Super-spreading events are gatherings at which one or more attendees are already harboring an infection and manage to transmit it to a number of others. These people, in turn, spread it to their close contacts, possibly at the same event. Super-spreading has dominated the transmission of COVID-19. These transmissions have almost always taken place indoors in spaces with limited ventilation, and they have usually involved close or prolonged contact. In addition, super-spreading originates with a small subset of infected individuals. That’s essentially what the chart above shows. It ranks individual subjects by their exhaled quantity of aerosolized particles per liter of air.
For more than a year, we’ve also known that obesity and age are associated with more severe COVID infections. Now, it’s startling to learn that obese and/or older, infected individuals are more prone to transmitting virus: this study found that a high body mass index (BMI) is associated with significantly greater quantities of exhaled aerosol, and that age has a similarly strong association. So called BMI-years, or age x BMI, has an extremely powerful association with the exhalation of aerosol-borne particles. The authors, David A. Edwards, et al, believe this is a consequence of the properties of mucus produced by different individuals in response to infections and how their lungs and airways handle it. The authors say:
“Our findings indicate that the capacity of airway lining mucus to resist breakup on breathing varies significantly between individuals, with a trend to increasing with the advance of COVID-19 infection and body mass index multiplied by age (i.e., BMI-years). Understanding the source and variance of respiratory droplet generation, and controlling it via the stabilization of airway lining mucus surfaces, may lead to effective approaches to reducing COVID-19 infection and transmission. … ”
“Surfactant and mucin compositional and structural changes, driven, in part, by physiological alterations of the human condition—including diet (10), aging (11), and COVID-19 infection itself (12)—may therefore be anticipated to alter droplet generation and droplet size (7) during acts of breathing.”
So there is substantial variation in the exhalation of aerosol-borne particles across individuals. In the study, less than 20% of healthy subjects produced more than 156 particles per liter of air, accounting for 80% of the exhaled particles. This defined their so-called “super-spreader” cohort. The association of BMI-years and exhaled particles was less pronounced but still positive within the “low-spreader” cohort.
Edwards, et al speculate that these fine droplets might help explain the greater severity of COVID infections among the elderly and obese. Not only does the breakup of mucus into tiny droplets cause these individuals to exhale aerosols more profusely, it probably also leads to deep penetration into their lung tissue.
This knowledge might be broadly applicable to infectious diseases, and SARS viruses in particular. The elderly know they are vulnerable. It’s not clear that the obese have viewed themselves as vulnerable, but they should, even in the age of “body positivity“. And not only are they vulnerable: they appear to pose an elevated hazard to others. I came across a couple of sardonic comments that got right to the apparent elephant in the room: “Instead of a mask mandate, how about a push-up mandate?”; and “Instead of a vaccine passport, how about a BMI passport?”
The debate about how to care for the most vulnerable is ongoing, but the mere mention of regularities like those identified by the study might lead to proposals for coercive policies. But first, a few practical points to bear in mind: 1) while the study identifies a major risk factor for transmission, it must be replicated by others, and there must be research into the underlying reasons for the phenomenon; 2) while the obese and seniors may be more likely to super-spread, not all of them are super-spreaders; and 3) as a matter of policy, how would “super-spreaders” be defined? What would be the cutoff BMIs at various ages? No matter what was decided, restrictive policies predicated on mere statistical associations would involve gross injustices to a large number of individuals.
With the degree of acquired immunity already in the population and fairly widespread voluntary vaccination (since alarmists have scared the bejeezus out of everyone), the whole issue might seem moot. It’s not, however, because COVID-19 is likely to become endemic, the immunities of some individuals might erode more quickly than expected, new and more dangerous variants might arise, and new SARS viruses are likely to emerge with time.
In a pandemic, however, and even without knowing who is infected, it is ethically barbaric to probabilistically isolate classes of individuals, whether based on age, BMI, or anything other than contagious status. The social cost is simply unacceptable. Instead, public health authorities should provide information to those at high risk, facilitate vaccination for those who desire it, and promote rapid, at-home tests. This is essentially a deregulatory agenda relative to the mindless lockdown approaches favored by so many public health experts.
Everyone must balance their own personal risks and rewards. Based on the study of exhaled particles discussed above, some might shun the obese and seniors until the threat has passed. Some of the obese and elderly might shun each other. That might be another regrettable dimension of the costs of a pandemic. On the other hand, perhaps more of us will respond to the unquestionably positive incentives for weight loss, of which we’re almost all aware.
The CDC’s new study on dining out and mask mandates is a sham. On its face, the effects reported are small. And while it’s true most of the reported effects are statistically significant, the CDC acknowledges a number of factors that might well have confounded the results. This study should remind us of the infinite number of spurious and “significant” correlations in the world. Here, the timing of the mandates (or their removal) relative to purported effects and seasonal waves is highly suspicious, and as always, attributing causality on the basis of correlation is problematic.
On one hand, the CDC’s results are contrary to plentiful evidence that mandates are ineffective; on the other hand, the results are contrary to earlier CDC “guidance” that masks and limits on indoor dining are “highly effective”. Nevertheless, the latest report has massive propaganda value to the CDC. The media lapped up the story and provided cover for Democrats eager to pass the COVID (C19) relief package. Likewise, the Biden Administration is apparently committed to the narrative of an ongoing crisis as cover for continued attempts to shame political opponents in states that have elected to “reopen” or remain open.
Right off the bat, the study’s authors assert that the primary mode of transmission of C19 is from respiratory droplets. This is false. We know that aerosols are the main culprit in transmission, against which cloth masks are largely ineffective.
Be that as it may, let’s first consider the findings on dining. There was no statistically significant effect on the growth rate of cases or deaths up to 40 days after restrictions were lifted, according to the report. In fact, case growth declined slightly. There was, however, a small but statistically significant increase after 40 days. The fact that deaths seemed to “respond” faster and with greater magnitude than cases makes no sense and suggests that the results might be spurious.
