong by a factor of 6 for the non-elderly without pre-existing conditions (their risk is 0.1%, not 0.016%).

Anatoli P
Anatoli P
Reply to  David Wallace
3 years ago

I agree with your estimate but even in your example you get the same number within a factor of 2. Also I was checking myself regularly Boston.com and the data works in the opposite way from your example, they published list of deaths by age every day, around 100 during the peak and for many more older people preexisting condition was marked as “no”. Conversely the number of younger people with no preexisting conditions was basically zero (it was either yes or unknown at that time). So if we assume non-uniform distribution of death of people with preexisting conditions by age, then younger people without them are even safer. But in any case I am not insisting on the number 98%, let us say it is 95%.. Whatever you do with your models, this 98% is so high that you will not be able to lower the ratio much. Even if you say that 100% of older people (80% of death) had preexisting conditions then you will have to say that 90% of remaining 20 have it. This is the lowest possible number you can get and as I wrote the numbers seem to work in the opposite direction.

David Wallace
Reply to  David Wallace
3 years ago

It’s a factor of 6, not of 2, but in any case my underlying numbers are just invented for the example. It’s there just to illustrate the mathematical error you were making in the previous post, nothing deeper.

Anatoli P
Anatoli P
Reply to  David Wallace
3 years ago

I understand, but 6 is the worst case scenario, which can not be true this asymmetry likely works in the opposite direction an older person without preconditions is more likely to die from the virus. In any case I was doing estimates on the log scale 1:10,000 vs 1:1000. These factors of 2 you will get from various models are not changing them much. If you take data from https://data.cdc.gov/NCHS/Provisional-COVID-19-Death-Counts-by-Sex-Age-and-S/9bhg-hcku
then even without preconditions and again with mild assumptions that only 5% of population got infected by now you will get that for average person under 25 the chance of dying is 1:100,000 . Without preconditions even if it is not 1:50 but 1:10, which is probably wrong, we will get 1:1000,000. One roundtrip to Worcester!

David Wallace
Reply to  David Wallace
3 years ago

I find some of those numbers dubious. But since the context of my comment in this bit of the thread was simply the fact that you were making a straightforward math error – and since as far as I can see you are not acknowledging that error – I don’t see a lot of point pursuing this further.

ehz
ehz
Reply to  David Wallace
3 years ago

Anatoly, you wrote:

“If out of every 100 people who died 98% had preconditions we only need one more reasonable assumption that infection rates are the same irrespective of the precondition to conclude that there is a 50 times higher chance to die from preconditions.”

I think that’s incorrect — you also need the assumption that the there is the same amount of people with precondition and people with no preconditions. Otherwise, the ratios vary. It’s mathematically consistent with these figures that 100% people with *no* preconditions die, for instance, if there’s a lot more people with preconditions.

ehz
ehz
Reply to  David Wallace
3 years ago

Sorry for misspelling your name! I just noticed.

Anatoli P
Anatoli P
Reply to  David Wallace
3 years ago

If 100% percent of people are with preconditions then this indeed does not make sense. I do not know what exactly the percentile of people with preconditions in their counting, but I checked definitions on google and they only list chronic deceases. So we are talking about a fraction of population, how high I do not know. It is hard to find exact statistics, here is some info I found about Louisiana: “Forty-one percent of the first 83 COVID-19 deaths in the state were patients with diabetes while 31% had chronic kidney disease. Twenty-eight percent had obesity as an underlying health condition while 23% had cardiac issues.”

These are very specific and chronic/lasting conditions. I am not a medical doctor to know how common they are, but this could be of course a point of discussion how to define high risk groups and protect them. Once the discussion moves from close everything to how we protect high risk groups and start making some meaningful choices I will be very happy. Right now, I think we are not going anywhere with this, just slowly ruining our lives and most importantly lives of our kids, who have next to zero risk.

Anatoli Polkovnikov
Anatoli Polkovnikov
Reply to  David Wallace
3 years ago

These are high numbers of course. I do not really know how preconditions we’re counted in 98%. My reading of other sources is that they were chronic, around 60% or 70% (forgot) of those who died and had preconditions requiring prior hospitalization. It is obvious that people who have more severe conditions disproportionally bear very high risk. Mild conditions are still low risk (depending on definition of low of course).

