1) Every once a while, when I find (or come up with) something that I think is REALLY important, which I don’t want to get lost in the deluge of information out there (much of it from me! 😉), I write STOP THE PRESSES to get your attention.
Here’s why I’m doing it here:
The number of Americans testing positive for the coronavirus is growing exponentially, roughly doubling every two days for the last two weeks, as you can see in this chart (the trends are similar across much of Europe):
If this rate of growth continues, the number of cases will grow by more than 30 times in the next 10 days, resulting in 750,000 infected Americans. And if it continues for 10 more days after that, we’ll have 23 million cases! Even if we get lucky and have only a 1% mortality rate, that would result in the death of nearly a quarter million of our fellow citizens. In the face of this grim math, it’s no wonder our political leaders are shutting down our economy, irrespective of the cost.
And the cost will be terrible. New York Times columnist Bret Stephens, a Republican, writes in this column, It’s Dangerous to Be Ruled by Fear, that if the shutdown extends for a few months:
…we might face not a recession but a full-blown depression, which would be financially ruinous for hundreds of millions and have its own disastrous knock-on effects in mental, emotional, and physical health, including for the elderly and sick who already face the greatest risks from the virus.”
And if it:
lasts 18 months — the estimate of how long it might take to develop and distribute a safe and effective vaccine?
In that case, we are looking at a century-defining calamity that could bankrupt the government; wreck nearly every business in America, large or small; disrupt supply chains and create critical shortfalls of food, medicines, and other essential items; lead to dramatic increases in deaths of loneliness and despair; tempt political leaders, including the president, to disavow democratic norms in the name of public health (including by seeking to postpone elections); and create widespread, perhaps deadly, civil unrest.
This editorial in the Wall Street Journal, Rethinking the Coronavirus Shutdown, expresses similar concerns.
But what choice do we have? After all, that’s what simple math tells us, right?
Not so fast…
There’s a huge error in this analysis – one that could cost us trillions of dollars if we don’t quickly recognize it and act accordingly.
You see, the chart above (and the accompanying hysterical media coverage) shows the number of positive coronavirus test results, not the number we really care about: how many people are actually infected.
To understand the difference, let’s imagine that an infected person arrives undetected in a particular city and starts an outbreak of the coronavirus. And, hypothetically, let’s assume that it spreads steadily, such that 1,000 people are infected at the end of the month. This is called linear growth, and here’s what it looks like on a chart:
Now, let’s imagine that nobody knew about the virus until the end of the third week, at which point testing ramps up quickly, such that by the end of the month every infected person has been detected. This is what that chart would look like:
This is exponential growth. Much scarier isn’t it? Imagine the headlines…
But, of course, it doesn’t reflect what’s really happening with the spread of the virus.
Now, in reality, the various, limited case studies we have indicate that the rate the coronavirus actually spreads is somewhere in between these two hypothetical examples I’ve given.
Unfortunately, however, it’s nearly impossible to know the exact replication rate (called the R0) because the coronavirus is so new, and the R0 appears to be highly dependent on many factors like population density, age and health of the population, smoking rate, and the degree of social distancing.
So the reality is, until there is more widespread testing, we’re sort of flying blind. Nobody knows how many people are actually infected here in the U.S. right now and how rapidly that number is growing.
So in the absence of this knowledge, those who are making the most extreme predictions are getting the most attention – and nobody can prove them wrong.
But one thing I know for sure is that the rising number of Americans testing positive for the coronavirus tells us only one thing: that there are more people getting tested – which is a good thing! This chart shows the enormous growth in tests over the last 17 days:
Looks similar to the earlier chart, right? So let’s put them on the same chart:
It’s hard to see what’s going on if total tests and positive tests are on the same axis because only a small percent of tests come back positive (consistently between 10-15% for 10 of the last 11 days; 13% overall). So, here they are, using two axes – as you can see, they are virtually identical:
So, when you see headlines that “Coronavirus cases hit new high in [name a city, state, or country]!”, keep in mind that this doesn’t necessarily mean that the number of newly infected people is going up.
In fact, it’s possible that all of the steps we’ve taken have dramatically slowed the spread of the virus, and the number of newly infected people is going down. But in the short-run, we wouldn’t know it because the reported number of new cases is certain to keep rising as our testing ramps up.
