Month: December 2020
1. Computer science rejects computer science (link corrected).
3. A new paper from Treasury suggesting the job effects of PPP loans were significant, WSJ Op-Ed version here.
4. New strain in South Africa? Probably a red herring, but not entirely reassuring to read that it affects young people more. Boris Johnson is saying that the new strain in the UK is 70% more contagious (whatever that means). That too may be speculative, more information here. The general point is you wish to minimize the “reservoir” for such mutations, and that is another reason to want to keep cases low.
5. “It’s not clear why the U.S. Army, the most powerful fighting force in the world, required nearly a year to develop a mask that would have taken the civilian sector mere days—if not hours—to develop.”
Walmart will use fully autonomous box trucks to make deliveries in Arkansas starting in 2021. The big-box retailer has been working with a startup called Gatik on a delivery pilot for 18 months. Next year, the two companies plan on taking their partnership to the next level by removing the safety driver from their autonomous box trucks.
Gatik, which is based in Palo Alto and Toronto, outfitted several multitemperature box trucks with sensors and software to enable autonomous driving. Since last year, those trucks have been operating on a two-mile route between a “dark store” (a store that stocks items for fulfillment but isn’t open to the public) and a nearby Neighborhood Market in Bentonville, Arkansas. Since then, the vehicles have racked up 70,000 miles in autonomous mode with a safety driver.
Next year, the companies intend to start incorporating fully autonomous trucks into those deliveries. And they plan on expanding to a second location in Louisiana, where trucks with safety drivers will begin delivering items from a “live” Walmart Supercenter to a designated pickup location where customers can retrieve their orders. Those routes, which will begin next year, will be longer than the Arkansas operation — 20-miles between New Orleans and Metairie, Louisiana.
Here is the article.
Among groups at higher risk of dying from COVID-19, such as people with diabetes, people with DS stand out: If infected, they are five times more likely to be hospitalized and 10 times more likely to die than the general population, according to a large U.K. study published in October. Other recent studies back up the high risk.
Researchers suspect background immune abnormalities, combined with extra copies of key genes in people with DS—who have three copies of chromosome 21 rather than the usual two—make them more vulnerable to severe COVID-19. “This is a vulnerable population that may need protective policies put in place,” says Julia Hippisley-Cox, a clinical epidemiologist at the University of Oxford’s medical school and senior author on the U.K. study.
On 2 December, the United Kingdom’s Joint Committee on Vaccination and Immunisation recommended prioritizing people with DS for speedy vaccination. But the more than 200,000 Americans with DS so far are not slated for early vaccination. Nor has the U.S. Centers for Disease Control and Prevention (CDC) included DS in its list of conditions it says boost the risk for severe COVID-19.
Here is the article, surely this merits further discussion as we allocate vaccines? And note this higher mortality risk holds even after controlling for other factors, such as living in group homes. And here you will find the original study.
Many professors at universities routinely quizzed their students too, although not as commonly as faculty at smaller colleges did. [In 1910]…a questionnaire of University of Chicago faculty revealed that 25 of 122 replying professors gave some kind of quiz each day; 31 gave them each week, and 10 others did so every other week. The following year, in 1911, a survey of 188 economics professors around the country showed that 171 of them employed “oral quizzes” in class; only 60 of them used written tests. Surveying undergraduates alongside faculty, the 1910 University of Chicago survey found that four of five students favored written tests over oral ones.
That is from Jonathan Zimmerman’s quite interesting The Amateur Hour: A History of College Teaching in America.
6. Whether you agree or not, the method and approach of the government/CDC here on vaccine allocation are so low quality as to almost defy belief. Again, whether or not you are convinced by Matt Yglesias, his simple Substack on this same question did a better job (and at a profit, presumably!). A useful reminder for those of you who “blame the CDC problems on Trump.” Here is related NYT coverage. And never forget Glazer. Why do we not prioritize men, who are at higher risk? And from “an expert in ethics.”
8. “Atlantic City has launched an auction in which the winner will get to virtually push the button that starts the long-anticipated implosion of the former Trump Plaza Hotel and Casino building.” Link here.
I’ve been arguing that we should delay the second dose (or at least not hold back first doses) in order to hit the virus hard and inoculate more people on the first dose. I wrote:
We should vaccinate 6 million people with first dose NOW. It is deadly cautious to hold second dose in *reserve*. Supply chain will be ok and the exact timing of the second dose is not magical and likely not critical. In the accidental low-dose, standard-dose regime for the AZ vaccine, people got the second dose 7 to 8 weeks after the first dose and that was the 90% efficacious regime. [A different vaccine obviously but ] Exact timing of the second-dose does not seem critical, although everyone should get a second dose.
