Saturday assorted links
2. Why isn’t Florida doing worse?
3. Podcast with Chris Murray on the IHME model (have not heard it yet).
4. Iceland update, lots of data, useful pictures.
5. A French cluster: “The overall IAR was 25.9% (95% confidence interval (CI) = 22.6-29.4), and the infection fatality rate was 0% (one-sided 97.5% CI = 0-2.1).”
Social distancing should never be too restrictive
That is the topic of a new paper by Farboodi, Jarosch, and Shimer, published version in here. They favor ” Immediate social distancing that ends only slowly but is not overly restrictive.” Furthermore, they test the model against data from Safegraph and also from Sweden and find that their recommendations do not depend very much on parameter values.
Here is an excerpt from the paper:
…social distancing is never too restrictive. At any point in time, the effective reproduction number for a disease is the expected number of people that an infected person infects. In contrast to the basic reproduction number, it accounts for the current level of social activity and the fraction of people who are susceptible. Importantly, optimal policy keeps the effective reproduction number above the fraction of people who are susceptible,although for a long time only mildly so. That is, social activity is such that, if almost everyone were susceptible to the disease, the disease would grow over time. That means that optimal social activity lets infections grow until the susceptible population is sufficiently small that the number of infected people starts to shrink. As the stock of infected individuals falls,the optimal ratio of the effective reproduction number to the fraction of susceptible people grows until it eventually converges to the basic reproduction number.
To understand why social distancing is never too restrictive, first observe that social activity optimally returns to its pre-pandemic level in the long run, even if a cure is never found. To understand why, suppose to the contrary that social distancing is permanently imposed, suppressing social activity below the first-best (disease-free world) level. That means that a small increase in social activity has a first-order impact on welfare. Of course, there is a cost to increasing social activity: it will lead to an increase in infections. However,since the number of infected people must converge to zero in the long run, by waiting long enough to increase social activity, the number of additional infections can be made arbitrarily small while the benefit from a marginal increase in social activity remains positive.
Recommended, one recurring theme is that people distance a lot of their own accord. That means voluntary self-policing brings many of the benefits of a lockdown. Another lesson is that we should be liberalizing at the margin.
If I have a worry, however, it has to do with the Lucas critique. People make take preliminary warnings very seriously, when they see those warnings are part of a path toward greater strictness. When the same verbal or written message is part of a path toward greater liberalization however…perhaps the momentum and perceived end point really matters?
For the pointer I thank John Alcorn.
Big G
Yes, that is the title of a new paper and it is excellent indeed. Lydia Cox et.al bring you a fresh and original look at some properties of government spending:
“Big G” typically refers to aggregate government spending on a homogeneous good. In this paper, we open up this construct by analyzing the entire universe of procurement contracts of the US government and establish five facts. First, government spending is granular, that is, it is concentrated in relatively few firms and sectors. Second, relative to private expenditures its composition is biased. Third, procurement contracts are short-lived. Fourth, idiosyncratic variation dominates the fluctuation of spending. Last, government spending is concentrated in sectors with relatively sticky prices. Accounting for these facts within a stylized New Keynesian model offers new insights into the fiscal transmission mechanism: fiscal shocks hardly impact inflation, little crowding out of private expenditure exists, and the multiplier tends to be larger compared to a one-sector benchmark aligning the model with the empirical evidence.
Via the still excellent Kevin Lewis.
Fast Grants update
Fast Grants has now made over 100 grants and contributed over $18 million in funding biomedical research against Covid-19, all in a little over two weeks’ time since project conception. If you scroll down the home page, you can see a partial list of winners (we are more concerned with getting the money out the door than keeping the list fully updated, but it will continue to grow).
Fast Grants is part of Emergent Ventures, a project of the Mercatus Center, George Mason University. And I wish to thank again all of those who have contributed to this project, either financially or otherwise. A partial list of financial contributors can be found at the above link as well.