The CDC offers possible explanations the long delay in the purported impact, such as the time required by restaurants to resume operations and early caution on the part of diners. These are speculative, of course. More pertinent is the fact that the data did not distinguish between indoor and outdoor dining, nor did it account for other differences in regulation such as rules on physical distancing, intra-county variation in local government mandates, and compliance levels.
Finally, the measurement of effects covered 100 days after the policy change, but this window spans different stages of the pandemic. There were three waves of infections during 2020, which correspond to the classic Hope-Simpson pattern of virus seasonality. One was near year-end, but as each of the first two waves tapered (April-May, August-September), it should be no surprise that many restrictions were lifted. Within two months, however, new waves had begun. Karl Dierenbach notes that most of the reopenings occurred in May. Here’s how he explains the pattern:
“The map on the left shows counties where there was no on-premises dining (pink) in restaurants as of the beginning of May (4/30). … The map on the right shows that by the end of May, almost the entire country moved to allow some on-premises dining (green).”
“In the 100 days after May 1, cases nationwide fell slightly, then began to rise, and then plateaued.”
“And what did the CDC find happened after restaurants were allowed (changing mostly in May) to have on-premises dining? … Surprise! The CDC found that cases fell slightly, then began to rise, and then plateaued.”
The summer “mini-wave” is typical of mid- and tropical-latitude seasonality. Thus, the CDC’s findings with respect to dining restrictions are likely an artifact of the strong seasonality of the virus, rather than having anything to do with the lifting of restrictions between waves.
What about the imposition of mask mandates? The CDC’s findings show a much faster response in this case, with statistically significant changes in growth during the first 20 days. Another indicator of spurious correlation is that the growth response of deaths did not lag that of cases, but in fact deaths have reliably lagged cases by over 18 days during the pandemic. Again, the CDC’s caveats apply equally to its findings on masks. A large share of individuals adopted mask use voluntarily before mandates were imposed, so it’s not even clear that the mandates contributed much to the practice.
It’s a stretch to believe that mask mandates would have had an immediate, incremental effect on the growth of cases and deaths, given probable lags in compliance, exposure, and onset of symptoms. Moreover, a number of mask mandates in 2020 were imposed near the very peak of the seasonal waves. Little wonder that the growth rates of cases and deaths declined shortly thereafter.
We’ve known for a long time that masks do little to stop the spread of viral particles. They become airborne as aerosols which easily penetrate the kind of cloth masks worn by most members of the public, to say nothing of making contact with their eyes. The table below contains citations to research over the past 10 years uniformly rejecting the hypothesis of a significant protective effect against influenza from masks. There is no reason to believe that they would be more effective in preventing C19 infections.
The CDC’s report on dining restrictions and mask mandates is a weak analysis. They wish to emphasize their faith in non-pharmaceutical interventions (NPIs) to minimize risks. They do so at a time when the vaccinated share of the most vulnerable population, the elderly, has climbed above 50% and is increasing steadily. Thus, risks are falling dramatically, so it’s past time to weigh the costs and benefits of NPIs more realistically. The timing of the report also seemed suspicious, coming as it did in the heat of the battle over the $1.9 trillion COVID relief bill, which subsequently passed.
It’s also a good time to note that zero risk, including “Zero COVID”, is not a realistic or worthwhile goal under any reasonable comparison of costs and benefits. Furthermore, NPIs have proven weak generally (also see here); claims to the contrary should always make us wary.
The CDC choked on a new analysis estimating COVID-19’s impact on U.S. life expectancy as of year-end 2020: they reported a decline of a full year, which is ridiculous on its face! As explained by Peter B. Bach in STAT News, the agency assumed that excess deaths attributed to COVID in 2020 would continue as a permanentaddition to deaths going forward. Please forgive my skepticism, but isn’t this too basic to qualify as an analytical error by an agency that subjects its reports to thorough vetting? Or might this have been a deliberate manipulation intended to convince the public that COVID will be an ongoing public health crisis. Of course the media has picked it up; even Zero Hedgereported it uncritically!
Bach does a quick calculation based on 400,000 excess deaths attributed to COVID in 2020 and 12 life-years lost by the average victim. I believe the first assumption is on the high side, and I say “attributed to COVID” as a reminder that the CDC’s guidance for completing death certificates was altered in the spring of 2020 specifically for COVID and not other causes of death. Furthermore, if our objective is to assess the impact of the virus itself, under no circumstances should excess deaths induced by misguided lockdown policies enter the calculation (though Bach entertains the possibility). Bach arrives at a reduction in average life of 5.3 days! Of course, that’s not intended to be a projection, but it is a reasonable estimate of COVID’s impact on average lives in 2020.
The CDC’s projection essentially freezes death rates at each age at their 2020 values. We will certainly see more COVID deaths in 2021, and the virus is likely to become endemic. Even with higher levels of acquired immunity and widespread vaccinations, there will almost certainly be some ongoing deaths attributable to COVID, but they are likely to be at levels that will blend into a resumption of the long decline in mortality rates, especially if COVID continues to displace the flu in its “ecological niche”. I include the chart at the top to emphasize the long-term improvement in mortality (though the chart shows only a partial year for 2020, and there has been some flattening or slight backsliding over the past five years or so). As Bach says:
“Researchers have regularly demonstrated that life expectancy projections are overly sensitive to evanescent events like pandemics and wars, resulting in considerably overestimated declines. … And yet the CDC published a result that, if anything, would convey to the public an exaggerated toll that Covid-19 took on longevity in 2020. That’s a problem.”
There were excess deaths from other causes in 2020, which Bach acknowledges. Perhaps 100,000 or more could be attributed to lockdowns and their consequences like economically-induced stress, depression, suicide, overdoses, and medical care deferred or never sought. The Zero Hedge article mentioned above discusses findings that lockdowns and their consequences, such as unemployment spells and lost education, will have ongoing negative effects on health and mortality for many years. The net effect on life expectancy might be as large as 11 to 12 days. Again, however, I draw a distinction between deaths caused by the disease and deaths caused by policy mistakes.
