The question: What’s the difference between Italy and the rest of the Western world?
The FT provides a graph:
Here are a few misconceptions I see:
1. It’s not that bad because the death rate is only around 1%, and it mostly kills old people.
2. It’s not that bad because it’s likely to fade away once the weather gets warmer.
The “misconception” part of each claim is the “It’s not that bad because”, not what comes later. Wuhan’s medical system was overwhelmed by the disease, despite less that one percent of its population becoming infected (maybe a bit more unreported, but at most a few percent.) Some experts believe that half of the entire population would be infected without social distancing. Italy’s medical system was also stressed, especially in Lombardy, despite a much lower infection rate than in Wuhan. If it truly became widespread, medical systems might be unable to handle the caseload. As for only one percent dying, if 50% of Americans become infected that might be somewhere around 1.6 million deaths. I certainly don’t think that will happen, but that raises the question of why won’t it happen?
The second claim sounds reassuring, but at a 33% daily growth rate we’d be overwhelmed long before we reached summer, (and we don’t even know yet whether the virus will slow down in the summer.) Again, I don’t expect the 33% growth rate to continue, but that raises the question of why won’t it continue?
We are already seeing quite a few shutdowns of universities, large social events, travel, etc., despite a relatively low rate of infection in the US, thus far. That tells me that when the highly risk averse US public becomes fully aware of the severity of the crisis, there will be much stronger measures taken, much more social distancing.
This is a frustrating problem because it’s not obvious to me that there are any attractive solutions. There might have been if the US government had had a plan for dealing with pandemics, as did the (quite small) governments in Singapore, Taiwan, and Hong Kong. But we did not. Nor did the Europeans. It’s like the old joke about the farmer who told the motorist, “First of all, if I were going there I wouldn’t start from here.” It’s too late for the Taiwan solution.
Instead we seem to be faced with two unpleasant choices:
1. Lots of deaths and lots of overburdened medical systems.
2. Lots of social distancing, and perhaps a recession.
Because I believe the second option is far more likely, we need to focus our attention on that outcome. In my view, the stock market is already assuming that outcome, and also assuming that this outcome will last for longer than a few months. Even if the caseload falls back in the summer months, it’ll spike again in the fall and winter. (Given that the caseload in Florida is rising rapidly, there’s no reason to assume it will go away entirely in the summer.)
So I expect social distancing to last until there is a vaccine, and perhaps a few months after that. We may face 18 months of greatly reduced social interaction.
Is there a more optimistic scenario? The caseload in China is down to very low levels. The main risk in China right now (outside Wuhan) is from international travelers coming into China. And even they get quarantined for 14 days. The Chinese are now restarting their economy. The optimistic outcome would occur if China were able to restart their economy without cases spiking upwards. So if we were able to achieve Chinese levels of social distancing and reduce the outbreak to extremely low levels (a big if) and if we could then restart our economy without a resurgence in the epidemic (another big if) we could get a much better outcome than the 12 to 18 month plague that I’ve described.
But can we?
PS. If you think my criticism of our pandemic preparation is just Monday morning quarterbacking, consider that Tyler Cowen (who is not known as an especially negative complainer) made this observation on February 23:
Here is further information about the obstacles facing the rollout of testing. And read here from a Harvard professor of epidemiology, and here. Clicking around and reading I have found this a difficult matter to get to the bottom of. Nonetheless no one disputes that America is not conducting many tests, and is not in a good position to scale up those tests rapidly, and some of those obstacles are regulatory. Why oh why are we messing around with this one?
That was three days before the very first case of community transmission in the US. It was more than a month after the epidemic had reached crisis proportions in China. Other countries were doing widespread testing. And it turns out that widespread testing is essential in preventing a wider outbreak of community transmission.
PPS. This NYT article on government incompetence is horrifying:
To repurpose the tests for monitoring the coronavirus, they would need the support of state and federal officials. But nearly everywhere Dr. Chu turned, officials repeatedly rejected the idea, . . . By Feb. 25, Dr. Chu and her colleagues could not bear to wait any longer. They began performing coronavirus tests, without government approval.
What came back confirmed their worst fear. They quickly had a positive test from a local teenager with no recent travel history. The coronavirus had already established itself on American soil without anybody realizing it.
“It must have been here this entire time,” Dr. Chu recalled thinking with dread. “It’s just everywhere already.”