China fact of the day
The U.S. higher education sector will also be hard hit, with U.S. universities increasingly dependent on tuition from Chinese students. According to the Institute of International Education, China has remained the largest source of international students for ten years running,44 with 369,548 Chinese students enrolled in U.S. higher education programs in 2018 and contributing $15 billion in tuition payments.45 The postponement or cancellation of U.S. college entrance examinations in China, indefinite travel restrictions, and continued uncertainty surrounding when U.S. college campuses will reopen are expected to reduce Chinese demand for U.S. higher education in the 2020-2021 academic year.46 University administrators report that cancelled recruitment events in China and inability to work with local recruitment agencies could further depress Chinese student enrollment in U.S. university programs.
Here is the full document, on cascading economic impacts from China more generally. For the pointer I thank a loyal MR reader.
People are dying from coronavirus because research is too slow
For years, there’s been talk about making the clinical trial process more standardized, and cheaper, so that the same rules would apply each time a study needed to be run. There’s even been discussion that what are known as pragmatic trials — large, simple, randomized studies in which less data are collected — might be conducted using electronic health records. But that hasn’t happened at the pace it should.
The reason involves another part of the problem. Clinical trials are principally run by drug and medical device companies in order to obtain regulatory approvals, with public health authorities only picking up the slack in rare examples. But the result is that we have not built a system that would make studies simpler; most patients have little opportunity to participate in research; and we are too slow to figure out what works.
What would the system look like if we fixed it? It would make it easier to study drugs for heart disease, where studies are so large and expensive that many companies don’t test their medicines. It would ease studies for rare cancers, which are currently problematic because the right patients are hard to find. And it could create a medical information superhighway that would power health care through the next century.
That is from Matthew Herper in StatNews. Via Malinga Fernando.
Are humans constantly but subconsciously smelling themselves?
Here is the opening of a lengthy abstract of a new paper by Ofer Perl, et.al., and it may help explain why it is so hard to avoid touching your face:
All primates, including humans, engage in self-face-touching at very high frequency. The functional purpose or antecedents of this behaviour remain unclear. In this hybrid review, we put forth the hypothesis that self-face-touching subserves self-smelling. We first review data implying that humans touch their faces at very high frequency. We then detail evidence from the one study that implicated an olfactory origin for this behaviour: This evidence consists of significantly increased nasal inhalation concurrent with self-face-touching, and predictable increases or decreases in self-face-touching as a function of subliminal odourant tainting. Although we speculate that self-smelling through self-face-touching is largely an unconscious act, we note that in addition, humans also consciously smell themselves at high frequency.
File under Questions that are Rarely Asked, via Michelle Dawson.
It is better to do lots of tests, even if they are not entirely accurate
We find that the number of daily tests carried out is much more important than their sensitivity, for the success of a case-isolation based strategy.
Our results are based on a Susceptible-Exposed-Infectious-Recovered (SEIR) model, which is age-, testing-, quarantine- and hospitalisation-aware. This model has a number of parameters which we estimate from best-available UK data. We run the model with variations of these parameters – each of which represents a possible present state of circumstances in the UK – in order to test the robustness of our conclusion.
We implemented and investigated a number of potential exit strategies, focusing primarily on the effects of virus-testing based case isolation.
The implementation of our model is flexible and extensively commented, allowing us and others to investigate new policy ideas in a timely manner; we next aim to investigate the optimal use of the highly imperfect antibody tests that the United Kingdom already possesses in large numbers.
There is much more at the link, including the model, results, and source code. That is from a team led by Gergo Bohner and also Gaurav Venkataraman, Gaurav being a previous Emergent Ventures winner.
Quarantining at Work
The Washington Post has good piece on one factory practicing an idea I mentioned a few weeks ago in my post on safety protocols, quarantining at work:
For 28 days, they did not leave — sleeping and working all in one place.