Today epidemiologist Michael Mina and writer Zeynep Tufekci, who has been ahead of the curve on much of the discussion, make the case even more strongly in the NYTimes:
First, the science. While the vaccine trials were designed to evaluate a two-dose regimen, some immunity might be acquired before a second dose is administered. We know, for instance, that a Covid-19 infection appears to yield protection for at least six months. While infections are not vaccinations, and while we need more data on this, it’s plausible that the immunity gained from a vaccination may turn out to be even stronger than what comes from an infection. The reason we do a second — booster — vaccination is that these later doses help to solidify immune memory, in part by giving extra training to the cells that produce antibodies, a process called affinity maturation. But this process begins with the single dose, and the evidence collected between the time of the first and second doses in tens of thousands of people in the Phase 3 trials suggests that the level of affinity maturation may provide enough protection to meet the standards we have set for vaccine approval during this pandemic even without the second dose.
While we know that the single dose can protect against disease, we don’t yet know how long this immune protection will last, and at what level. However, there is no rule that says that vaccines must be boosted within weeks of each other. For measles, the booster dose is given years after the first dose. If the booster dose could be given six months or a year after the first dose, while maintaining high efficacy before the second dose, that would allow twice as many people to get vaccinated between now and later next year, accelerating herd immunity — greatly helping end the crisis phase of the pandemic in the United States.
… we should begin immediate single-dose trials, recruiting volunteers from low-risk populations who are first in line for the vaccinations. For example, among health care workers protective equipment works, rates of infection among this group have fallen sharply and severe disease is much more rare.Younger essential workers without risk factors are less likely to be severely affected if they are exposed since this disease’s impact rises steeply with age. Just as tens of thousands of people volunteered for the earlier vaccine trials, many may well volunteer to test a placebo against a second dose, allowing us to quickly ascertain questions of durability and effectiveness of the single dose.
Two additional points. First, mix and match, as I argued earlier, may be beneficial:
…we could mix and match vaccines. The UK will run a trial on this question. Mix and matching has two potentially good properties. First, mix and matching could make the immune system response stronger than either vaccine alone because different vaccines stimulate the immune system in different ways. Second, it could help with distribution. It’s going to be easier to scale up the AZ vaccine than the mRNA vaccines, so if we can use both widely we can get more bang for our shot.
Second, an economics issue. If we want Pfizer and Moderna on board we need to pay them not just to run the clinical trials but to be happy with potentially selling half as many doses. Incentives matter.
In the present wreck of empires, and under the extinction of all international law, no small state can hope to maintain its independence. Great Britain and Ireland, from their situation, their language, and their mutual necessities, seem naturally destined to support each other’s strength, and supply each other’s wants; and we are quite convinced, that nothing but extreme misgovernment can separate them. Heavy indeed, then, will be the responsibility of those men, under who administration, or by whose previous unconciliatory measures such a separation is effected — whether the immediate cause of it be foreign conquest, or internal commotion.
That is Thomas Robert Malthus, “On the State of Ireland (II), published in the Edinburgh Review in 1809. It made perfect sense back then — and today — and yet for entirely different and indeed almost opposite reasons.
You may recall I already posted my best non-fiction books and best fiction books of 2020. But, unlike on previous lists, I didn’t pick a very best book of the year because in my gut I felt it had not yet arrived. Now I have a top three, all of which came after I posted my original list. Here are my top three picks for the year:
David S. Reynolds. Abe: Abraham Lincoln in His Times. At some point I vowed never to read another Lincoln biography, but this one won me over with its readability and also grasp of the broader cultural and political context. You may know Reynolds from his excellent Walt Whitman book — could there be a better background to write on Lincoln? Conceptual throughout. At 932 pp. every page of this one is instructive, even if you feel sated in Lincoln as I did.
Heather Clark, Red Comet: The Short Life and Blazing Art of Sylvia Plath. This is like the Lincoln biography — I was convinced I didn’t want to read a thousand pages about her (though I am a fan). And yet I keep on reading, now at about the halfway mark and I will finish with joy. This is one of the best and most gripping biographies I have read, covering growing up as a brilliant young woman in the 1950s, poetry back then, dating and gender relations amongst the elite at that time, how mental health problems were dealt with, and much more.
Jan Swafford, Mozart: the Reign of Love. Self-recommending. A wonderful biographer covers one of the most important humans, to produce the best Mozart biography of all time. You may recall I also had high praise for Swafford’s Beethoven biography from 2014.