I would not be Personally objecting total lockdown if I would see it leads us anywhere. I only see that it will lead to higher total damage. Maybe you can describe me a possible successful scenario, say a year from now beyond hoping for vaccine. Maybe I do not have enough imagination. There should be a purpose of a huge sacrifice, which we forth many people to make, especially kids who have a zero risk, even if we ignore the precondition factor. It used to be “flatten the curve”. What is it now?

Best, Anatoli

Anatoli Polkovnikov
Anatoli Polkovnikov
Reply to  David Wallace
3 years ago

Also to add to this as you mentioned a large fraction of older people eventually develop preconditions. So I guess these numbers will be significantly lower for non-retired population, maybe still significant, but not as high. At least all cases I saw of young people who died had very severe preconditions like cancer, but I do not know statistics.

Anatoli P
Anatoli P
Reply to  Anatoli P
3 years ago

David, we do not need to continue indeed. But I do not think there is any error in my argument (some uncertainty). Here is another estimate for people under age 24 (student population). Total deaths (same sources as before) =137, total population in that age group 104 million. With 6% infection rate (assumption) 6.2 M are infected. The mortality rate is 2 10^{-5}, i.e. one per 50 thousands. You do not like my factor of 50 for preconditions. Use whatever you like, 20? It is still one per million. There is really no escape for student population from kindergarten to graduate school this virus poses zero risk on everyday scale. I did not analyze other severe outcomes, which are not deaths but I suspect the situation is very similar as statistics of hospitalizations by age is not drastically different from statistics of death. By the way I am also in mid forties and also have some mild preconditions. I am not advocating anything for myself, just trying to find some common sense.

skeptical skeptic
skeptical skeptic
3 years ago

Irina,
Your comment is a prime example of the sorts of exaggerations and conflations that make these debates go in circles. Two things:

1. “And that is how we end up back at (1): needing to flatten the curve so as to avoid overwhelming hospitals.”

That makes no sense in this context. You think that with states reopened, the question of whether classes are in person or online is going to be the deciding factor in determining whether hospitals are overwhelmed? What possible reason could we have for thinking this is true? If people are interacting (hopefully while wearing masks etc) anyway, why think that in person classes pose a sufficiently large additional risk to merit banning them? Again, it is not the case that most people are deciding between (a) staying at home all the time and only venturing outside to go to class, and (b) staying at home all the time and watching lectures on zoom. Most people are deciding between (c) engaging in an ordinary range of activities, hopefully while wearing masks, and also going to in person classes, and (d) engaging in an ordinary range of activities, hopefully while wearing masks, and watching zoom lectures. You may not like it, but that’s the reality, as a glance at any newspaper with stories on reopening will show you. Posing this as if it’s a choice between (a) and (b) is either naive or intentionally misleading. To be perfectly clear: the vast majority of people are not going to stay locked up at home for the next 6-18 months waiting for a vaccine.

2. Second, in response to my claim ““If you’re doing those things, the additional risk that you take by going to an in-person class is probably very small,” you write, “It may be small risk *relative* to the huge risks you are already taking, but it is not small in absolute terms. Driving drunk may pose a *comparatively* small additional risk for someone who (e.g.) skydives for a living, but we wouldn’t say that the skydiver therefore takes a small risk when she decides to drive home drunk.”

This is just rhetoric. First, what “huge” risks am I already taking? I mentioned normal activities that most individuals are undertaking (shopping, seeing friends, sending kids to school, etc). Again, let’s be clear: there is zero chance that the majority of people are going to stay at home for the next 6-18 months waiting for a vaccine. You are again posing a false dichotomy. You’re also greatly exaggerating the risks, as the comments above indicate. If you’re a healthy person under the age of 65, the risks you incur from covid infection are comparable to risks that we undertake daily without thinking about it.

Second, do you think there might be a slight difference between drunk driving and going to classes in person? For example, might there be (a) a difference in how risky these activities are, (b) a difference in the costs and benefits, (c) a difference in the moral permissibility, (d) a difference in the legality? Do you think our intuitions about the impermissibility of drunk driving might be based on factors other than the degree of risk? These shouldn’t be hard questions.