In what they called a “live-in” at the factory, the undertaking was just one example of the endless ways that Americans in every industry have uniquely contributed to fighting coronavirus. The 43 men went home Sunday after each working 12-hour shifts all day and night for a month straight, producing tens of millions of pounds of the raw materials that will end up in face masks and surgical gowns worn on the front lines of the pandemic.
…Nikolich said the plants decided to launch the live-ins so employees could avoid having to worry about catching the virus while constantly traveling to and from work, and so the staff at the factory could be closed off to nonessential personnel.
The article also indicates why price increases are critical to increase supply:
They were paid for all 24 hours each day, with a built-in wage increase for both working hours and off time, the company said. It did not disclose the specific percentages.
Hat tip: Jonathan Meer.
Friday assorted links
1. What do we know about superspreader events? And indoor transmission in China. And what Arnold Kling has come to believe.
2. “We show irradiance and in particular solar zenith angle in combination with cloudopacity explain COVID-19 morbidity and mortality growth better than temperature.” Interesting, though still more interpretation is needed there.
3. Covid-19 in Haiti. And what is Wuhan like right now?
4. In Germany, they consult humanities scholars about how to end the lockdown. And from a French philosopher. And we need blogs back for the pandemic.
5. “Southern New Hampshire University, known for being on the cutting edge of collegiate learning, plans to slash tuition for incoming freshmen as it drastically revamps how it conducts on-campus learning beginning in the fall.As part of the changes, tuition will be cut 61%, from $31,000 to $10,000 starting in the 2021-2022 academic year.” Link here.
6. No strong statistical evidence for the BCG vaccine claim.
7. Is the internet economy going to crash as the real economy shrinks? Several interesting points in that one.
8. de Rugy and Kling on government-backed lines of credit for small business.
9. Bundled insurance markets in everything: “COVID-19 insurance comes free with food delivery in Hong Kong now.”
10. Thread on the meaning of the new NY results.
11. Toward a theory of Tyrone.
12. Test different recovery levers. And Zeynep speaks sense.
How things are, in a few short words
If we keep the economy closed at current levels, it will continue to decay, and at some point turn into irreversible, non-linear damage. No one knows when, or how to model the course of that process. That decay also will eat into our future public health capacities, and perhaps boost hunger and poverty around the world.
If we keep people locked up at current levels, fewer of them will be exposed to the virus, and in the meantime we can develop better treatments, and also improve test and trace capabilities. No one knows how quickly those improvements will come, or how to model the course of that process, or how much net good they will do.
The relative pace of those two processes should determine our best course of action. No one knows the relative pace of either of those two processes. Yet commentators pretend to be increasingly knowledgeable, moralizing based on the pretense of knowledge they do not have.
That is where we are at! And here is my earlier post Where We Stand.
The essential Ben Thompson on isolation and quarantine
…while I have written about Taiwan’s use of cellphone-enforced quarantines for recent travelers and close contacts of those infected, I should also note that every single positive infection — symptomatic or not — is isolated away from their home and family. That is also the case in South Korea, and while it was the case for Singaporean citizens, it was not the case for migrant workers, which is a major reason why the virus has exploded in recent weeks.
Here’s the thing, though: isolating people is hard. It would be very controversial. It would require overbearing police powers that people in the West are intrinsically allergic to. Politicians that instituted such a policy would be very unpopular. It is so much easier to let tech companies build a potential magic bullet, and then demand they let government use it; most people wouldn’t know or wouldn’t care, which appears to matter more than whether or not the approach would actually work (or, to put it another way, it appears that the French government sees privacy as a club with which to beat tech companies, not a non-negotiable principle their citizens demand).
So that is why I have changed my mind: Western governments are not willing to take actions that we know work because it would be unpopular and controversial (indeed, the fact that central quarantine is so clearly a violation of liberties is arguably a benefit, because there is no way people would tolerate it once the crisis is over). And, on the flipside, that makes digital surveillance too dangerous to build. Politicians would rather leverage tech companies to violate liberty on the sly, and tech companies, once they have the capability, are all too willing to offload the responsibility of using it wisely to whatever government entity is willing to give them cover. There just isn’t much evidence that either side is willing to make hard choices.