The CDC’s estimate should not be taken seriously when, as Kyle Smith says, there is every indication that the battle against COVID is coming to a successful conclusion. Public health experts have not acquitted themselves well during the pandemic, and the CDC’s life expectancy number only reinforces that impression. Here is Smith:
“We have learned a lot about how the virus works, and how it doesn’t: Outdoor transmission, for the most part, hardly ever happens. Kids are at very low risk, especially younger children. Baseball games, barbecues, and summer camps should be fine. Some pre-COVID activities now carry a different risk profile — notably anything that packs crowds together indoors, so Broadway theater, rock concerts, and the like will be just about the last category of activity to return to normal.”
But return to normal we should, and yet the CDC seems determined to poop on the victory party!
It’s been said that many of the so-called “heroes” of the COVID pandemic who’ve been celebrated by the media are actually villains, and perhaps Governor Andrew Cuomo of New York should top the list. He saw to it that retirement homes were seeded with infected patients by ordering them returned their care homes rather than admitted to hospitals. Deaths in these facilities mounted, and they mounted faster than Cuomo’s administration was willing to admit. But the media and even Democrat state legislators have begun to take note, which is practically a miracle!
It seems equally true that some vilified by the media for their COVID response are actually heroes. Governor Ron DeSantis of Florida might deserve top honors here. Having spent the last month in Florida, I can attest that the business and social environment here is quite open compared to my home state (despite the presence of a few freaked out northerners who can’t quite fathom how stupid they look wearing masks on the beach). Florida’s infections, hospitalizations, and deaths have been lower than in California, New York, and many other states where lockdown measures have been stringent. (The first chart below is just a little busy…)
This approach to saving lives is obvious, yet critics at outlets like NBC News insist that DeSantis must be pandering to the senior population in Florida. Well, one wouldn’t want to be responsive to voters who happen to face high mortality risks, right? Others such as horror writer Stephen Kinghave jumped onboard to offer their bumbling public health expertise as well.
There were many experts and the usual collection of numbskulls on social media who were wrong about Florida. DeSantis handled the pandemic as it should have been handled elsewhere. But the propaganda to the contrary goes unabated. For example, this article is pathetic. Can these people be serious? Or are they really that stupid? This goes for the Biden Administration as well, which had entertained the notion of imposing federal travel restrictions on Florida!
The political attacks on Florida and its governor reveal the extent to which opponents wish to ignore the evidence in plain sight. The data on COVID outcomes put the lie to the narrative of a public health emergency requiring massive restrictions on personal liberty. We know those policies are powerless to control the course of the contagion. The pandemic, however, was the key to convincing the public to accept a more authoritarian role for government. It’s a blessing that not everyone bought in, and that there are places like Florida where you can still go about your business in approximate normalcy.
Anthony Fauci has repeatedly increased his estimate of how much of the population must be vaccinated to achieve what he calls herd immunity, and he did it again in late December. This series of changes, and other mixed messages he’s delivered in the past, reveal Fauci to be a “public servant” who feels no obligation to level with the public. Instead, he crafts messages based on what he believes the public will accept, or on his sense of how the public must be manipulated. For example, by his own admission, his estimates of herd immunity have been sensitive to polling data! He reasoned that if more people reported a willingness to take a vaccine, he’d have flexibility to increase his “public” estimate of the percentage that must be vaccinated for herd immunity. Even worse, Fauci appears to lack a solid understanding of the very concept of herd immunity.
There is so much wrong with his reasoning on this point that it’s hard to know where to start. In the first place, why in the world would anyone think that if more people willingly vaccinate it would imply that even more must vaccinate? And if he felt that way all along it demonstrates an earlier willingness to be dishonest with the public. Of course, there was nothing scientific about it: the series of estimates was purely manipulative. It’s almost painful to consider the sort of public servant who’d engage in such mental machinations.
Immunity Is Multi-Faceted
Second, Fauci seemingly wants to convince us that herd immunity is solely dependent on vaccination. Far from it, and I’m sure he knows that, so perhaps this too was manipulative. Fauci intimates that COVID herd immunity must look something like herd immunity to the measles, which is laughable. Measles is a viral infection primarily in children, among whom there is little if any pre-immunity. The measles vaccine (MMR) is administered to young children along with occasional boosters for some individuals. Believe it or not, Fauci claims that he rationalized a requirement of 85% vaccination for COVID by discounting a 90% requirement for the measles! Really???
In fact, there is substantial acquired pre-immunity to COVID. A meaningful share of the population has long-memory, cross-reactive T-cells from earlier exposure to coronaviruses such as the common cold. Estimates range from 10% to as much as 50%. So if we stick with Fauci’s 85% herd immunity “guesstimate”, 25% pre-immunity implies that vaccinating only 60% of the population would get us to Fauci’s herd immunity goal. (Two qualifications: 1) the vaccines aren’t 100% effective, so it would take more than 60% vaccinated to offset the failure rate; 2) the pre-immune might not be identifiable at low cost, so there might be significant overlap between the pre-immune and those vaccinated.)
Vaccinations approaching 85% would be an extremely ambitious goal, especially if it is recommended annually or semi-annually. It would be virtually impossible without coercion. While more than 91% of children are vaccinated for measles in the U.S., it is not annual. Thus, measles does not offer an appropriate model for thinking about herd immunity to COVID. Less than half of adults get a flu shot each year, and somewhat more children.
Fauci’s reference to 85% – 90% total immunity is different from the concept of the herd immunity threshold (HIT) in standard epidemiological models. The HIT, often placed in the range of 60% – 70%, is the point at which new infections begin to decline. More infections occur above the HIT but at a diminishing rate. In the end, the total share of individuals who become immune due to exposure, pre-immunity or vaccination will be greater than the HIT. The point is, however, that reaching the HIT is a sufficient condition for cases to taper and an end to a contagion. If we use 65% as the HIT and pre-immunity of 25%, only 40% must be vaccinated to reach the HIT.