David Wallace
Reply to  skeptical skeptic
3 years ago

Hang on. If you thought that systematically we should be minimizing large in-person gatherings with the goal of crushing the pandemic down to the level that contact tracing could keep a lid on it, it wouldn’t be a good objection to say ‘this given thing, in isolation, won’t achieve that’. And while I agree that it’s not realistic to suppose that everyone will stay at home for 6-18 months, it is a lot more realistic to suppose that large in-person events (e.g., lectures) can be kept to a minimum in that time.

That said, I don’t think that works as a defense of the actual BU letter. It would be a case for the university actually banning large in-person lectures, not for making it optional for faculty. The letter is mostly focused around individual health risks for faculty, not on epidemiology.

Anatoli P
Anatoli P
Reply to  David Wallace
3 years ago

My main problem with the latter was that it did not say who will pay. Even if all risks are true, asking other people: doctors, plumbers, grocery workers to take them and ask these people who save for year for tuition to cover full salary for not showing up is deeply immoral, I am sorry for using this word but I stand by it. Additionally it is also impractical; students will not pay $50+k a year for Zoom. What for? They can listen lectures in YouTube for free or simply defer for a year and do something else. That letter was explaining why someone else should pay high salaries of faculty without offering any sacrifice in exchange. If they would say that they are willing to take 50% pay cut for not taking the risk, or do other things in addition, it would possibly be a starter. Otherwise it is not.

David Wallace
Reply to  Anatoli P
3 years ago

The coherent case I’m presenting (which, I’ve acknowledged, is not quite the case in the BU letter) is that we can control the pandemic without total lockdown if we fairly systematically minimize those in-person gatherings that we can minimize and use contact tracing to squash those outbreaks that happen anyway, and that higher education teaching is in the list of such activities. That’s not really a moral argument: it’s not about individual risk but about epidemiology. I don’t even know if I accept that case, though I’m pretty sympathetic to it.

ehz
ehz
Reply to  Anatoli P
3 years ago

“Additionally it is also impractical; students will not pay $50+k a year for Zoom. What for? They can listen lectures in YouTube for free or simply defer for a year and do something else.”

Students will not pay $50+k a year just for in-person lectures, either. One big motivator in paying such large sums to attend some university is that at the end you get a stamped document that says that you have graduated from Prestigious University with a degree in Impressive Studies. For better or worse such things are highly valued in our society, so people pay a lot of money for them.

I’m also not sure that in-person lectures are much higher quality than online lectures provided over Zoom or similar platforms. My impression, for instance, is that you can find many youtube lectures that are much better than your average university lecture on the same subject. Why would people still choose the latter? See paragraph above.

Anatoli P
Anatoli P
Reply to  ehz
3 years ago

I do not think you are right, students do learn a big deal at Universities. I think your remark would be very offensive to most of them. They will get degree even if they defer, Anyway I know from my Harvard colleague that this is reality even there, where stamp is much more valuable, most students deferred after Harvard announced no in person classes. I also know that Cambridge UK has another model with online large classes but in person small group instructions and labs. This is their model anyway. If we implement something like that in BU, faculty will need to spend more time at BU. Anyway, I think this is a reality and BU administration is fully aware if it. The moment BU announces online only classes most students defer and even if half do, there is a 500, million gap in budget, But my understanding that it will be more than half.

Jamie
Jamie
Reply to  Anatoli P
3 years ago

“most students deferred after Harvard announced no in person classes.”

I hadn’t seen that announcement. Can you point me to the announcement that Harvard won’t have any in-person classes?

Anatoli Polkovnikov
Anatoli Polkovnikov
Reply to  Jamie
3 years ago

This is what I heard privately from a faculty member there that Harvard is moving in that direction and that most students are deferring. This is a private information so I do not know if it is final already.

Best, Anatoli

Irina Mikhalevich
Irina Mikhalevich
Reply to  skeptical skeptic
3 years ago

Skeptical Skeptic:

David Wallace is right that we should keep the epidemiological questions and the moral questions distinct. You and Anatoli believe that the risks from COVID are too low to justify giving faculty the freedom to decide for themselves whether to teach remotely. Our disagreement seems to be (largely) empirical. With that said:

1. Apologies for the vague referent. By “this,” I meant the large-scale relaxation of physical distancing, including by seeing friends, sending children to school, and resuming in-person instruction prematurely and without adequate safeguards. Each of these may be a contributing cause to a rise in cases; none are likely to be the deciding factor. The more of contributing causes we have, the larger the spread of the disease will be. It may not overwhelm hospitals, or it may; either way, the rise in cases translates into a rise in deaths and life-long disabilities.