Eventually, the testing will catch up with the reality, as my hypothetical example above demonstrates. And the sooner, the better, as we can’t make decisions about what actions to take – and what burdens to bear (and not bear) – when we’re stumbling around in the dark, with no idea what the real numbers are regarding the number of infected people, where they are, and how rapidly the virus is spreading.
2) It’s amazing how many people – smart people who should know better! – don’t understand that the number of positive coronavirus tests has little do do with the numbers that really matter.
The New York Times and the Columbia University researchers it quotes sure don’t, based on this front page story in today’s paper: Coronavirus Could Overwhelm U.S. Without Urgent Action, Estimates Say. This is the panic-inducing image that dominated the front page:
Under the researchers’ best-case scenario:
Even if the country cut its rate of transmission in half — a tall order — some 650,000 people might become infected in the next two months…
New York City, Seattle, Boston and parts of California already have such large outbreaks that they will probably see significant growth even after taking extraordinary measures over the past week, the researchers say. New York City’s outbreak, the nation’s largest, grew to more than 4,000 known cases on Friday and is likely to increase many times over even in a favorable scenario.
On the other hand, parts of the country without large clusters of cases could still avoid the worst of the outbreak — if they impose measures like closing schools, banning mass gatherings and testing and quarantining sick people and their contacts. The epidemic would then spread inland at a much slower pace and strike with less severity, the estimates say.
But controls would need to be put in place immediately, and everywhere…
“You have to think of this as an insurance for the future: The earlier you do it, the greater effect you have on the virus,” said Alessandro Vespignani, director of the Network Science Institute at Northeastern University, who said the estimates were in line with his own projections. “It’s better to take excessive precautions than not.”
This is among the worst research and worst journalism I have ever seen.
To be clear: the three scenarios that the researchers lay out are all possible. We shouldn’t dismiss them, and should take reasonable measures to head off even their best-case scenario, much less their worst-case one.
But there is no real-world evidence whatsoever for any of their scenarios. In fact, nearly all of it is the exact opposite of what they’re modeling.
The best real-world case study is China, where the virus originated in the city of Wuhan in December. The Chinese initially didn’t recognize the problem, and then the government tried to cover it up, so it spread like wildfire initially.
By the time the government realize its mistake and started locking down Wuhan and surrounding cities on January 23rd, 25 people had already died and new cases were growing exponentially, as this chart shows:
But only three weeks later, the strong measures China took began to kick in and the number of new cases began to fall – and only weeks later went to almost zero and have stayed there, as you can see in this chart:
In total, only 81,000 Chinese were infected – less than 1/100th of 1% of their population – and fewer than 3,300 people have died, nearly all of them elderly.
China’s Asian neighbors had even better experiences. Travelers from Wuhan carried the coronavirus to Taiwan, South Korea, Japan, Hong Kong, and Singapore, but each of these countries recognized the risk and took strong countermeasures, focused on testing to identify who’s infected and then taking targeted steps to quarantine and treat those people. All of these countries have effectively stopped the coronavirus in its tracks – and, importantly, did so without shutting down completely and quarantining all of their citizens.
Now, some people think that we’ll never be able to do what these Asian countries have done because neither our government nor our people are taking this seriously enough. They think we’re doomed to follow in the footsteps of Italy, which has seen many of its hospitals overwhelmed, had more deaths than any other country, and is now in a nationwide lockdown.
They point to this chart, which shows the number of new cases by day for the U.S. and Italy, with the U.S. 10 days behind:
Scary, isn’t it? We’re now above Italy’s trajectory!
But the problem with this chart is that the U.S. is a much larger country – our population, at 331 million, is 5.5 times larger than Italy’s 60 million. If we adjust for cases per million people, the chart looks different – we’re still on a similar trajectory, but are 16 days behind:
Those extra six days may not sound like much, but when the number of people infected by the virus, left unchecked, doubles every few days, it’s huge.
Here’s my back of the envelope math. Italy declared a national lockdown on March 9. Similarly, starting a week or two ago here, we started taking strong measures: closing schools, cancelling all large group events like professional and college sports, Broadway shows, and concerts; shuttering most businesses in cities with outbreaks; and initiating social distancing.
