Did this happen? Were Spain’s hardest hit provinces in the spring spared in the second wave?
To get a quick sense of the answers to those questions I plotted the cumulative number of cases per 100,000 population in the Spanish provinces since June 15 against the proportion of the population in the provinces that tested positive for antibodies after the first wave. If herd immunity were playing a large role in suppressing cases in the second wave, we would expect to see a negative relationship between provinces with high levels of antibodies in the population at the end of May and total case counts since that time…
Instead of a negative correlation, there is a positive, although weak, correlation between having higher prevalence of antibodies in the population and having a higher case rate in the second wave.
…Take Madrid for example, if roughly 13% of the population had antibodies after the first wave, at least one of the low-HIT models estimates the Rₑ would be approximately 60% lower than than the unmitigated reproductive rate (R₀). If population immunity were reducing transmission in the Madrid area by 60% below unmitigated levels, it seems unlikely Madrid would again have one of the highest rates of infection in the second wave [yet it does].
…Ultimately, the strongest conclusion that can be drawn from this look at infection rates is that there is not clear evidence herd immunity is playing a significant role, yet.
Also take a look at a deeper dive looking for herd immunity in Sweden (spoiler alert: no signs of it yet).
It is fine to call this inconclusive, but still the pattern predicted by standard herd immunity claims is not yet showing up. Here is the whole piece from Kbenes, very useful.
And elsewhere, this was not supposed to happen, as New York Orthodox Jews also have been cited as a “herd immunity” community:
Officials this week released statistics showing that the positivity rate in some Orthodox Jewish neighborhoods [in NYC] had grown to anywhere from 3 percent to 6 percent, significantly more than the city’s overall rate of between 1 percent and 2 percent. Officials are especially worried about the positivity rates in the Brooklyn neighborhoods of Borough Park, Midwood and Gravesend, which they have referred to as the “Ocean Parkway Cluster.”
Here is that full story (NYT).
Once upon a time, there were some herd immunity theorists. They claimed that once a certain percentage of the population had been infected, the R for Covid would fall below one and the disease would become far less common and less significant. Since these analysts were especially aware of heterogeneity issues (though common in the broader scholarly literature), these same herd immunity theorists tended to be less pessimistic than many of the mainstream forecasts.
To be clear, everyone knew that herd immunity was a general and universally accepted concept in the literature. But these particular herd immunity theorists were the ones saying it really would matter, and they did so in the bold and fearless manner. As I mentioned earlier, the NYT didn’t really start covering this issue until this August, a kind of unbelievable (and appalling) communications failure from public health experts who didn’t want to say anything that might be construed as minimizing expected risk.
Now, I don’t recall many of those theorists early on making a prediction about a specific number required for the herd immunity threshold to be reached. Nonetheless, when deaths and hospitalizations collapsed in Sweden, London, and New York at about 20 percent seroprevalence, obviously it seemed that might be the critical level for herd immunity to kick in. (Higher measured levels of seroprevalence, such as for the slums of Mumbai might just come from the speed of ripping through a very dense and exposed community.) And a lot of the observed later waves were in fact coming in other parts of these countries or regions, such as Barcelona following Madrid, or Arizona following New York.
These herd immunity theorists were correct in predicting an “earlier than the mainstream is telling you” collapse in deaths and hospitalizations in the hard hit regions. And that is very much to their credit.
You will note that part of their prediction or implied prediction was that past the herd immunity point cases should fall, not just deaths. Transmission just would not be very effective or speedy any more, so cases should be low whether or not people die in the hospitals or the hospitals can save them. I’ll be coming back to this.
Then things started to go askew in the last few weeks. First, it seems like a bad second wave came to an already fairly hard hit Madrid. OK, you could say Madrid was never had 20% seroprevalence to begin with. And then what appears to be a second wave has started coming to Israel, with rising hospitalizations. Finally, it is believed that in Britian R equals about 1.7, and that a second wave of cases is on the verge of hitting London and Southeast England. That hasn’t quite happened yet, but the informed authorities greatly fear it, and the numbers so far seem to indicate that as the trend.
Added all up, those data points are not decisive in rejecting the claims of these herd immunity theorists. But they do make the herd immunity theorists look less correct than they did say three weeks ago. Those “partial second waves,” or whatever they turn out to be, seem more active than one might have expected. Again, though, the story is still unfolding and we should not rush to final conclusions. But in the meantime we should update!