Read the whole thing. And then weep.
READER COMMENTS
Rebes
Mar 11 2020 at 4:11pm
Never thought I would say that in my lifetime: the Italian government seems way more competent than the US government.
Alan Goldhammer
Mar 11 2020 at 5:15pm
I’ve stated this several times before. We cannot ‘fully’ understand what is going on without good epidemiology numbers. Nobody knows what the real mortality rate from this is as the cases are still small compared to the seasonal flu. Testing needs to be ramped up as fast as possible. We need a lot of data by age, sex, and co-morbities. It’s still striking to me that those <20 years of age seem to be unaffected in terms of severity of illness.
In our area, suburban DC, things are now being cancelled in terms of large and in some cases not so large gatherings.
Let’s hope that the remesdivir (Gilead) clinical trial has a positive outcome. I also believe Pfizer has a drug against a different viral target. Regeneron has monoclonals against the virus and should be ramping up production but those will have to be infused.
robc
Mar 11 2020 at 6:24pm
Do we know the numbers for the flu? How many people are undiagnosed every year?
Alan Goldhammer
Mar 11 2020 at 7:24pm
Yes, there is an extensive flu detection network.
Scott Sumner
Mar 11 2020 at 7:44pm
The basic points I make here would be valid under any plausible mortality rate (say 0.5% to 2.0%)
DeservingPorcupine
Mar 11 2020 at 5:28pm
Not saying we didn’t screw up or that the net result won’t be bad, but it seems quite hard to justify a mortality estimate of 1%. Taking S. Korea as the place with the “best” denominator, I think .6% is a ceiling.
Scott Sumner
Mar 11 2020 at 7:43pm
That number may turn out to be correct, but you can’t get that from South Korea. The latest figures show deaths there are 0.77% of cases. Cases will rise, but so will deaths. The cruise ship death rate was 1% (albeit they may have been more elderly than average.) But I recall the Korean sample was younger than average.
Also, the death rate would rise if hospitals were overwhelmed with patients. Certainly if half of Americans got the virus then death rates would rise, as most would not have access to ventilators, etc.
Aleksander
Mar 12 2020 at 1:41pm
We actually have some reason to believe the death rate is increasing in Italy right now, as number of deaths is increasing quadratically, rather that exponentially as you would expect from normal spreading patterns.
Matthias Görgens
Mar 13 2020 at 4:10am
There’s lot of talking about getting the right denominator.
But a bigger deal to me seems to be controlling for age distribution and smoking. Those two variables differ a lot between countries.
Aleksander
Mar 14 2020 at 9:06am
Number of confirmed infections also tends to lag behind A LOT, as governments test more and more the more people die. And it is biased in all other kinds of ways. If you estimate infections by using deaths/day, you will get a very accurate number up until the point where a country starts taking heavy measures; after that, the real number may be lower, to the extent that the measures are effective. But again, in most cases we have every reason to believe that official “confirmed infections” numbers are also incredibly and ridiculously low in almost every single country.
Alexander Turok
Mar 12 2020 at 9:26pm
The denominator for South Korea counts a lot of people who are currently afflicted and could die in the future.
Thaomas
Mar 11 2020 at 9:32pm
Agree conceptually with the alternatives but wonder how much good social distancing/self quarantining/event cancellations will do. What’s the marginal benefit? To what extent can we avoid reactions with negative NPV’s?
As for the recession, the Fed’s job (and it will fall almost exclusively to the Fed) is to keep the supply shock of deaths and days lost from work from knocking NGDP (aggregate demand) growth off target. [See Money Illusion “Do it now!”]
All “fiscal policy” can do is things that need doing anyway: tax credits to firms for days of sick/family leave granted, expand Medicaid in the states that have not already done so, re-open ACA enrollment, remove work requirement for SNAP and make SNAP, FICA holiday, and unemployment benefits dependent on the state of the economy.
Mark Bahner
Mar 12 2020 at 11:06pm
My “fiscal policy” would include at least two huge (multi-billion-dollar) technology prizes for inexpensive and rapidly deployed testing (e.g.,”$5 billion for the first test that is proven to have x specificity and y sensitivity in 1 million tests”) and rapidly deployed vaccines (e.g., “$5 billion for a vaccine that first meets x protectiveness and less than y percent significant adverse side effects in 1 million people”).
