My friend and fellow economist Ben Powell, head of the Free Market Institute at Texas Tech University, wrote on Facebook this morning:
Like many people, I’ve been following the various COVID numbers with interest and have become aware of the various data problems with many of them. However, it is becoming obvious to me that as talk turns to “re-opening” the economy most of these numbers aren’t relevant, yet people are selectively using them to support their arguments. The reason for the shutdown was not simply to “flatten the curve” for its own sake but to flatten it so that hospitals would not be overwhelmed. It was not to decrease the total number of people who get the disease. It was to spread out the cases to prevent hospitals from being overwhelmed so that there would not be deaths from COVID because of lack of available medical care (regardless of what the death rate from COVID is otherwise). It was a particular cause of death (lack of hospital capacity) that the shutdown was to avoid. So, the relevant data for re-opening is hospital capacity. That’s it. And we can get reliable data for that in the US and it varies by region. But most of the country has plenty of hospital capacity and in fact, due to bans on non-essential medical proceedures, some hospitals have so much excess capacity that they are hurting financially and laying off workers. Other data like cases and case fatality are interesting and (poorly) inform other decisions such as whether and how much I choose to social distance. But it’s not relevant for the re-opening debate.
This is a case where, within a month, the goalposts have been moved and not just on a small issue but on what will likely be the biggest economic/regulatory issue of the decade and could be the biggest economic/regulatory issue in the lives of many of our readers.
I wish I had made Ben’s point in my recent article on liberating the economy.
Now that the curve seems at or close to flattening in the relevant sense, that is, cases that require hospitalization relative to hospital capacity, I would hope that those whose rationale was to “flatten the curve” would call for ending the lockdowns, certainly in much of the country. I haven’t seen it yet.
READER COMMENTS
Gary Short
Apr 15 2020 at 1:53pm
Great point, David (and Ben). But, the main argument for flattening the curve was not limited to a shortage of hospital beds, but also a shortage of ventilators. While I don’t know the current ventilator numbers, your analysis should apply equally to ventilators (at least in most parts of the country).
Mark Bahner
Apr 15 2020 at 4:17pm
If we’re going to base ending “stay at home” on whether ventilators are available, it would be good to consider: a) whether ventilators are even the best treatment option, and b) even if they are, how much better are they than other treatment options (for example bi-level positive airway pressure machines plus oxygen)?
I have no medical background, but it appears to me from my reading as a layperson, that the answers to both “a” and “b” aren’t really clear:
Some doctors move away from ventilators for COVID patients
Daniel
Apr 15 2020 at 1:54pm
That is a really good point!
But what it’s missing is that as we re-open, the curve will not stay as flat as it is now, our monitoring/testing is not sufficient to provide information with a short-enough lag, and our masks are still limited. Re-open too early, and we’ll find out we are behind the curve again, forced to shutdown again. We at least have bought time to build up testing and prophylactic capacity, so it’s not like this would be an immediate return to the pre-shutdown exponential curve, but it’s definitely important to realize that the curve we have transitioned to is not the permanent state of the world and we should abandon it (i.e., re-open) with care.
Phil H
Apr 15 2020 at 2:01pm
This is right, though it’s not the only strategy. There may still be time to stop tens of thousands more people dying, with a better lockdown. Americans are absolutely free to follow the pure flatten the curve strategy, but should do so in the knowledge that other countries have done much, much better, and have effectively prevented the outbreak from reaching the entire population.
Art Carden
Apr 15 2020 at 2:06pm
Prediction: COVID shutdown:climate policy as WWI War Industries Board:Roosevelt’s Brain Trust & the New Deal.
Jon Murphy
Apr 15 2020 at 2:06pm
Of course, the underlying assumption here is that the lockdown is actually flattening the curve. That is still unknown. We want to avoid spurious reasoning.
Alan Goldhammer
Apr 15 2020 at 2:24pm
Of course, the underlying assumption here is that the lockdown is actually flattening the curve. That is still unknown. We want to avoid spurious reasoning.
Seriously? This is unquestionably the case and has been shown to be true in a number of countries.
Jon Murphy
Apr 15 2020 at 2:29pm
Yes, seriously. Other countries that didn’t do lockdown to flatten the curve, like Sweden and Korea, have had similar patterns.
robc
Apr 15 2020 at 3:51pm
Yes, there was a paper out of Israel written in a language I dont read and I dont trust the website reporting on it that attempted to measure this and found no difference based on level of lockdown. The virus followed same curve.
robc
Apr 15 2020 at 3:52pm
I have been eyeballing death data by state and the same applies, I cant spot an obvious difference, other than density matters somewhat.