In response, many of the herd immunity theorists strike back and ask “where are the deaths“? But that is not the right question for testing herd immunity claims. Those claims were about transmission slowing down, and those claims should be true about Covid-19 cases whether or not more people are surviving in the hospital. (Imagine for instance a perfect antiviral that saved everybody — would that mean herd immunity was true a priori? Nope.)
Another claim from some of the less careful herd immunity theorists is that cases are rising again because testing is rising. That doesn’t seem to explain observed patterns in Israel, Spain, or England, where in all instances actual Covid cases are rising above and beyond what is going on with testing policy, and by some considerable margin. It probably does explain some parts of America, however.
It is very likely that death rates will be much lower this time around, because of better procedures, younger victims, lower doses, and possible (speculative!) variolation through mask use over time, exposing people to lower doses repeatedly and boosting their immune responses.
There is a temptation to say “few deaths, we don’t need lockdowns!” Indeed, the more partisan of the herd immunity theorists are obsessed with the lockdown issue. Lockdowns are important questions, but don’t let your lockdown views skew your interpretation of the numbers, and furthermore there are many other important Covid questions of interest, for instance:
1. How much more should we invest in better hospital procedures? Better biomedical fixes? And how much should we hurry? If transmission is mostly over, you can relax much more, but ongoing cases both will bring some long-term damages (short of death) and also some ongoing panic, whether rational or not.
2. How do we deal with the fact that case numbers per se will scare people for a long time to come? Again, if transmission is winding down, you don’t need as big a long-term plan here.
3. Should you let large swarms of tourists into your currently semi-protected region, say it is Venice, Italy or the less infested parts of Hawaii?
4. To the extent there is current herd immunity or semi-herd immunity as I call it, how fragile is that arrangement with respect to a possible rotation of potential super-spreaders? And what might set off those fragilities?
For those questions, and indeed many others, it matters a great deal whether the original herd immunity prediction about “permanently low cases past the herd immunity threshold” is correct, or not. Whether the death rate is high or low. You really do need to understand about the cases in their own right, once you see this broader spectrum of issues at stake.
The more partisan herd immunity theorists wish to debate “how terrible will this be and will that justify a lockdown?”, and then they seek to talk you into a mood of not being so terrified, because frequently they are lockdown skeptics. Again, that is a super-important question. But don’t let it distract you from the other important questions at hand.
And for those other questions, as I’ve already stated above, the trajectory caseload predictions of the herd immunity theorists are looking worse than they did a few weeks ago.
Of course I will be giving you updates on this matter as time passes. But this is the very latest, namely that some of the herd immunity theorists are on the precipice of being dogmatically wrong about matters of real import, just as were some of the most pessimistic mainstream predictions from March and April.
We examine the impact of criminalizing sex work, exploiting an event in which local officials unexpectedly criminalized sex work in one district in East Java, Indonesia, but not in neighboring districts. We collect data from female sex workers and their clients before and after the change. We find that criminalization increases sexually transmitted infections among female sex workers by 58 percent, measured by biological tests. This is driven by decreased condom access and use. We also find evidence that criminalization decreases earnings among women who left sex work due to criminalization, and decreases their ability to meet their children’s school expenses while increasing the likelihood that children begin working to supplement household income. While criminalization has the potential to improve population STI outcomes if the market shrinks permanently, we show that five years post-criminalization the market has rebounded and the probability of STI transmission within the general population is likely to have increased.
That is from a new NBER working paper by Lisa Cameron, Jennifer Seager, and Manisha Shah.
Tata group has received approval from the Drug Controller General of India (DCGI) for the commercial launch of the country’s first CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats) coronavirus test ‘Feluda’, the Council of Scientific and Industrial Research (CSIR) said on Saturday. This test uses an indigenously developed, cutting-edge CRISPR technology for detection of the genomic sequence of SARS-CoV-2 virus, CSIR said in a statement.
The Tata CRISPR test achieves accuracy levels of traditional RT-PCR tests with quicker turnaround time, less expensive equipment and better ease of use.
Here is the full story, via Alex HR.
Buggert’s study in Sweden seems to support this position. Investigating close family members of patients with confirmed covid-19, he found T cell responses in those who were seronegative or asymptomatic. While around 60% of family members produced antibodies, 90% had T cell responses. (Other studies have reported similar results.) “So many people got infected and didn’t create antibodies,” concludes Buggert.