2. I think you misunderstood me. I wasn’t drawing a comparison between in-person instruction and drunk driving. That would be silly. The example was meant to illustrate a mistake that I thought you were making: namely, of confusing relative risk (or perhaps subjective risk) for absolute risk (or perhaps objective risk).

Irina Mikhalevich
Irina Mikhalevich
Reply to  Irina Mikhalevich
3 years ago

Addendum: I would like to second DW’s call for epistemic humility for all the reasons he cites, given the unfortunate politicization and recent failures of the CDC and other American institutions, and because none of us in this conversation have the relevant expertise.

skeptical skeptic
skeptical skeptic
3 years ago

Yes, that’s true. I was reacting to the idea that the arguments above seem to be going in circles.

If the argument is “in order to minimize infections, we should stay at home until there’s a vaccine”, it’s unrealistic and in my view not supportable. The costs are too high.

If the argument is “in order to minimize infections, we should minimize large gatherings until there’s a vaccine”, this seems to me to be an argument for minimizing the number of large lectures, rather than minimizing in person classes in general. From what I can tell, BU and other universities are already doing the former; no university that I know of is planning to cram hundreds of students into ordinary classrooms.

If the argument is “in order to minimize the risk of infection to individual faculty, we should allow them to teach online until there’s a vaccine,” we’d need to be clear about what the risks actually are and what the costs of the risk-reduction are. If it turns out that the risk to most faculty is quite small, as I think the evidence indicates, and if the costs of risk-reduction are very high, then this argument won’t entail that we should allow (healthy, under 65) faculty to teach online.

In general I think we need to be careful about unsupported claims about “huge risks”, comparing teaching in person to drunk driving, assuming that lockdowns are going to continue, etc.

David Wallace
Reply to  skeptical skeptic
3 years ago

I agree that ‘everyone stays home until there’s a vaccine’ is not viable.

I am less sure about the value of avoiding in-person teaching as a method of restricting spread. It is true that small-group teaching is going to be safer, but it is also true that university teaching involves mixing people around a lot according to their individual curricula, and that that’s potentially bad news epidemiologically. (Compare the situation with school re-opening, where it’s more realistic to create a ‘bubble’ of 10-15 students who don’t interact with students outside that bubble. My take here probably depends on the details, in particular what the background level of the virus is and how robust the contact-tracing framework is at the time.

I agree with the principle that we need to be clear what the risks actually are. And I agree with you that some of those risks have been over-catastrophized. But it is possible to overreact to that by being too sanguine. There is a lot that we simply don’t know, and estimates in the literature of case fatality rates and the effects of various pre-existing conditions on those rates vary extremely widely and don’t show much sign of convergence yet. From a philosophy of science point of view, anyone who’s not themselves an expert on infectious diseases should probably react to that by having quite a lot of epistemic humility about the real rates, not by trying to make their own judgement on which rate is most plausible. (That’s a general comment, not aimed specifically at you.) And of course we know next to nothing about the long-term morbidity, because COVID-19 hasn’t been around for long enough.

I’m male, in my early forties, in fairly good health but with some relevant pre-existing conditions (like a pretty large fraction of people my age). It is plausible that when the dust settles the fatality rate for people in my situation who contract COVID-19 will be around 0.1%, and that I have an 90% chance of mild symptoms for a couple of weeks and a 10% chance of being quite ill for a couple of weeks, but with very low risks of long-term effect. But it is also plausible that the fatality rate will be closer to 1%-2%, and that there will turn out to be a10%-20% chance that I will get seriously ill for weeks, be in substantial distress during that period, and end up with lung damage that affects my health for the rest of my life. (I share others’ views that these long-term health risks may turn out to be the most severe consequences of the pandemic.) If the former is correct, in retrospect it would be silly of me to avoid in-person teaching. If the latter is correct, in retrospect it would be pretty sensible.