A recent innovation in epidemiological models is the recognition that there are tremendous differences between individuals in terms of transmissibility, pre-immunity, and other factors that influence the spread of a particular virus, including social and business arrangements. This kind of heterogeneity tends to reduce the effective HIT. We’ve already discussed the effect of pre-immunity. Suppose that certain individuals are much more likely to transmit the virus than others, like so-called super-spreaders. They spur the initial exponential growth of a contagion, but there are only so many of them. Once infected, no one else among the still-susceptible can spread the virus with the same force.
Researchers at the Max Planck Institute (and a number of others) have gauged the effect of introducing heterogeneity to standard epidemiological models. It is dramatic, as the following chart shows. The curves simulate a pandemic under different assumptions about the degree of heterogeneity. The peak of these curves correspond to the HIT under each assumption (R0 refers to the initial reproduction number from infected individuals to others).
Moderate heterogeneity implies a HIT of only 37%. Given pre-immunity of 25%, only an additional 12% of the population would have to be infected or vaccinated to prevent a contagion from gaining a foothold for the initial exponential stage of growth. Fauci’s herd immunity figure obviously fails to consider the effect of heterogeneity.
How Close To the HIT?
We’re not as far from HITs as Fauci might think, and a vaccination goal of 85% is absurd and unnecessary. The seasonal COVID waves we’ve experienced thus far have faded over a period of 10-12 weeks. Estimates of seroprevalence in many localities reached a range of 15% – 25% after those episodes, which probably includes some share of those with pre-immunity. To reach the likely range of a HIT, either some additional pre-immunity must have existed or the degree of heterogeneity must have been large in these populations.
But if that’s true, why did secondary waves occur in the fall? There are a few possibilities. Of course, some areas like the upper Midwest did not experience the springtime wave. But in areas that suffered a recurrance, perhaps the antibodies acquired from infections did not remain active for as long as six months. However, other immune cells have longer memories, and re-infections have been fairly rare. Another possibility is that those having some level of pre-immunity were still able to pass live virus along to new hosts. But this vector of transmission would probably have been quite limited. Pre-immunity almost surely varies from region to region, so some areas were not as firmly above their HITs as others. It’s also possible that infections from super-spreaders were concentrated within subsets of the population even within a given region, in certain neighborhoods or among some, but not all, social or business circles. Therefore, some subsets or “sub-herds” achieved a HIT in the first wave, but it was unnecessary for other groups. In other words, sub-herds spared in the first wave might have suffered a contagion in a subsequent wave. And again, reinfections seem to have been rare. Finally, there is the possibility of a reset in the HIT in the presence of a new, more transmissible variant of the virus, as has become prevalent in the UK, but that was not the case in the fall.
Tyler Cowen has mentioned another possible explanation: so-called “fragile” herd immunity. The idea is that any particular HIT is dependent on the structure of social relations. When social distancing is widely practiced, for example, the HIT will be lower. But if, after a contagion recedes, social distancing is relaxed, it’s possible that the HIT will take a higher value at the onset of the next seasonal wave. Perhaps this played a role in the resurgence in infections in the fall, but the HIT can be reduced via voluntary distancing. Eventually, acquired immunity and vaccinations will achieve a HIT under which distancing should be unnecessary, and heterogeneity suggests that shouldn’t be far out of reach.
Anthony Fauci has too often changed his public pronouncements on critical issues related to management of the COVID pandemic. Last February he said cruises were fine for the healthy and that most people should live their lives normally. Oops! Then came his opinion on the limited effectiveness of masks, then a shift to their necessity. His first position on masks has been called a “noble lie” intended to preserve supplies for health care workers. However, Fauci was probably repeating the standing consensus at that point (and still the truth) that masks are of limited value in containing airborne pathogens.
This time, Fauci admitted to changing his estimate of “herd immunity” in response to public opinion, a pathetic approach to matters of public health. What he called herd immunity was really an opinion about adequate levels of vaccination. Furthermore, he neglected to consider other forms of immunity: pre-existing and already acquired. He did not distinguish between total immunity and the herd immunity threshold that should guide any discussion of pandemic management. He also neglected the significant advances in epidemiological modeling that recognize the reality of heterogeneity in reducing the herd immunity threshold. The upshot is that far fewer vaccinations are needed to contain future waves of the pandemic than Fauci suggests.
There are currently two vaccines in limited distribution across the U.S. from Pfizer and Moderna, but the number and variety of different vaccines will grow as we move through the winter. For now, the vaccine is in short supply, but that’s even more a matter of administering doses in a timely way as it is the quantity on hand. There are competing theories about how best to allocate the available doses, which is the subject of this post. I won’t debate the merits of refusing to take a vaccine except to say that I support anyone’s right to refuse it without coercion by public authorities. I also note that certain forms of discrimination on that basis are not necessarily unreasonable.
The vaccines in play all seem to be highly effective (> 90%, which is incredible by existing standards). There have been a few reports of side effects — certainly not in large numbers — but it remains to be seen whether the vaccines will have any long-term side effects. I’m optimistic, but I won’t dismiss the possibility.
Despite competing doctrines about how the available supplies of vaccine should be allocated, there is widespread acceptance that health care workers should go first. I have some reservations about this because, like Emma Woodhouse, I believe staff and residents at long-term care facilities should have at least equal priority. Yet they do not in the City of Chicago and probably in other areas. I have to wonder whether unionized health care workers there are the beneficiaries of political favoritism.
Beyond that question, we have the following competing priorities: 1) the vulnerable in care homes and other elderly individuals (75+, while younger individuals with co-morbidities come later); 2) “essential” workers of all ages (from police to grocery store clerks — decidedly arbitrary); and 3) basically the same as #2 with priority given to groups who have suffered historical inequities.