That is from Peter Doshi, mostly a survey on pre-existing immunity, interesting and useful and properly agnostic throughout. Here is a version of the Buggert piece, also with a link to the published version.
Note two things. First, “the kooks” saw this possibility first, and insisted on its relevance, to their credit. Second, many of “the kooks” are overly dogmatic, not always to be trusted, and they commonly shift the goalposts (when predictions about cases are falsified, they switch to pretending those were predictions about deaths). Often the non-kooks do that too of course.
For a sobering worry, here are some recent numbers from Spain.
The key to interpreting the literature is to focus on the data, and to keep an open frame of mind, rather than digging in to a particular position. Right now I am focused on observing what kind of second wave London is going to have, and how mild or bad it will be, as that is most likely to induce me to update my positions, in one direction or another.
For the pointer I thank E. Ward.
But Congress did ultimately chop off a leg when it repealed the mandate penalties in 2017 — and, despite these predictions, the Affordable Care Act still stands. New federal data and economic research show the law hasn’t collapsed or entered the “death spiral” that economists and health insurers projected.
Many experts now view the individual mandate as a policy that did little to increase health coverage — but did a lot to invite political backlash and legal challenges.
The newest evidence comes from census data released Tuesday, which shows health coverage in the United States held relatively steady in 2019, even though Congress’s repeal of the mandate penalties took effect that year.
“The stool might be a bit rocky, but you can get away with two legs,” said Evan Saltzman, a health economist at Emory University who studies the topic. “It’s like the table at the restaurant that is a little wobbly. You can still sit at it, even if it’s not quite as pleasant.”
That is from Sarah Kliff at the NYT, the whole piece is excellent and full of substance. And:
Mr. Saltzman went on to earn a doctorate in economics after his job at RAND, and focused his research on the mandate. He has found that the mandate isn’t a very effective tool for increasing enrollment. One recent paper of his estimated that eliminating the mandate penalties would reduce marketplace enrollment by 2 percent and increase premiums by 0.7 percent.
“My viewpoint on the mandate has changed,” he said. “Back in 2012, my sense was it was essential. The evidence indicates that the marketplaces are doing about the same as they were before the mandate was set to zero.”
Separately, in The New England Journal of Medicine last year, researchers concluded that “the individual mandate’s exemptions and penalties had little impact on coverage rates.”
To be clear, this surprises me too. Was it Ross Douthat who once said on Twitter that it was the Trump administration and the Republican courts that saved Obamacare? The Krugman line, pushed without qualification for over a decade (and with incessant moralizing), that all of the legs of the stool are necessary, seems…wrong. I would say be careful with this one, as sometimes elasticities don’t kick in for a long time (as maybe with the corporate income tax cuts as well?…let’s be consistent here…). Still, it seems that an update of priors is in order. As you will see in the piece, even Jonathan Gruber thinks so.
And here are useful comments from John Graves.
That is the topic of my latest Bloomberg column, here is one excerpt:
Ideally, a government will wish to publicize the announcement of a vaccine while slow-walking the actual distribution. That way, if there is something wrong with the brew, it can stop distribution before too many of its citizens experience adverse side effects. In essence, the approving countries are doing a version of their Phase III trials with fewer scientific controls and more out in the open. For Russia in particular, it is not obvious how much it is really ahead of other countries.
One possible American strategy would be to encourage the early approvers to distribute and test their vaccines on a broader scale, and then make their data freely available. Given close working relations, this may be easier to accomplish with the UAE than with China or Russia. If one of those vaccines turned out to be good enough, the U.S. has the resources either to buy doses or to reverse engineer it.
U.S. decisions on approval speed, meanwhile, will depend on what other nations do. For instance, if the early approvers are gathering useful data through their experiments, U.S. officials might decide not to hurry so much, instead preferring to let foreigners take the risks. That sounds good, but it could be counterproductive for the world as a whole. America is the country most likely to come up with the highest quality vaccine. Slowing down the U.S. will mean that more of the world gets the (possibly) lower-quality but more readily available Chinese product.
One tension in “vaccine relations” is that richer countries and poorer countries do not want exactly the same thing. Typically, the wealthier the country, the more risk-averse its citizens, and the less need to hurry.
China may be unique: It has some properties of a rich country (a big, advanced scientific establishment), but it has a poor country’s willingness to take risks. That’s one reason China might end up leading on vaccines. The U.S. is ahead of China technologically, but Chinese priorities are more in sync with those of many other countries in the world.
There are further arguments at the link.