Thaomas
Apr 15 2020 at 2:47pm
Don’t get dragged into a discussion of “re-open”/”no reopen” dichotomy. We should be talking about how and when not yes and no. This is basically a local cost benefit analysis, albeit a very unsatisfactory one as we do not have a good model of the health-economic system. Cochrane and McArdle point in the right direction.
Kevin Erdmann
Apr 15 2020 at 4:20pm
I second this, and Josiah’s point below.
Case growth is declining rapidly. Now, we can test the details of getting as many things back to some functional norm without triggering exponential case growth again. And, since we are flattening the curve with lockdowns, in the cases where normalization overshoots and triggers new case growth, those outbreaks will be far fewer and they will involve a few dozen infections instead of a few thousand.
The most visible activities will need to be normalized more slowly because, by definition, those are activities where many people are in close proximity. But, there is a ton of activity in the background that is most responsible for our daily economic lives. Some of that is closed down now just because overzealous lockdowns in certain locations have created supply chain discontinuities. But other facilities that randomly didn’t suffer from supply chain breakdowns have safely been proceeding through the lockdowns. A lot can be restarted safely.
There should be a large muddy middle of economic activity where a consensus about how to normalize shouldn’t be that hard. This is a case where things are a lot easier in detail than they seem in the abstract.
Jon Murphy
Apr 15 2020 at 8:03pm
I agree. And these things should be tested. I am urging that these tests be conducted, not simply assuming that lockdowns are necessary
AMT
Apr 15 2020 at 9:58pm
This is the article for you!
https://www.bmj.com/content/327/7429/1459?ijkey=c3677213eca83ff6599127794fc58c4e0f6de55a&keytype2=tf_ipsecsha
Alan Goldhammer
Apr 15 2020 at 2:22pm
For those of us who have been tracking this for some time, David’s (and Ben who I don’t know) point is both obvious and requiring quick action in the event hot spots pop up (and they surely will; that’s probably a bet even Bryan Caplan will take that side of). The critical problem is the testing regime is in no way robust enough for a broad reopening. I’ve probably commented on this matter more than anyone so I was surprised to hear Larry Summers say essentially the same thing today.
You cannot link to individual episodes of Noah Feldman’s ‘Deep Background’ podcast (I subscribe via Stitcher) so you will have to go to the website to listen to it. It’s only 30 minutes and Summers opines on the economics of toilet paper among other things. He believes that $80B is lost to the economy for every week that things are closed at the present level. If we just donated a small portion of this to implementing a testing regime that was broad and timely, shortening the shutdown by 3-4 weeks is big $$$$$s. The failure of the initial test development was bad but this failure is more expensive by billions of dollars. I think this is the same point Paul Romer has made.
There are so many incompetent people in charge of things that I’ve lost count.
robc
Apr 15 2020 at 3:54pm
Thomas Massie suggested using 1/2 of 1% of the stimulus to develop and implement widespread testing. I don’t know if his number was right, but it is probably in the right ballpark, so even if he was off by a factor of 10, we are still talking about 5% of the stimulus, which is a nice savings.
Mark Z
Apr 16 2020 at 12:10am
Larry Summers? I did a little ‘back of the envelope’ math and also looked up what’s been published on the economic value of testing, and saw Paul Romer’s tweet storm on it, and it seems undeniable that vastly expanded testing is very much worth it. This paper – by a French economist – tries to compute the value of an individual test (well, applying optimized group testing too) due to the value of ‘deconfining’ workers: https://www.tse-fr.eu/sites/default/files/TSE/documents/doc/by/gollier/group_testing.pdf. The numbers he reaches are pretty impressive, many times the cost of a test even if these turn out to be drastic overestimates. It is nice to see economists converging to a consensus that testing is very valuable.
One response to Romer I saw was that it could take months to expand the production of reagents necessary for the test. His reply was that there are tests in production that use fewer of these reagents, but even if it took months, it’s still easily worth it to do while awaiting vaccine development, still over a year away.
robc
Apr 15 2020 at 3:29pm
I have been making this point for a while now.
Governors need to start loosening restrictions (they dont have to do everything at once) and watching the hospital numbers. When capacity (however the best way to measure it) becomes close to an issue, stop loosening for a bit. If you unwind one week of restrictions per week and nothing bad happens on the hospital front, then we have unwound the last 6 weeks over the next six. They can go faster of slower than that as their relative capacity calls for.
We need to unflatten the curve now to speed up time to herd immunity.