#1 is clearly the way to save the most lives, at least in the short-run. Over 40% of the deaths in the U.S. have been in elder-care settings, and COVID infection fatality ratesmount exponentially with age:
To derive the implications of #1 and #2, it’s more convenient to look at the share of deaths within each age cohort, since it incorporates the differences in infection rates and fatality rates across age groups (the number of “other” deaths is much larger than COVID deaths, of course, despite similar death shares):
The 75+ age group has accounted for about 58% of all COVID deaths in the U.S., and ages 25 – 64 accounted for about 20% (an approximate age range for essential workers). This implies that nearly three times as many lives can be saved by prioritizing the elderly, at least if deaths among so-called essential workers mimic deaths in the 25 – 64 age cohorts. However, the gap would be smaller and perhaps reversed in terms of life-years saved.
Furthermore, this is a short-run calculation. Over a longer time frame, if essential workers are responsible for more transmission across all ages than the elderly, then it might throw the advantage to prioritizing essential workers over the elderly, but it would take a number of transmission cycles for the differential to play out. Yes, essential workers are more likely to be “super-spreaders” than work-at-home, corporate employees, or even the unemployed, but identifying true super-spreaders would require considerable luck. Moreover, care homes generally house a substantial number of elderly individuals and staff in a confined environment, where spread is likely to be rampant. So the transmission argument for #2 over #1 is questionable.
The over-riding problem is that of available supply. Suppose enough vaccine is available for all elderly individuals within a particular time frame. That’s about 6.6% of the total U.S. population. The same supply would cover only about 13% of the younger age group identified above. Essential workers are a subset of that group, but the same supply would fall far short of vaccinating all of them; lives saved under #2 would then fall far short of the lives saved under #1. Quantities of the vaccine are likely to increase over the course of a few months, but limited supplies at the outset force us to focus the allocation decision on the short-term, making #1 the clear winner.
Now let’s talk about #3, minority populations, historical inequities, and the logic of allocating vaccine on that basis. Minority populations have suffered disproportionately from COVID, so this is really a matter of objective risk, not historical inequities… unless the idea is to treat vaccine allocations as a form of reparation. Don’t laugh — that might not be far from the intent, and it won’t count as a credit toward the next demand for “justice”.
For the sake of argument, let’s assume that minorities have 3x the fatality rate of whites from COVID (a little high). Roughly 40% of the U.S. population is non-white or Hispanic. That’s more than six times the size of the full 75+ population. If all of the available doses were delivered to essential workers in that group, it would cover less than half of them and save perhaps 30% of minority COVID deaths over a few months. In contrast, minorities might account for up to two-thirds of the deaths among the elderly. Therefore, vaccinating all of the elderly would save 58% of elderly COVID deaths and about 39% of minority deaths overall!
The COVID mortality risk to the average white individual in the elderly population is far greater than that faced by the average minority individual in the working age population. Therefore, no part of #3 is sensible from a purely mathematical perspective. Race/ethnicity overlaps significantly with various co-morbiditiesand the number of co-morbidities with which individuals are afflicted. Further analysis might reveal whether there is more to be gained by prioritizing by co-morbidities rather than race/ethnicity.
Megan McArdle has an interesting column on the CDC’s vaccination guidelines issued in November, which emphasized equity, like #3 above. But the CDC walked back that decision in December. The initial November decision was merely the latest of the the agency’s fumbles on COVID policy. In her column, McArdle notes that the public has understood that the priority was to save lives since the very start of the pandemic. Ideally, if objective measures show that identifiable characteristics are associated with greater vulnerability, then those should be considered in prioritizing individuals who desire vaccinations. This includes age, co-morbidities, race/ethnicity, and elements of occupational risk. But lesser associations with risk should not take precedence over greater associations with risk unless an advantage can be demonstrated in terms of lives saved, historical inequities or otherwise.
The priorities for the early rounds of vaccinationsmay differ by state or jurisdiction, but they are all heavily influenced by the CDC’s guidelines. Some states pay lip service to equity considerations (if they simply said race/ethnicity, they’d be forced to operationalize it), while others might actually prioritize doses by race/ethnicity to some degree. Once the initial phase of vaccinations is complete, there are likely to be more granular prioritizations based on different co-morbidities, for example, as well as race/ethnicity. Thankfully, the most severe risk gradient, advanced age, will have been addressed by then.
One last point: the Pfizer and Moderna vaccines both require two doses. Alex Tabarrok points out that first doses appear to be highly effective on their own. In his opinion, while supplies are short, the second dose should be delayed until all groups at substantially elevated risk can be vaccinated…. doubling the supply of initial doses! The idea has merit, but it is unlikely to receive much consideration in the U.S. except to the extent that supply chain problems make it unavoidable, and they might.
For clarity, start with this charming interpretive one-act on public health policy in 2020. You might find it a little sardonic, but that’s the point. It was one of the more entertaining tweets of the day, from @boriquagato.
A growing body of research shows that stringent non-pharmaceutical interventions (NPIs) — “lockdowns” is an often-used shorthand — are not effective in stemming the transmission and spread of COVID-19. A compendium of articles and preprints on the topic was just published by the American Institute for Economic Research (AEIR): “Lockdowns Do Not Control the Coronavirus: The Evidence”. The list was compiled originally by Ivor Cummins, and he has added a few more articles and other relevant materials to the list. The links span research on lockdowns across the globe. It covers transmission, mortality, and other health outcomes, as well as the economic effects of lockdowns. AIER states the following:
“Perhaps this is a shocking revelation, given that universal social and economic controls are becoming the new orthodoxy. In a saner world, the burden of proof really should belong to the lockdowners, since it is they who overthrew 100 years of public-health wisdom and replaced it with an untested, top-down imposition on freedom and human rights. They never accepted that burden. They took it as axiomatic that a virus could be intimidated and frightened by credentials, edicts, speeches, and masked gendarmes.
The pro-lockdown evidence is shockingly thin, and based largely on comparing real-world outcomes against dire computer-generated forecasts derived from empirically untested models, and then merely positing that stringencies and “nonpharmaceutical interventions” account for the difference between the fictionalized vs. the real outcome. The anti-lockdown studies, on the other hand, are evidence-based, robust, and thorough, grappling with the data we have (with all its flaws) and looking at the results in light of controls on the population.”