Fans of horror films exhibit less psychological distress during COVID-19.
Fans of “prepper” films reported being more prepared for the pandemic.
Morbidly curious people exhibit greater positive resilience during COVID-19.
Morbidly curious people are more interested in pandemic films during the pandemic.
Speculative, and yes replication crisis, but consistent with my intuitions, and in any case a question worthy of further study. Here is the full paper, by Coltan Scrivner, et.al., via the excellent Kevin Lewis.
Life expectancy in the US increased 3.3 years between 1990 and 2015, but the drivers of this increase are not well understood. We used vital statistics data and cause-deletion analysis to identify the conditions most responsible for changing life expectancy and quantified how public health, pharmaceuticals, other (nonpharmaceutical) medical care, and other/unknown factors contributed to the improvement. We found that twelve conditions most responsible for changing life expectancy explained 2.9 years of net improvement (85 percent of the total). Ischemic heart disease was the largest positive contributor to life expectancy, and accidental poisoning or drug overdose was the largest negative contributor. Forty-four percent of improved life expectancy was attributable to public health, 35 percent was attributable to pharmaceuticals, 13 percent was attributable to other medical care, and −7 percent was attributable to other/unknown factors. Our findings emphasize the crucial role of public health advances, as well as pharmaceutical innovation, in explaining improving life expectancy.
A Chinese pharmaceutical company has injected hundreds of thousands of people with experimental Covid-19 vaccines, as its Western counterparts warn against administering mass vaccinations before rigorous scientific studies are complete.
China National Biotec Group Co., a subsidiary of state-owned Sinopharm, has given two experimental vaccine candidates to hundreds of thousands of people under an emergency-use condition approved by Beijing in July, the company said this week. Separately, Chinese drugmaker Sinovac Biotech Ltd. said it has inoculated around 3,000 of its employees and their family members, including the firm’s chief executive, with its experimental coronavirus vaccine.
The three vaccine candidates are still undergoing Phase 3 clinical trials, which involve testing a vaccine’s safety and effectiveness on thousands of people. Six other leading Covid-19 vaccine candidates are also in this final phase, according to the World Health Organization.
I am agnostic on this! Of course we will see how it goes, and you should note that if the Chinese vaccines turn out to be “good enough,” they will spread to poorer countries rather quickly.
I see so much not so high quality moralizing from public health figures on Twitter, backed only by adjectives or appeals to authority. Until they “show their work” with actual numbers and probabilities, my current view is to think this Chinese policy stands a reasonable (but by no means certain) chance of passing the Benthamite test.
Please note: this does not mean America should do the same! In fact, China rushing may well lower the benefits from an American rush, because the major gains at stake here are the easing of non-Covid deaths and deprivations in South Asia and other poor parts of the world. Maybe the optimal portfolio is indeed a “China + Russia rush,” followed by some good’ ol American patience. (Is that what we do? Who said that!?)
Here is the underlying WSJ piece.
Trouble in the Madrid region is brewing again, even though earlier seroprevalance had clocked in at about 20 percent:
Good for New York of course, here is a thread discussing the comparison, to me the conclusions seem premature. The important point in any case is that Covid-protected time periods need not last forever, and you can end up in multiple rounds of “let it rip.” As far as I know, this is the first established case of a major “second wave” in a previously hard-hit area.
The good news is that Madrid cases seem to have peaked, and furthermore the death rate is much lower the second time around, the latter being one good reason for postponing cases into later time periods rather than taking them all up front.
Note also that England has had months of open pubs, and a very quiet situation, but now cases there are doubling every six to seven days (FT). Don’t switch back to talk of deaths! The “simple” theory of herd immunity is surprised to see that new trend in cases. What I call semi-herd immunity suggests a high degree of protection for the current configuration of social relations, after some point. As those social relations change, some of that temporary herd immunity dissolves, as new infecting connections are being created and new superspreaders arise and do their thing. But that takes a while, possibly months.
The herd immunity theorists downplay the possible temporariness of the equilibrium they pinpoint. They instead prefer to focus on the (correct) point that most of the mainstream approaches did not forecast the collapse in deaths and hospitalizations found in England, Sweden, New York, and now parts of the American South. In reality, you need to put both sides of the picture together, and grasp both the insights and limitations of the herd immunity theorists.
