My general view is that social distancing is better than an explosion of coronavirus cases. I believe we were too slow to begin social distancing, at least in hindsight. At the same time, I expect that after the worst phase of the epidemic is over we’ll do too much social distancing.
Suppose you saw a sign in a restaurant window saying, “Our restaurant has a smaller than average number of rats and cockroaches in our kitchen.” Does that make you more inclined to eat there? Obviously it should make you more inclined to eat there, but would it? Many people don’t even give any thought to rats and cockroaches in the kitchen when deciding where to eat. They’d be put off by that sign. Gross!
The Chinese movie theatre industry was about to re-open last weekend, and then the government suddenly reversed course. This picture of a Chinese movie theatre caught my attention:
Those barriers that keep people separated obviously make the theatre safer. But I wonder if they make it seem less safe to the average filmgoer. I can picture my wife saying, “Well, if it’s so dangerous they need to separate people, then perhaps I’ll just stay home and watch Netflix.”
Our society doesn’t react very rapidly to hard to understand risks discussed by experts. But when the risk becomes obvious, we tend to overreact. In a previous post, I argued that overreaction (from an individual risk perspective) was socially optimal in this case. But that Chinese movie theatre picture (and the likely reaction of theatre patrons) makes me think that eventually we’ll go too far. It would be optimal at some point to reboot our economy if we could keep the R0 factor (virus reproduction) below 1.0. Do lots of little things like what we see in the picture above. But I wonder if our extreme risk aversion will demand an even higher standard, closer to zero risk.
READER COMMENTS
Alan Goldhammer
Apr 3 2020 at 3:19pm
As Paul Romer said, we need massive testing to get people back to work. The main policy makers as is evidenced by the daily briefings are not thinking outside the box. A quick and accurate blood test can show the true number of infected and recovered individuals (antibodies to SARS-CoV-2). Only ONE test has received an Emergency Use Authorization from the FDA. Romer wants to test 15M people day so that everyone gets tested at two week intervals. I think that number is too high but as we find those who have recovered, they don’t need to be tested any more.
Most of the modeling papers I’ve read come up with an infection rate 10 times the reported figure. A priori, this means that there are lots of recovered people cooped up in their homes and not productive (a bit of humor – anyone who writes or responds to a blog is by definition, productive).
This is my big problem with the response to this pandemic.
Scott Sumner
Apr 3 2020 at 9:06pm
I completely agree about testing. I wish I had been tested when I was sick in February.
Thaomas
Apr 3 2020 at 4:29pm
The question is, will people be too risk averse with their supply of labor and other factors of production? If they are just too averse with their demand, – consumption and investment (demand too many safe assets) — then the Fed can (and should, right?) just sate the safe asset demand and keep NGDP on target.
Scott Sumner
Apr 3 2020 at 9:07pm
Supply will partly depend on the unemployment insurance program.
Thaomas
Apr 4 2020 at 7:01am
What proportion of slow recovery after 2008 was due to unemployment insurance and how much to the Fed failing to keep NGDP or even the price level growing at the pre 2008 trends?
Scott Sumner
Apr 4 2020 at 12:47pm
Most was due to NGDP, but that fact has no bearing on today, as the UI program is now far more generous, with far stronger work disincentives. That’s not a problem at the moment, but could become one when jobs are again available.
Thaomas
Apr 4 2020 at 10:36pm
I’ll believe it (macro economically important, not statistically detectable) when we see the Fed achieving 4-5% NGDP growth with slow real growth and high unemployment.
Phil H
Apr 3 2020 at 9:32pm
“Suppose you saw a sign in a restaurant window saying, “Our restaurant has a smaller than average number of rats and cockroaches in our kitchen.” Does that make you more inclined to eat there? Obviously it should make you more inclined to eat there,”
I think a mild protest at this example is warranted. That conclusion is not obvious at all, because it assumes that the sign is accurate. If you don’t assume that the sign is accurate, and you put together a fairly reasonable set of assumptions about how that sign came to be there, you can easily build a case for avoiding that restaurant.
A lot of examples of supposedly bad logical thinking break down on points like this. They demand that at some point, people accept some piece of data as true and accurate. But of course, in the world, there is no data that we uncritically accept as true and accurate. Under those conditions, it’s far from clear to me that our risk aversion (which I agree exists) is not rational.
Scott Sumner
Apr 4 2020 at 12:44pm
OK, an “accurate sign that says . . . ” I think you know the point I was trying to make.
Robert EV
Apr 5 2020 at 1:38pm
Oh mean, median, and mode.
A numerate person would seriously consider avoiding that place.
Phil H
Apr 6 2020 at 1:38am
Scott: I do know the point, and I really, really disagree with it! I don’t know if you know this comic Chinese fable: A man has some treasure to hide, and needs to remember where it is; but he also doesn’t want anyone else to find the treasure. So he erects a sign where he’s buried it saying: There are not 300 taels of silver in the ground here (此地无银三百两). Just as prices are both exchange rates and signals, any communication is both a signal of its content, and a signal of the processes that brought it to be there (symbol and index, in the language of Peirce; face-value interpretation and Straussian interpretation might be how Tyler Cowen would put it).
BB
Apr 4 2020 at 2:07am
Not sure if it’s 100% accurate (the math seems to make sense though), but I saw a discouraging article today that getting R0 below 1 may not be nearly enough:
http://www.overcomingbias.com/2020/04/beware-r0-variance.html
Jens
Apr 4 2020 at 7:26am
Most important post. People either ignore/forget about risks or want 100% security. Probably due to what Richard Dawkins calls the tyranny of the discontinuous mind. Or our limited attention span. Reducing the transmission risk in certain situations “a good portion” and reducing the number of these situations “a good portion” is absolutely sufficient and cost efficient to let the virus run short. Total lockdown, total security is ridiculously expensive. Furthermore improve monitoring, testing and modelling in summer to prepare for autumn/winter. The police should step a bit on the feet of wannabe superspreaders. Vaccine and/or efficient treatment will arrive sooner or later. It’s not that hard after all. Just stop being blind.
Thomas Sewell
Apr 4 2020 at 8:28pm
The isolated seats are counter-productive in another way.
Most people who live together will also want to see a movie together. There is no reasonable additional risk to them of sitting together watching a movie as well. It makes more sense to isolate family groups from one another, but you’d have to rely on individuals to self-regulate for that to happen.
Comments are closed.