We are constantly told that public intervention constitutes “leadership”, as if our well being depends upon behavioral control by the state. Unfortunately, it’s all too typical of research on phenomena deemed ripe for intervention that computer models are employed to “prove” the case. A common practice is to calibrate such models so that the outputs mimic certain historical outcomes. Unfortunately, a wide range of model specifications can be compatible with an historical record. This practice is also a far cry from empirically testing well-defined hypotheses against alternatives. And it is a practice that usually does poorly when the model is tested outside the period to which it is calibrated. Yet that is the kind of evidence that proponents of intervention are fond of using to support their policy prescriptions.
Here’s some incredible data on PCR tests demonstrating a radically excessive lab practice that generates false positives. I’m almost tempted to say we’d do just as well using a thermometer and the coffee ground test. Open a coffee tin and take a sniff. Can you smell the distinct aroma of the grounds? If not, and if you have other common symptoms, there’s a decent chance you have an active COVID infection. That test is actually in use in some parts of the globe!
The data shown below on PCR tests are from the Rhode Island Department of a Health and the Rhode Island State Health Lab. They summarize over 5,000 positive COVID PCR tests (collected via deep nasal swabs) taken from late March through early July. The vertical axis in the chart measures the cycle threshold (Ct) value of each positive test. Ct is the number of times the RNA in a sample must be replicated before any COVID-19 (or COVID-like) RNA is detected. It might be from a live virus or perhaps a fragment of a dead virus. A positive test with a low Ct value indicates that the subject is likely infected with billions of live COVID-19 viruses, while a high Ct value indicates perhaps a handful or no live virus at all.
The range of red dots in the chart (< 28 Ct) indicates relatively low Ct values and active infections. The yellow range of dots, for which 28 < Ct <= 32, indicates possible infections, and the upper range of green dots, where Ct > 32, indicates that active infections were highly unlikely. It’s important to note that all of these tests were recorded as new COVID cases, so the range of Ct values suggest that testing in Rhode Island was unreasonably sensitive. That’s broadly true across the U.S. as well, which means that COVID cases are over-counted by perhaps 30% or more. And yet it is extremely difficult for subjects testing positive to learn their Ct values. You can ask, but you probably won’t get an answer, which is absurd and counterproductive.
Notice that the concentration of red dots diminished over time, and we know that the spring wave of the virus in the Northeast was waning as the summer approached. The share of positives tests with high Ct values increased over that time frame, however. This is borne out by the next chart, which shows the daily mean Ct of these positive tests. The chart shows that active infections became increasingly rare over that time frame both because positive tests decreased and the average Ct value rose. What we don’t know is whether labs bumped up the number of cycles or replications to which samples were subjected. Still, the trend is rather disturbing because most of the positive cases in May and the first half of June were more likely to be virus remnants than live viruses.
It’s also worth noting that COVID deaths declined in concert with the upward trend in Ct values. This is shown in the chart below (where the Ct scale is inverted). This demonstrates the truly benign nature of positive tests having high Ct values.
This is also demonstrated by the following data from a New York City academic hospital, which was posted by Andrew Bostom. It shows that a more favorable “clinical status” of COVID patients is associated with higher Ct values.
It’s astounding that the U.S. has relied so heavily on a diagnostic tool that gets so many subjects wrong. And it’s nearly impossible for subjects testing positive to obtain their Ct values. Instead, they are subject to self-quarantine for up to two weeks. Even worse, until recently there were delays in reporting the results of these tests of up to a week or more. That made them extremely unhelpful. On the other hand, thecoffee ground test is fast and cheap, and it might enhance the credibility of a subsequent positive PCR test, if one is necessary … and especially if the lab won’t report the Ct value.
The PCR test has identified far too many false infections, but it wouldn’t have been quite so damaging if 1) a reasonably low maximum cycle threshold had been established; 2) test results had not been subject to such long delays; and 3) rapid retests had been available for confirmation. The cycle threshold issue is starting to receive more attention, quite belatedly, and more rapid tests have become available. As I’ve emphasized in the past, cheap, rapid tests exist. But having dithered in February and March in approving even the PCR test, the FDA has remained extremely grudging in approving newer tests, and it persists in creating obstacles to their use. The FDA needs to wake up and smell the coffee!
What does it take to shake people out of their statist stupor? Evidently, the sweet “logic” of universal confinement is very appealing to the prescriptive mindset of busybodies everywhere, who anxiously wag their fingers at those whom they view as insufficiently frightened. As difficult as it is for these shrieking, authoritarian curs to fathom, measures like lockdowns, restrictions on business activity, school closures, and mandates on behavior have at besta limited impact on the spread of the coronavirus, and they are enormously costly in terms of economic well-being and many dimensions of public health. Yet the storm of propaganda to the contrary continues. Media outlets routinely run scare stories, dwelling on rising case numbers but ignoring them when they fall; they emphasize inflated measures of pandemic severity; certain researchers and so-called health experts can’t learn the lessons that are plain in the data; and too many public officials feel compelled to assert presumed but unconstitutional powers. At least the World Health Organization has managed to see things clearly, but many don’t want to listen.
I’ll be the first to say I thought the federalist approach to COVID policy was commendable: allow states and local governments to craft policies appropriate to local conditions and political preferences, rather than have the federal government dictate a one-size-fits-all policy. I haven’t wavered in that assessment, but let’s just say I expected more variety. What I failed to appreciate was the extent to which state and local leaders are captive to provincial busybodies, mavens of precautionary excess, and fraudulent claims to scientific wisdom.
Of course, it should be obvious that the “knowledge problem” articulated by Friedrich Hayek is just as dangerous at low-levels of government as it is in a central Leviathan. And it’s not just a knowledge problem, but a political problem: officials become panicked because they fear bad outcomes will spell doom for their careers. Politicians are particularly prone to the hazards of “do-somethingism”, especially if they have willing, status-seeking “experts” to back them up. But as Scott Sumner says:
“When issues strongly impact society, the science no longer ‘speaks for itself’.
Well, the science is not quite as clear as the “follow-the-science” crowd would have you believe. And unfortunately, public officials have little interest in sober assessments of the unintended effects of lockdown policy.