Matthias Görgens
Mar 13 2020 at 4:18am
The simpler first policy for testing is to remove the silly bans on non-CDC tests.
The prizes you suggest might be a good idea. But more important than government activities, is first to stop standing in the way.
Mark Bahner
Mar 13 2020 at 1:16pm
Yes, absolutely.
As I thought about this after I made the comments, I realized that what economists could do (a friendly hint to the economists who post and comment here on Econlog ;-)) is to answer the question, “What is economic damage over time of not having a widely available and effective vaccine?” I think the answer would be so incredibly high that there should be a massive worldwide coordinated effort (coordinated, not necessarily non-competitive) to develop a vaccine in months or even weeks.
Just look at what has happened with NCAA and pro sports events here in the U.S. We’re probably talking about billions of dollars in lost revenue right there. I guess an even better way to view it would be to look at the S&P being off by…about 25%? (it’s so bad I hate to even look anymore)…since the start of the virus.
We’re really talking about this virus probably costing trillions of dollars in damage, globally. So why isn’t the globe coordinating to spend at least tens of billions of dollars just in the next month to bring a vaccine to market much earlier than the 12-18 month value I’ve seen?
What would it cost to bring the vaccine to market in 2 months? 4 months? 8 months? 12-18 months? If we drew a cost curve and a damage curve (damage over 2 months, 4 months, 8 months, and 12-18 months) my guess is that the two curves would cross at very close to the 2 month mark, indicating that, globally, humanity should be spending much more money right now on developing a vaccine.
Phil H
Mar 11 2020 at 10:07pm
I’m going to stick to my guns on saying: you can’t judge government responses till they’re over.
In the early days of this outbreak, here in China, we heard nother but how badly it had been messed up. In China and outside, that was the focus: boy, those Chinese (or sometimes Wuhan) officials really got this wrong, Chinese government incompetence. And there was a *lot* of evidence for this – Li Wenliang being the most obvious symble.
Now, we’re all woah! The Chinese government really sorted things out. Within China, everyone is marveling at how rubbish the US/European response is, and realising that messed up as their government is, everyone else’s is the same.
To Scott and other libertarians, I would suggest this: the US government moving slowly isn’t a bad thing. I know slowness looks bad when events move fast, but this slowness is *part of the mechanism that stops government being tyrannical*. It’s a positive. To everyone else, I just want to suggest this: What if everything we’re seeing is the result of 100% natural progression. There is a strong chance that all of the different human interventions across the world are making, in sum, zero difference to this epidemic.
Singapore and Taiwan? Small islands. Exactly the kind of place you would expect to escape lightly. Large land masses like the US and China? Much more susceptible. Africa? It’s hot, less virus. China’s “win”? It’s pure S-curve.
Rather than responding to day-by-day news, the way to think about this is, what would be happening if nothing were being done?
(Example: where I live, Xiamen, we have a city of 5 million people on lockdown. The whole city has been sprayed with chlorine. And there was never any covid here. We’ve had 30 cases, all of them people from elsewhere. Incidentally, we’re also a small island.)
Scott Sumner
Mar 11 2020 at 10:18pm
There’s nothing inconsistent about saying the Wuhan government messed up badly in January and the national government did better in February. It happens to be true.
Phil H
Mar 12 2020 at 2:12am
I know what you mean, but I regard that position as classic Zhongnanhai kool-aid. It is exactly the narrative that the Chinese state wants you to believe: sure, the localities mess up sometimes, and when they do, you can trust Daddy Xi to step in and sort it all out.
Whatever incentives led Wuhan to mess it up in the first place were put there by the central government. Even the individuals who messed it up in Wuhan were put their by the central government – did you see this comic detail: the guy who was in charge of the Wuhan Red Cross, which apparently embezzled half the donations sent by a concerned nation, was the Tianjin mayor who got fired when they had the biggest oil depository explosion in the world. He’d been parachuted into a nice little sinecure…
And like I say, how good have the national policies been? The virus seems to have burnt itself out… but at what cost? And did the national precautions actually make the difference.
What you’re saying isn’t impossible, but when it fits the CCP’s narrative so perfectly, I’m going to demand a higher bar of truth before I start believing it.
Mark Z
Mar 11 2020 at 10:30pm
I don’t think being an island is really that big of a deal in an age of airplanes. Once the virus arrives, and it’s arrived on those islands, most of the spread is from within. After all, the Pacific Ocean didn’t do much to protect the US from it, it’s doubtful the Singapore Strait makes a world of difference.