Those are my top three books of the year. I think you can make a good case for Joe Henrich’s WEIRD book having the most important ideas of the year in it, but, perhaps because I already had read much of the material in article form, I didn’t love it as a book the way I do these.
Finally, I will note that the “best books lists” of other institutions have grown much worse, even over the last year. A good list has never been more valuable, and please note my recommendations are never done to fill a quota, “achieve balance,” right previous wrongs, or whatever. They are what I think are the best books. Scary how rare that has become.
Young adults are dying at historic rates. In research published on Wednesday in the Journal of the American Medical Association, we found that among U.S. adults ages 25 to 44, from March through the end of July, there were almost 12,000 more deaths than were expected based on historical norms.
In fact, July appears to have been the deadliest month among this age group in modern American history. Over the past 20 years, an average of 11,000 young American adults died each July. This year that number swelled to over 16,000.
The trends continued this fall. Based on prior trends, around 154,000 in this demographic had been projected to die in 2020. We surpassed that total in mid-November. Even if death rates suddenly return to normal in December — and we know they have not — we would anticipate well over 170,000 deaths among U.S. adults in this demographic by the end of 2020.
That is from Jeremy Samuel Faust, Harlan M. Krumholz and at the NYT. To be clear, this is not the main problem, but it is not a nothingburger either. 3,656 deaths per day right now, no matter what the ages how many other American catastrophes can rival that? #1 cause of death in the country right now, bar none.
I have pondered this matter further, and have upon reflection abandoned my previous concern with the high M2 figures. Here is one excerpt from my latest Bloomberg column:
To understand forthcoming inflation rates further, consider why monetary aggregates such as M2 have been rising so rapidly. It is not that the U.S. Federal Reserve wishes to relive the experience of the 1970s. It’s that many businesses have been borrowing while they can, fearing they may end up strapped for cash as the pandemic continues.
From that point onward, there are two possible paths. The first is that businesses will actually need that extra cash, due to low consumer demand and a stalled recovery. In that case, the potentially inflationary boost from a higher money supply would be offset by lower demand elsewhere in the economy. The velocity of money would be weak, and there would be no reason to expect major inflationary pressures.
The second possible path is that the recovery will proceed at a rapid clip, and businesses will have extra cash on their hands for a while. That likely would translate into lower borrowing for the rest of the year and a smoothing of monetary flows — and no major increase in inflationary pressures.
There will, however, be some areas of the economy in which inflation will likely be high. Consider a sector in which demand is largely seasonal and has been stifled by the pandemic, and supply in the short run is inelastic. One obvious example is summer vacation travel.
You also might say that I am agreeing with the market forecasts embedded in interest rates, indexed securities prices, and so on.
1. Those new service sector jobs: death witnesses (Canada).
3. An NPR segment with Cardiff Garcia and me: after the Great Stagnation ends, what will be overrated and underrated?
I know, I know, you are tired of me beating on the FDA. Too bad. The beatings will continue until mortality improves.
Today’s beating, however, concerns hearing aids rather than than vaccines. The current system, much like the optometry racket, requires that hearing aids be purchased through an audiologist:
WEBMD: A pair of hearing aids runs $5,600 on average — a cost that health insurance doesn’t typically pick up. In order to get the devices, people with hearing loss must have a medical exam to rule out the slim chance of a serious medical problem that is causing the hearing loss, or they must sign a waiver opting out of the exam. You can only buy hearing aids through an audiologist or a licensed hearing aid dispenser, who is authorized to test hearing and sell hearing aids.
…choosing a hearing professional limits a person’s hearing aid choices, and changing audiologists as a result of location or personal preference could require a person to change hearing aids, too.
Due in part to these hurdles, once a person starts to lose their hearing, they wait an average of 7 years before they seek help.
The NYTimes continues the story:
By now, we were supposed to be swiftly approaching the day when we could walk into a CVS or Walgreens, a Best Buy or Walmart, and walk out with a pair of quality, affordable hearing aids approved by the Food and Drug Administration….in 2017, Congress passed legislation allowing the devices to be sold directly to consumers, without a prescription from an audiologist. The next step was for the F.D.A. to issue draft regulations to establish safety and effectiveness benchmarks for these over-the-counter devices.
Of course, the FDA is likely to miss the deadline. But even I am willing to cut them some slack, given other more pressing concerns. The real issue, however, is one that the NYTimes doesn’t confront and that is why does the FDA regulate hearing aids at all?