In my last post, I presented a simple framework for thinking about the benefits and costs of lockdown measures, or non-pharmaceutical interventions (NPIs). I also emphasized the knowledge problem: even if there is some point at which NPI stringencies are “optimized”, government does not possess the knowledge to find that point. It lacks detailed information on both the costs and benefits of NPIs, but individual actors know their own tolerance for risk, and they surely have some sense of the risks they pose to others in their normal course of affairs. While voluntary precautions might be imperfect, they accomplish much of what interventionists hope will be gained via coercion. But, in an effort to “sell” NPIs to constituents and assert their authority, officials vastly over-estimate benefits of NPIs and under-estimate the costs.
NPI Stringency and COVID Outcomes
Let’s take a look at a measure of the strength of NPIs by state — the University of Oxford Stringency Index — and compare those to CDC all-cause excess deaths in each state. If it’s hard to read, try clicking on the image or turn your phone sideways. This plot covers outcomes through mid-November:
The chart doesn’t suggest any benefit to the imposition of greater restrictions, or more stringent NPIs. In fact, the truth is that people will do most of the work on their own based on perceptions of risk. That’s partly because government restrictions add little risk mitigation to what can be accomplished by voluntary social distancing and other precautions.
Here’s a similar chart with cross-country comparisons, though the data here ended in early October (I apologize for the fuzzy image):
But what about reverse causality? Maybe the imposition of stringency was a response to more severe contagions. Now that the virus has swept most of the U.S and Europe in three distinct waves, and given the variety and timing of NPIs that have been tried, it’s harder to make that argument. States like South Dakota have done fairly well with low stringency, while states like New Jersey with high stringency have fared poorly. The charts above provide multiple pair-wise examples and counter-examples of states or countries having faced hard waves with different results.
But let’s look at a few specific situations.
The countries shown above have converged somewhat over the past month: Sweden’s daily deaths have risen while the others have declined to greater or lesser degrees, but the implications for mask usage are unaltered.
And of course we have this gem, predicated on the mental gymnastics lockdown enthusiasts are fond of performing:
But seriously, it’s been a typical pattern: cases rise to a point at which officials muster the political will to impose restrictions, often well after the “exponential” phase of the wave or even the peak has passed. For the sake of argument, if we were to stipulate that lockdowns save lives, it would take time for these measures to mitigate new infections, time for some of the infected individuals to become symptomatic, and more time for diagnosis. For the lockdown arguments to be persuasive, the implementation of NPIs would have to precede the point at which the growth of cases begins to decline by a few weeks. That’s something we’ve seldom observed, but officials always seem to take credit for the inevitable decline in cases.
More informed lockdown proponents have been hanging their hats on this paper in Nature by Seth Flaxman, et al, published in July. As Philippe LeMoine has shown, however, Flaxman and his coauthors essentially assumed their result. After a fairly exhaustive analysis, Lemoine, a man who understands sophisticated mathematics, offers these damning comments:
“Their paper is a prime example of propaganda masquerading as science that weaponizes complicated mathematics to promote questionable policies. Complicated mathematics always impresses people because theydon’t understand it and it makes the analysis look scientific, but often it’s used to launder totally implausible assumptions, which anyone could recognize as such if they were stated in plain language. I think it’s exactly what happened with Flaxman et al.’s paper, which has been used as a cudgel to defend lockdowns, even though it has no practical relevance whatsoever.”
The Economic Costs of Stringency
So the benefits of stringent lockdowns in terms of averting sickness and death from COVID are speculative at best. What about the costs of lockdowns? We can start with their negative impact on economic activity:
That’s a pretty bad reflection on NPI stringency. In the U.S, a 10% decline in GDP in 2020 amounts to about $2.1 trillion in lost goods and services. That’s just for starters. The many destroyed businesses and livelihoods carry an ongoing cost that could take years to fade, as this graphic on permanent business closures shows:
If you’re wondering about the distributional effects of lockdowns, here’s more bad news:
It’s possible to do many high-paying jobs from home. Not so for blue-collar workers. And distributional effects by size of enterprise are also heavily-skewed in favor of big companies. Within the retail industry, big-box stores are often designated as “essential”, while small shops and restaurants are not. The restaurant industry has been destroyed in many areas, inflicting a huge blow to owners and workers. This despite evidence from contact tracing showing that restaurants and bars account for a very small share of transmission. To add insult to injury, many restaurants invested heavily in safety measures and equipment to facilitate new, safer ways of doing business, only to be double-crossed by officials like Andrew Cuomo and Eric Garcetti, who later shut them down.
Public Health Costs of Stringency
Lives are lost due to lockdowns, but here’s a little exercise for the sake of argument: The life value implied by individual willingness-to-pay for risk reduction comes in at less than $4 million. Even if the supposed 300,000 COVID deaths had all been saved by lockdowns, that would have amounted to a value of $1.2 trillion, about half of the GDP loss indicated above. Of course, it would be outrageously generous to concede that lives saved by NPI’s have approached 300,000, so lockdowns fall far short at the very outset of any cost-benefit comparison, even if we value individual lives at far more than $4 million.
As AJ Kay says, the social and human costs go far beyond economic losses:
I cited specific examples of losses in many of these categories in an earlier post. But for the moment, instead of focusing on causes of death, take a look at this table provided by Justin Hart showing a measure of non-COVID excess deaths by age group in the far right-hand column:
The numbers here are derived by averaging deaths by age group over the previous five years and subtracting COVID deaths in each group. I believe Hart’s numbers go through November. Of greatest interest here is the fact that younger age groups, having far less risk of death from COVID than older age groups, have suffered large numbers of excess deaths NOT attributed to COVID. As Hart notes later in his thread:
These deaths are a tragic consequence of lockdowns.