(For the record, if Pitt gives me the choice I’ll probably choose to teach in-person, at least on the basis of what I know at present. But I don’t think the reverse assessment would be indefensible given the current state of the science.)

Irina Mikhalevich
Irina Mikhalevich
3 years ago

Anatoli:

Here is a helpful blog post about the flaws in the Stanford study you cite.

https://statmodeling.stat.columbia.edu/2020/04/19/fatal-flaws-in-stanford-study-of-coronavirus-prevalence/

Anatoli P
Anatoli P
Reply to  Irina Mikhalevich
3 years ago

This is a different paper. I do not understand the link. Are you implying that all Stanford papers are wrong? I can try to find others.

I think the factor of 10 discrepancy at least between reported and actual Covid cases is confirmer by many many studies and agrees with what we know. If you call the doctor here and describe your Covid symptoms but say that you are OK they will tell you to stay home. So none of mild cases is detected unless they are in a very small group of people, where everyone is tested. And then from examples I know the percentage of infected people is a way higher than average reported. So everything is consistent. You do not need to be a specialist to understand how numbers work.

Also I got the reference from people I trust (also faculty at BU) who do study Covid and publish research papers, so I trust their selection of papers. Those numbers agree with my own analysis based on published data.

Lisa S
Lisa S
3 years ago

The Ioaniddis preprint (not yet peer-reviewed) is flawed, as discussed in this twitter thread here: https://twitter.com/gidmk/status/1270490491600003072?s=21

Here’s a peer-reviewed preprint that estimates the U.S. IFR at 1.3%: https://www.healthaffairs.org/doi/full/10.1377/hlthaff.2020.00455?utm_campaign=covid19fasttrack&utm_medium=press&utm_content=basu&utm_source=mediaadvisory

Cherry-picking studies is a bad idea. “Trusting a selection of papers” –cherry-picking. Much better is admitting the data is in flux, and that studies currently have different results.

Finally, the sheer numbers of the dead in different countries–not relying on estimated rates of asymptomatic or pre-symptomatic infection as necessary to calculate the IFR, just the sheer number of excess deaths, without any extrapolations–are terrifying enough. Look at this New York Times graphic, and scroll to the bottom. Mortality was almost SIX TIMES the usual number in New York in April, which is comparable to war or famine. Bergamo was 6.67. This is comparable to the 6.85 times the usual rate following the earthquake and tsunami in Japan. With a few super-spreaders, one of the greatest disasters in the past 100 years could be coming to a community near you. Why the need to downplay this? Everyone knows there are tradeoffs to be made in responding to this disaster–but that’s what it is, a disaster. Why minimize?

Nicolas Delon
Nicolas Delon
Reply to  Lisa S
3 years ago

The 1.3% figure is for the IFR *among symptomatic cases* (IFR-S), which almost by definition will be much higher than the ‘true’ IFR, since the denominator is probably many times higher than the number of symptomatic cases.

Lisa S.
Lisa S.
Reply to  Nicolas Delon
3 years ago

My mistake. Too-quickly looking for an outlier to show dangers of picking and choosing studies. They later point out that in the Diamond Princess the percentage of asymptomatic infections turned out to be 17.9%, which would bring their estimate down by about 20%. Of course that depends on accepting those numbers. It’s very difficult to find good estimates of the IFR right now. Here’s a preprint showing 1.3 for Europe. O.6 for China, based on age of population. Notably it assigns 50-year olds a .25 IFR, and 60-year olds 0.84.

https://arxiv.org/pdf/2006.02757.pdf

David Wallace
Reply to  Nicolas Delon
3 years ago

By definition it’s higher, but I’m less sure about ‘much higher’. “Asymptomatic” in their paper means “no symptoms”, not “has not tested positive for COVID-19”. They’re doing some fairly sophisticated mathematical modelling to estimate the number of symptomatic cases from the data (how reliable those methods are, I’ve no idea). It remains one of the huge unknowns whether 80% of people with COVID are symptomless, or just 20%.
If you take that 1.3% and reverse-engineer the number of NYC-area COVID cases from the 40,000 deaths, you get about 3 million cases. That’s not far off the 4 million cases that random antibody testing seems to imply, even before allowing for the undercounting of fatalities in the NYC data, which would suggest a relatively small adjustment due to asymptomatic cases. (Note that the fatality rate in NYC estimated from confirmed cases would be around 10%, which is obviously much higher than anyone really predicts.)
Of course, the 4 million estimate for NYC could also be wrong. I repeat my previous call for epistemic humility on these numbers.