So herd immunity does seem to be fragile, and if other developments (treatment, antivirals, steroids, masks and thus lower dosage) lower death rates, bravo, but case behavior still moves against the simple herd immunity theory, at least in Madrid. How fragile we still do not know, and I readily grant and indeed would emphasize that Madrid is the only major counterexample to date. Appreciate the limits of knowledge!
If you listen to Ivor Cummins, a darling of the herd immunity theorists, he doesn’t seem to grasp these problems of possible temporariness (he loves to switch to talk of deaths at just the wrong time), but rather treats herd immunity as “it’s over,” with a few vague qualifiers tossed in at the very end. We will see.
Yes, in short. Here is a new paper from Corey Deangelis and Christos Makridis:
The COVID-19 pandemic led to widespread school closures affecting millions of K-12 students in the United States in the spring of 2020. Groups representing teachers have pushed to reopen public schools virtually in the fall because of concerns about the health risks associated with reopening in person. In theory, stronger teachers’ unions may more successfully influence public school districts to reopen without in-person instruction. Using data on the reopening decisions of 835 public school districts in the United States, we find that school districts in locations with stronger teachers’ unions are less likely to reopen in person even after we control semi-parametrically for differences in local demographic characteristics. These results are robust to four measures of union strength, various potential confounding characteristics, and a further disaggregation to the county level. We also do not find evidence to suggest that measures of COVID-19 risk are correlated with school reopening decisions.
And please do note that last sentence again:
We also do not find evidence to suggest that measures of COVID-19 risk are correlated with school reopening decisions (emphasis added).
Via the excellent Kevin Lewis.
Here is a new paper from , , and :
Background Recent reports based on conventional SEIR models suggest that the next wave of the COVID-19 pandemic in the UK could overwhelm health services, with fatalities that far exceed the first wave. These models suggest non-pharmaceutical interventions would have limited impact without intermittent national lockdowns and consequent economic and health impacts. We used Bayesian model comparison to revisit these conclusions, when allowing for heterogeneity of exposure, susceptibility, and viral transmission. Methods We used dynamic causal modelling to estimate the parameters of epidemiological models and, crucially, the evidence for alternative models of the same data. We compared SEIR models of immune status that were equipped with latent factors generating data; namely, location, symptom, and testing status. We analysed daily cases and deaths from the US, UK, Brazil, Italy, France, Spain, Mexico, Belgium, Germany, and Canada over the period 25-Jan-20 to 15-Jun-20. These data were used to estimate the composition of each country’s population in terms of the proportions of people (i) not exposed to the virus, (ii) not susceptible to infection when exposed, and (iii) not infectious when susceptible to infection. Findings Bayesian model comparison found overwhelming evidence for heterogeneity of exposure, susceptibility, and transmission. Furthermore, both lockdown and the build-up of population immunity contributed to viral transmission in all but one country. Small variations in heterogeneity were sufficient to explain the large differences in mortality rates across countries. The best model of UK data predicts a second surge of fatalities will be much less than the first peak (31 vs. 998 deaths per day. 95% CI: 24-37)–substantially less than conventional model predictions. The size of the second wave depends sensitively upon the loss of immunity and the efficacy of find-test-trace-isolate-support (FTTIS) programmes. Interpretation A dynamic causal model that incorporates heterogeneity of exposure, susceptibility and transmission suggests that the next wave of the SARS-CoV-2 pandemic will be much smaller than conventional models predict, with less economic and health disruption. This heterogeneity means that seroprevalence underestimates effective herd immunity and, crucially, the potential of public health programmes.
This would appear to be one of the very best treatments so far, though I would stress I have not seen anyone with a good understanding of the potential rotation (or not) of super-spreaders, especially as winter comes and also as offices reopen. In that regard, at the very least, modeling a second wave is difficult.
Via Yaakov Saxon, who once came up with a scheme so clever I personally sent him money for nothing.
Let’s hope this is true!:
Policy makers in Africa need robust estimates of the current and future spread of SARS-CoV-2. Data suitable for this purpose are scant. We used national surveillance PCR test, serological survey and mobility data to develop and fit a county-specific transmission model for Kenya. We estimate that the SARS-CoV-2 pandemic peaked before the end of July 2020 in the major urban counties, with 34 – 41% of residents infected, and will peak elsewhere in the country within 2-3 months. Despite this penetration, reported severe cases and deaths are low. Our analysis suggests the COVID-19 disease burden in Kenya may be far less than initially feared. A similar scenario across sub-Saharan Africa would have implications for balancing the consequences of restrictions with those of COVID-19.
Here is the full paper.