Josiah
Apr 15 2020 at 3:33pm
I can only speak for myself, but my support for the lockdowns was never based on simply spreading out the number of infections (I can provide proof of this if anyone really cares). Instead, I supported the lockdowns because they would let us get the outbreak under control in the short term and give us time to develop a better strategy for ending the outbreak.
Michael
Apr 16 2020 at 6:51am
I think this is correct. A few weeks ago, there seemed to be some confusion (which still exists) between “mitigation”, which aims to spread the impact out so that hospital resources are not overwhelmed, and “suppression”, which aims to drive R0 below 1 to reduce the amount of the virus that is criculating in the population.
Like others, I’m tired of the “when should the lockdowns end?” debate and would like to move on to the “how can we safely end the lockdowns?” debate. Some states, including my home state (MA) seem to be working on this.
MG
Apr 17 2020 at 7:50am
I think someone should compile a granular list of all the facts/views/opinions expressed here. I think that there is would be a lot of consensus for many actions that would constitute progress from paralysis. I think that Ben Powell’s point is not to be interpreted so simplistically — that he is arguing for a complete reversal to the status quo ante. What Michael terms as how can we safely end the lock downs is what everyone is after. And it should be noted that these are decisions that depend on local conditions. And state by state analysis should give us comfort about how selective the current problem is, and that hot spots will have to be managed differently from the broad swath of American geography, demography, and political economy. (I feel that my state of FL has applied the right amount of suppression and economic restriction given its resources and challenges. Obviously New York City — and a few hot spots — did not, for whatever reasons. )
Secondly, I am sure that Ben is aware that continued future suppression is not assured initial suppression. But neither can one ignore that a good deal of the initial suppression involved changing habits and practices — personal, societal, medical — that will continue and will even increase as part of safe re-openings. Keeping landscapers unemployed could not have done nearly as much as restricting access to nursing homes. Additionally, as some have pointed out, our medical ability to flatten the hospitalization, ICU, death curve will only improve — it already has via a much better understanding of the co-morbidities and the way in which different cases should be treated.
Finally, we must not cowardly signal our compassion by avoiding an honest attempt to define what is a tragedy to be avoided at all costs and what is a negative fluctuation in humanity’s relation with adversity and death — which has accommodated much worse suffering than even the worst scenarios allow, that it seldom hits demographics evenly, and that is always deemed most tragic when it comes unannounced (by which I mean, medically).
Craig Richardson
Apr 16 2020 at 9:35am
It’s amazing how stuck we all got on having ventilators (including me) because of the endless discussion that ventilators were in short supply. The experts told us that is what was needed and we took that as a working assumption.
We (at least me) were sort of led to believe a ventilator was like a spare oxygen tank, and this would save people’s lives who would die otherwise. But going on a ventilator is a grim prognosis- between 25-90% of people die anyway, and those who come off are physically damaged.
There is new evidence that Covid-19 works differently and that a ventilator damages the lungs. We were chasing this technology under a deadline but the shortage in areas led to better techniques that cost little to nothing (lying prone, combined with gentle oxygenation). Not blaming anyone here, but it goes to show that working assumptions always need to be questioned.
*************
In terms of opening the economy, I think we should think about waves of groups who have the lowest risk factors- e.g. first people under 40 first- who can first reenter the economy. The risk to them is similar to the flu, I think. That of course must be accompanied by inexpensive and widespread testing.
As they develop herd immunity, then we allow the next wave and so forth. That would require a new way of thinking in terms of civil liberties and enforcement, but I’m willing to take slices rather than none at all as we work towards normalcy again.
Michael
Apr 16 2020 at 11:13am
25% and 90% is quite a range.
BS
Apr 18 2020 at 11:42pm
There are two curves: one for COVID cases requiring hospital care (the one everyone talks about), and one for hospital capacity to handle those cases (the one usually described or drawn as a flat line). The aim, noted by Powell, is to keep the former under the latter. Plots are mostly useful as a way of easily visualizing the relative velocities (ie. whether the case curve is catching up to the capacity curve). The case curve could be allowed to rise, if the capacity curve could also be raised, particularly if we can see that the velocity of the latter is greater than the velocity of the former. Raising the capacity curve is difficult because it depends on resources: buildings, beds, equipment, consumables, people. But the first thing to do is realize that the capacity curve can and should be bent upwards. It can also be bent downwards – supplies run out, workers become incapacitated. If workers are being furloughed, it’s definitely time to allow the case curve to rise. And while most medical providers can not be trained in short order, delegation of authority downwards should produce, at the bottom, a use for “orderlies” who might in fact be trainable in short order.
Comments are closed.