Nicolas Delon
Nicolas Delon
Reply to  David Wallace
3 years ago

Fair. It all depends what we mean by “much higher”, and what credence we have in the claims that seroprevalence is (much or somewhat) higher than we think. I was not making a specific claim about the overall denominator—like you, I call for epistemic humility.

Nicolas Delon
Nicolas Delon
Reply to  Nicolas Delon
3 years ago

There’s also huge variance in the asymptomatic rate. Relying on the Diamond Princess case tells you something but only so much. Again, we don’t really know anything.

Nicolas Delon
Nicolas Delon
Reply to  Nicolas Delon
3 years ago
Irina
Irina
Reply to  Lisa S
3 years ago

Yes, sorry, Anatoli, in my haste I linked to an analysis of the wrong pre-print from the Ioannidis lab. Many thanks to Lisa for the correction.

Anatoli P
Anatoli P
Reply to  Lisa S
3 years ago

I think the analysis is very simple and we do not need to go into papers. Let me take a group of people 45-54 years old, not too young not too old. According to CDC data the USA there are total 4588 people who died from COVID in this group. According to the census data for 2018: https://www.statista.com/statistics/241488/population-of-the-us-by-sex-and-age/ there are total 41 630 000 people in this age group. Now the only assumption I am making based on various sources that about 10 times more people are infected than actually reported, which makes the number of infected around 20 000 000, which is roughly 6% of the population. This factor of 10 was also found in Italy and is consistent with various testings done, probably the real number is even higher. Anyway let me use 6%. So the total number of infected people in this age group is roughly 2,500,000. Now we can get the death ratio 4588/2,500,000. We are getting roughly 0.2%, i.e. 1 in 500. But now going back to preconditions. For people without them the risk is 50 times lower (plus minus uncertainty), which brings it to 1 in 25,000. This is certainly very low on everyday scale. When we take students who are much younger the risks are much much lower according to the same data. Preconditions could be tricky, I do not know exactly which percentile of population have them. Let us say 25%, then the rest can easily get the virus with small risk, get immunity and most importantly slow down transmitting the virus much more efficiently than masks.

I raised many time side effects of prolonged lockdown. No one seems to be concerned. We will get soaring unemployment, lost health insurances and much higher health risks associated with these. Also we are sacrificing our kids offering them Zoom circus instead of schools. But if all these measures would save more lives I would agree, but I do not see how. The virus can not go away now with 50 million people infected, so we are simply postponing the problem creating huge side effects. In either case I did not see any serious analysis of these factors, which are on the other side of prolonged lockdown. I wonder how many people think the virus is gone by September, December, April, next June, June of 2025. Also sooner or later people will be fed up with this anyway. We got the virus and we have to live with it, not simply exist but live!

David Wallace
Reply to  Anatoli P
3 years ago

The chance that a randomly-selected person has (e.g.) undiagnosed cardiovascular disease is way higher than 1/25,000.

Anatoli P
Anatoli P
Reply to  David Wallace
3 years ago

But you were just teaching me probability. The chances of this random person of dying are still much less than 100%. There is no escape, either number of chronic conditions is small or even with conditions the chances of dying are not that high. Besides obviously people with more severe conditions are at higher risk and they are more likely to know about them.

David Wallace
Reply to  Anatoli P
3 years ago

But you are the one who keeps relying on drastically reduced risk estimates based on not having preexisting conditions. If you’re content to set that aside and use your 1/500 number rather than 1/25,000, fine.

I’d like to repeat, yet again, that these discussions of mortality rates don’t engage with morbidity. If (purely for illustration) among the 40-65 range there are 10 cases of long-term health damage for every death, that’s a 1/50 risk of such damage. That sounds worth taking seriously. Perhaps it’s higher than that. Perhaps it’s lower. We don’t know.