Phil H
Mar 12 2020 at 2:30am
I hear you, but that sounds to me like a triumph of logic over facts. Hainan, Singapore, Taiwan: three islands within the China sphere, different governments, different regimes, none of them badly hit. The UK, so far as I can tell, doing better than mainland Europe. Do you have figures for Malta or Cyprus?
Mark Z
Mar 12 2020 at 2:45am
The UK’s trajectory in the plot doesn’t really stand out among European countries, just not as far along yet. Modestly slower than average, but not the slowest. So again, it looks to me like being an island maybe buys you some time, but doesn’t seriously change the trajectory of the epidemic. Iceland and Ireland may tell us soon enough.
bill
Mar 12 2020 at 1:59pm
How about the Faroe Islands? 2 cases?
Phil H
Mar 22 2020 at 9:18pm
A week or so later, if we look at the four large countries in Europe with comparable populations, onfirmed infections stand at: Italy 60k, Germany 25k, France 16k, UK 6k. The UK’s response has been notably rubbish, with PM Boris Johnson flip-flopping, and no coordination across schools. So I don’t think the UK has a superior record there. And yet our numbers are still low. Ireland, 906 cases.
Caveat: the numbers may just be wrong, an artifact of testing rather than reflecting the real situation.
But until we know more, I’m sticking to my basic thesis. Islands get it easier. This is the knowledge of epidemiology that several posts on Econlog have been talking about in recent weeks.
Matthias Görgens
Mar 13 2020 at 4:21am
Singapore might be a small island geographically (just about, you can walk over to mainland Asia on foot), but socialogically it’s plugged into the main power spine of the world economy.
Far less of an ‘island’ than our closest mainland neighbour.
Michael Sandifer
Mar 11 2020 at 11:53pm
It shouldn’t be forgotten that the US has a rather high obesity rate, which will mean higher mortality, ceteris paribus.
Aleksander
Mar 12 2020 at 1:46pm
I strongly recommend using graphs with numbers of deaths/day, which is likely a much better proxy for number of infections (three weeks ago, but possible to extrapolate to today) than number of “confirmed” infections. For instance, Japan clearly has fewer infections now than before, as people have stopped dying; but the “confirmed infections” numbers don’t show that.
Matthias Görgens
Mar 13 2020 at 4:22am
That might be true, but it’s also lagging the action by at least two weeks.
dcpi
Mar 12 2020 at 8:56pm
The chart would be much more compelling if the number of cases were per capita and not absolute.
Matthias Görgens
Mar 13 2020 at 4:26am
Maybe. But to see anything interesting, you’d have to split the graph by metro area or so, not by country.
Eg the per capita load in Wuhan or its province is more interesting and comparable to Lombardi’s load than comparing per capita rates in China to Italy. At least at this stage of the spread of the virus.
Henry Morgan
Mar 13 2020 at 12:40am
I highly suggest the use of death / day graphs, which are probably a much better proxy for the number of infections (three weeks ago but can be extrapolated to today) than the number of “confirmed” infections. For example, Japan now obviously has less infections than before, since people have stopped dying; but the statistics of “confirmed infections” do not reflect this.
Hazel Meade
Mar 13 2020 at 1:00pm
The thing is, if it’s been here all along and nobody noticed, how bad can it really be?
Maybe the death rate isn’t much worse than the seasonal flu and people have been just assuming it’s the seasonal flu. Especially if nobody actually had been doing testing, how would anyone know if there’s a new virus going around?
John
Mar 13 2020 at 1:08pm
The entire point of Scott’s post is that exponential growth means things go from “nobody noticed” to “the hospital is overwhelmed and exhausted urologists are deciding which patient gets to survive” in just a few weeks. Right now we’re in “nobody noticed” land (or “very few people notice” land), but our trajectory doesn’t look good.
Mark Bahner
Mar 15 2020 at 9:42pm
If you look at in terms of cases growing exponentially at the rate of 30 percent per day, the answer is: “It could get really, really bad. In fact, it could get much worse than that.”
Day 1 = 100 new exposures
Day 2 = 130 new exposures
Day 3 = 169 new exposures…
…doesn’t seem like a big deal. But then:
Day 30 = 2,137,211 new exposures
Comments are closed.