Headphones and earbuds are not regulated as medical devices and hearing aid technology is similar to hearings buds with a sophisticated equalizer. Indeed, you can even buy “personal sound amplification products” that are not FDA regulated or approved. It’s only the somewhat arbitrary declaration that these products are “medical devices” that brought them under FDA regulation to the detriment of consumers. The National Academy of Sciences report agrees writing:
FDA has established regulations for hearing aids, including quality system regulation requirements, mandatory labeling, and pre-purchase medical evaluation (or a documented waiver). These regulations, along with a number of state regulations, have largely restricted the availability of hearing aids to being mainly dispensed through medical, audiology, or hearing instrument specialist venues.
…In examining the Food and Drug Administration’s (FDA’s) requirements for physician evaluation prior to obtaining hearing aids, the committee finds no evidence that the required medical evaluation or waiver of that evaluation provides any clinically meaningful benefit.
Addendum: In some positive news the FDA lifted its ban on the first non-prescription, at-home SARS-COV test and also on a prescription, paper-strip test. Now all we need is a non-prescription paper-strip test! Come on guys.
Maybe not, or so I argue in my latest Bloomberg column. As you may know, the first day price pops for Airbnb and Doordash were considerable, Airbnb more than doubling in its first day of trading: Here is one excerpt:
On IPO day, each prospective buyer is wondering what the shares will be worth, and to a great extent looking to the judgment of the other investors. A buyer might start the day willing to pay $60 a share, but upon seeing that many others are willing to pay more, maybe she will, too. It is like Keynes’s famed “beauty contest,” where investors are guessing as much about each other as about the company.
In such a setting, prices can rise or fall extremely quickly, as the very process of trading reveals information about the stock’s value. That in turn makes it possible for the share price to soar on the first day of trading, creating the “pop.”
Now consider this scenario from the perspective of the issuing investment bank. If it sets the IPO price too high, it may set off a downward spiral of negative enthusiasm. Traders will see that most of the other traders think it is overpriced, leading to a plunge.
It’s all a bit like a restaurant on a Saturday night. If the place is seen as “cool” — whether because of its food and service (the product), its setting (the physical asset) or its ambience (the brand) — there will be a line out the door. Otherwise it will be fairly empty. Furthermore, the presence of a line will draw continued interest over time. It is hard or maybe even impossible to set prices so that every table is filled yet there is no line. To deploy some technical language, the demand curve may be discontinuous.
In this position, the IPO issuer likely will set the initial price too low — leading to a “line,” excess demand, and a big run-up in price on the first day. If the price is super-high in the first place, the market mood would be nervousness rather than eagerness, and most investors wouldn’t be able to see the surges in demand visible in lower price ranges.
Keep in mind that this surge in buying interest only has to make investors modestly more enthusiastic about the quality of the firm to generate a potentially big increase in final valuation.
All this said, in the current case, there is the question of why the initial prices were so low. Several theories present themselves: The markets for DoorDash and Airbnb might be more “winner take all” than usual. The value of those companies might be more closely linked to the value of their intangible assets. Or maybe the future of online services might be especially hard to predict in the midst of a pandemic, thus inducing larger bandwagon effects.
It is worth noting that differing auction systems for IPOs have not produced obviously superior results.
This is all David:
I have a couple new papers on tax progressivity and redistribution that may be of interest to you. Both used CBO data to avoid the PSZ-AS differences. Abstracts below.
The first paper is about the ends of the distribution: tax progressivity has increased significantly since 1979 (and steadily since 1986) due to more generous tax credits for the bottom, while average tax burdens of the top have been relatively unchanged because lower marginal rates were offset by decreased use of tax shelters. The online appendix shows why the CBO estimates differ from those of Saez and Zucman (see Fig. B7 at the end; it’s mostly due to refundable credits at the bottom and imputed income at the top) and the Heathcote et al. paper you blogged about a couple months ago (it’s technical differences and their inclusion of some transfers, but their most similar measure of tax progressivity was not flat—it increased 21 percent since 1979).
The second paper, with Adam Looney and Jeff Larrimore, is about the middle of the distribution. Since 1979, we found that non-elderly middle-class market income increased 39 percent in real per person terms. The increase was 57 percent when accounting for taxes and transfers. This seems to fit with the “updated” view of stagnation—expanding male wages to also look at untaxed compensation and including female compensation and taxes/transfers shows larger median growth. But there was a structural break in 2000. Before then, middle-class incomes grew at the same rate before and after taxes and transfers, and since then income after taxes and transfers grew three times faster (Fig. 6 on page 19). We don’t discuss the recent market income slowdown (maybe related to the debated labor share break around 2000), but we show that the additional fiscal support that filled the gap looks like an unsustainable way to boost middle-class disposable incomes going forward.”