Anatoli Polkovnikov
Anatoli Polkovnikov
Reply to  David Wallace
3 years ago

I am not really trying to ignore health risks. I just think that scare of population is greatly exaggerated, if you do not like factor of 50 choose a smaller one. There are many healthy individuals even young people who are overly scared. The issue which no one discusses is that there are very serious consequences. Loosing health care benefits b y many with soaring unemployment will certainly increase death and serious illnesses by a sizable amount. We simply cannot focus on Covid and forget about other factors. If you extrapolate even current trends it will take years and years of continuous lockdown to eliminate it. And exponential extrapolations are certainly wrong as noise (e.g. people coming in and out of the country) will kill any exponents. We are in a bad situation and we need to face it. But if we as a society do not estimate risks correctly we definitely create more damage to us and to future generations.

My own expectation that the cases will start raising again, schools will close, in the fall when flu season returns, by that time people will be exhausted from the lockdown, free money will be over and states with stricter lockdowns will be in much worse position. We will have very high unemployment rate unsustainable for Social Security, businesses will slowly go to more open states, which will do well or at least much better. This will create further strain on Mass. budget. There will be no Covid surges as they were mostly due to nursing homes and older people who learned to protect themselves, but there will be steady increases because we are far from her immunity even among healthy people. No vaccine will be developed, or if there is something, it will be not very inefficient with unknown long term effects. I hope I am wrong, but I simply do not see any optimistic scenario and hence no benefits from lockdown.

Lisa S
Lisa S
3 years ago
Anatoli P
Anatoli P
Reply to  Lisa S
3 years ago

We (by we we mean all of us, society politicians, mass media…) should be ashamed of this graph. Instead of trying to protect elderly or seriously ill people, who needed this most we started to protect ourselves basically covering our buts. 60% of deaths in Massachusetts were in nursing homes. If we not have any lockdown but simply put resources in shielding those, we would get fewer deaths. Maybe there will come a moment of truth when we (again as a society) admit how selfish we were!

Irina
Irina
3 years ago

“That letter was explaining why someone else should pay high salaries of faculty without offering any sacrifice in exchange. If they would say that they are willing to take 50% pay cut for not taking the risk, or do other things in addition, it would possibly be a starter. Otherwise it is not.”

Consider what you are saying. First, you are asking people to take a pay cut without a corresponding reduction in their personal labor. It’s best not to conflate the amount of work that goes in to teaching remotely with the pedagogical value of remote learning. (Setting aside the question of whether face-to-face instruction amid a pandemic is indeed superior to temporary remote instruction, which I’m not sure is as clear cut as you suggest.) Second, are you implying that, rather than it being the employer’s responsibility to protect its employees from harm, that employees must PAY their employer for the privilege of receiving basic protections to their health and respect for their dignity? By this logic, everyone, including those in the most vulnerable categories, should be required to pay their employer back for the right to avoid exposure to a deadly virus. Think of what this would mean for your colleagues in their 70s and 80s, those undergoing (e.g.) chemotherapy, or who live with medically fragile family members.

Your argument might go through if we were to accept with high confidence that severe outcomes from COVID were extremely rare among those infected and if we didn’t consider how patterns of human behavior affect the spread of the virus. In such a case, working from home might be viewed as a luxury and a reduction in pay might be justified. As others have explained, however, we are not entitled to this assumption given the current state of the evidence.

Anatoli P
Anatoli P
Reply to  Irina
3 years ago

Sorry my argument is simpler. If even 50% of student will defer (what I heard the actual number will be higher), BU budget will loose $500 million dollars (may estimate). This means higher than 50% cut. BU is not printing money and students will not pay, like I suspect you will not pay for a Zoom plumber. Many students got these money through loans, many through savings of parents who work hard. Why on the earth they should pay as much for Zoom. It does not matter how hard you work, it matters how much you produce and you produce much less online. Asking students to sacrifice their experience and their money to pay for faculty convenience (justified or not) is immoral and simply will not work as students won’t pay, whatever we think. With this gap BU will be forced to simply furlough faculty and staff even those who are willing to work. Hiding behind backs of faculty who will teach and take the risk is again deeply immoral. There is no solution, I am afraid, but to come and work, find a substitute, or take an unpaid leave of absence. You can not charge for a product nobody wants…

DoubleA
DoubleA
3 years ago

What happens when so many faculty opt to teach remotely that not enough are left to teach the students who want (and paid for) in person instruction?