That is from Ben’s Stratechery email newsletter, gated but you can pay to get it. There is currently the risk that “test and trace” becomes for the Left what “chloroquine” has been for Trump and parts of the political right — namely a way to make otherwise unpalatable plans sound as if they have hope for more than “develop herd immunity and bankrupt the economy in the process.”
To be clear, I fully favor “test and trace,” and I’ve worked hard to help fund some of it. That said, I wonder if we will anytime soon reach the point where it is a game changer. So when people argue we should not reopen the economy until “test and trace” is in place, I increasingly see that as a kind of emotive declaration that others do not care enough about human lives (possibly true!), rather than an actual piece of advice.
Covid-19 liability reform for the eventual reopening
That is the topic of my latest Bloomberg column, here is one excerpt:
If an infected but asymptomatic worker shows up at work and sickens coworkers, for example, should the employer be liable? The answer is far from obvious. Liability exists not to shift unmanageable risk, but rather to induce management to take possible and prudent measures of precaution.
Another problem with liability law in this context is that the potential damages are high relative to the capitalization of most businesses. Covid-19 cases often pop up in chains; there have been many cases from a single conference, or in a single church choir, or on a single cruise ship. If a business or school is host to such a chain, it could be wiped out financially by lawsuits. In these cases the liability penalties do not have their intended deterrent effect because the money to lose simply isn’t there…
Another problem with liability in this setting has to do with jury expertise. Are random members of the public really the best people to determine acceptable levels of Covid-19 risk and appropriate employer precautions? Juries are better suited for more conventional applications of liability law, such as when the handyman fixing your roof falls off your rickety ladder. Given the unprecedented nature of the current situation, many Covid-19 risk questions require experts.
Finally, there is the issue of testing. Businesses could be of immeasurable help by testing their employees for Covid-19, as additional testing can help limit the spread of the virus (if only by indicating which workers should stay home or get treatment). Yet the available tests are highly imperfect, especially with false negatives. If businesses are liable for incorrect test results, and their possible practical implications, then business will likely not perform any tests at all, to the detriment of virtually everybody.
I recommend modest liability for some sectors, and zero liability, bundled with a New Zealand-like accident compensation system, for other sectors. And of course some very dangerous sectors should not be allowed to reopen at all, though I am more sympathetic to regional experimentation than are some people on Twitter.
Nursing home estimates of the day
From the (since updated) report: “In the remaining 5 countries for which we have official data (Belgium, Canada, France, Ireland and Norway), and where the number of total deaths ranges from 136 to 17,167, the % of COVID-related deaths in care homes ranges from 49% to 64%).”
“Residents of long-term care facilities account for half of the total deaths in Massachusetts”
Those are all from an email from Michael A. Alcorn.
Thursday assorted links
1. Why is the Eastern European response better? (WSJ) And how are Swedish hospitals doing?
2. Further doubts on the LOA and Santa Clara serology stories, it now seems they really do not establish any particular results.
3. Why a vaccine will be tough, a depressing thread. And are China’s early patients shedding coronavirus?
4. Ezra Klein on why we can’t build things.
5. Why two decades of pandemic planning failed.
6. New Joshua Gans book, Economics in the Age of Covid-19, MIT Press.
7. Hollis Robbins on why some saw this coming before others.
8. New Becker-Friedman Center podcast on Pandemic Economics.
9. Various forms of presenting state-level data. What exactly is going on with Ohio?
10. Are we prepping for vaccine state capacity?
11. Department of Why Not?: “Former Labradoodle breeder tapped to lead U.S. pandemic task force.”
12. The couple that meets on the German-Danish border during lockdown (NYT).
13. The Fed and saving cities (NYT).