Educational Costs of Stringency
Many schools have been closed to in-person instruction during the pandemic, leading to severe disruptions to the education f children. This report from the Fairfax County, VA School District is indicative, and it is extremely disheartening. The report includes the following table:
Note the deterioration for disabled students, English learners, and the economically disadvantaged. The surfeit of failing grades is especially damaging to groups already struggling in school relative to their peers, such as blacks and Hispanics. Not only has the disruption to in-person instruction been disastrous to many students and their futures; it has also yielded little benefit in mitigating the contagion. A recent study in The Lancet confirms once again that transmission is low in educational settings. Also see here and here for more evidence on that point.
It’s clear that the “follow-the-science” mantra as a rationale for stringent NPIs was always a fraud, as was the knee-jerk response from those who conflated lockdowns with “leadership”. Such was the wrongheaded and ultimately deadly pressure to “do something”. We can be thankful that pressure was resisted at the federal level by President Trump. The extraordinary damage inflicted by ongoing NPIs was quite foreseeable, but there is one more very ominous implication. I’ll allow J.D. Tucille to sum that up with some of the pointed quotes he provides:
“‘The first global pandemic of the digital age has accelerated the international adoption of surveillance and public security technologies, normalising new forms of widespread, overt state surveillance,’ warned Kelsey Munro and Danielle Cave of the Australian Strategic Policy Institute’s Cyber Policy Centre last month.
‘Numerous governments have used the COVID-pandemic to repress expression in violation of their obligations under human rights law,’ United Nations Special Rapporteur on Freedom of Expression David Kaye noted in July.
‘For authoritarian-minded leaders, the coronavirus crisis is offering a convenient pretext to silence critics and consolidate power,’ Human Rights Watch warned back in April.
There’s widespread agreement, then, that government officials around the world are exploiting the pandemic to expand their power and to suppress opposition. That’s the case not only among the usual suspects where authorities don’t pretend to take elections and civil liberties seriously, but also in countries that are traditionally considered ‘free.’ … It’s wildly optimistic to expect that newly acquired surveillance tools and enforcement powers will simply evaporate once COVID-19 is sent on its way. The post-pandemic new normal is almost certain to be more authoritarian than what went before.”
This post offers a simple representation of the argument against public non-pharmaceutical interventions (NPIs) to subdue the COVID-19 pandemic. The chart below features two lines, one representing the presumed life-saving benefits of lockdown measures or NPI stringency, and another representing the costs inflicted by those measures. The values on the axes here are not critical, though measures of stringency exist (e.g., the University of Oxford Stringency Index) and take values from zero to 100.
The benefits of lives saved due to NPI stringency are assigned a value on the vertical axis, as are the costs of lives lost due to deferred health care, isolation, and other stressors caused by stringency. In addition, there are the more straightforward losses caused by suspending economic activity, which should be included in costs.
One can think of the benefits curve as representing gains from forcing individuals, via lockdown measures, to internalize the external costs of risk inflicted on others. However, this curve captures only benefits incremental to those achieved through voluntary action. Thus, NPI benefits include only extra gains from coercing individuals to internalize risks, while losses from NPI stringency are captured by the cost curve.
My contention is that the benefits of stringency diminish and may in fact turn down at some point, and that costs always increase in the level of stringency. In the chart, for what it’s worth, the “optimal” level of stringency would be at a value of 2, where the difference between total benefits and total costs is maximized (and where the benefits of incremental stringency are equal to the marginal costs or losses). However, I am not convinced that the benefits of lockdown measures ever exceed costs, as they do in the chart above. That is, voluntary action may be sufficient. But if the benefits of NPIs do exceed costs, it’s likely to be only at low levels of stringency.
To the extent that people are aware of the pandemic and recognize risk, the external costs of possible infectiousness are already internalized to some degree. Moreover, there is mutual risk in most interactions, and all individuals face risks that are proportional to those to which they expose others: if your contacts are more varied and your interactions are more frequent and intimate, you face correspondingly higher risks yourself. After all, in a pandemic, an individual’s failure to exercise caution may lead to a very hard internalization of costs if an infection strikes them. This mutuality is an element absent from most situations involving externalities. And to the extent that you take voluntary precautions, you and your contacts both benefit. Nevertheless, I concede that there are individuals who face less risk themselves (the young or healthy) but who might behave recklessly, and they might not internalize all risk for which they are responsible. Yes, stringency may have benefits, but that does not mean it has net benefits.
Even if there is some meaningful point at which NPIs are “optimized”, government does not possess the knowledge required to find that point. It lacks detailed knowledge of both costs and benefits of NPIs. This is a manifestation of the “knowledge problem” articulated by Friedrich Hayek, which hampers all efforts at central planning. In contrast, individual actors know their own tolerance for risk, and they surely have some sense of the risks they create in their normal course of affairs. And again, there is a strong degree of proportionality and voluntary internalization of mutual risks.
While relying on voluntary action is economically inefficient relative to an ideal, full-information and perfectly altruistic solution, it is at least based on information that individuals possess: their own risk profile and risk preferences. In contrast, government does not possess information necessary to impose rules in an optimal way, and those rules are rife with unintended consequences and costs inflicted on individuals.
My next post will present empirical evidence of the weakness of lockdown measures in curbing the coronavirus as well as the high costs of those measures. The coronavirus is a serious infection, but it is not terribly deadly or damaging to the longer-term health of the vast majority of people. This, in and of itself, should be sufficient to demonstrate that the array of non-pharmaceutical interventions imposed in the U.S. and abroad were and are not worthwhile. People are capable of assessing risks for themselves. The externality argument, that NPIs are necessary because people do not adequately assess the risk they pose to others, relies on an authority’s ability to assess that risk, and they invariably go overboard on interventions for which they underestimate costs. COVID is not serious enough to justify a surrender of our constitutional rights, and like every concession to government authority, those rights will be difficult to recover.
In advanced civilizations the period loosely called Alexandrian is usually associated with flexible morals, perfunctory religion, populist standards and cosmopolitan tastes, feminism, exotic cults, and the rapid turnover of high and low fads---in short, a falling away (which is all that decadence means) from the strictness of traditional rules, embodied in character and inforced from within. -- Jacques Barzun