Unlike Italy, we haven’t had a national lockdown so it’s hard to choose a date for comparison, but I’m going to pick last Tuesday, March 17. That was day that, anecdotally, I felt that Americans really started to take this seriously, at least in part because President Trump finally reversed course the day earlier and issued guidelines that called for Americans to avoid social gatherings of more than 10 people and to limit discretionary travel.
So, we acted eight days after Italy – but our number of cases per capita is 16 days behind Italy, so that gives us an eight-day advantage.
If we assume the virus was doubling every four days, as Italy’s was at the time of its lockdown, that means we might have avoided two doubles. In other words, if the actions we’re taking prove to be effective, our peak infection rate could be only one-fourth Italy’s.
Now, it’s true that we haven’t done a complete, nationwide lockdown, but keep a few things in mind. First, Italy didn’t do a complete lockdown either. A lot of businesses remain open and many citizens have been violating the rules. Italy is infamous for this – for example, tax evasion there is among the highest in the world.
And there are other reasons to believe we’ll fare better than Italy, even without fully adopting their draconian measures. Their population density is nearly six times ours (though our cities are similarly dense). 23% of Italians are over age 65 versus only 16% of Americans. 21% of Italians smoke vs. 14% of Americans. If you’ve ever been to Italy – I go every summer for at least a week – you’ll see how they all hug and kiss each other far more than Americans. In addition, families, across generations, are far more likely to live together. And lastly, though I know some will dispute this, I think our hospitals will do a better job of treating infected patients than Italy’s.
All six of these factors indicate that the coronavirus won’t spread nearly as quickly or broadly in the U.S., so I think there’s a good chance that our more targeted lockdown will work. And when people do get infected, our mortality rate will likely be much lower.
3) It will be many weeks before we know, on a national scale, whether the steps we’ve taken to control the spread of the coronavirus are working. But, just as we can get hints from the countries that got hit first, there will be early indications in the cities and states when the virus first appeared.
Recall that the first outbreak in the U.S. was in the nursing home in Seattle. Experts estimate that the coronavirus was likely spreading in that area for at least a month before anyone realized what was happening. Then, before anywhere else in the U.S., the city and state took steps to control it and ramped up testing. As recently as nine days ago (March 12), more than one third (34.2%) of all coronavirus tests done to date in the country were in Washington, which has only 2.4% of the U.S. population.
Thus, there’s little doubt that Washington is ahead of the curve, meaning, among other things, that it’s closer than anywhere else in the U.S. to having new cases give at least a partial indication of what’s really happening with the actual spread of the virus.
And here, the news is promising. As you can see in this chart, the total number of cases doubled in the past five days (still rapid, but a lot better than every two days!), and new cases each day may be leveling off (let’s cross our fingers!):
4) Whether we can believe China’s numbers is a hugely important question.
Many are doubting them. For example, here’s an article in a conspiracy rag, the Epoch Times: Beijing’s Claim of No New Infections Contradicts Reality on the Ground. And here’s one in the New York Times that documents how the Chinese government is squelching negative stories: As China Cracks Down on Coronavirus Coverage, Journalists Fight Back.
I forwarded this to my best source in China, who replied:
No, I don’t believe this could be true, especially in Wuhan or in any of the major cities.
There is so much focus on testing, exposure monitoring/tracking, and prevention now, with a particular emphasis on imported cases from other parts of the world. The procedures at the international airports for incoming passengers are quite incredible, lasting 6-8 hours or more. Anyone who shows any symptoms is quarantined.
I, too, believe the Chinese data – not that there isn’t some noise, as there is everywhere. But I think it’s HIGHLY unlikely that the coronavirus is, in reality, still spreading out of control, but the government is hiding this fact, for two reasons:
a) As authoritarian as China is, there’s no way they could hide this even if they wanted. They already tried and failed – and every person in China and around the world is now super-attuned to this.
b) Like most authoritarian regimes, they initially tried to hide bad news (as, tragically, we did as well), but they couldn’t bury such a serious crisis. And once the masses discovered it, they threatened to revolt, forcing the government to completely reverse course (again, just like here).
The government now recognizes that this virus could topple it (again, just like here), so it’s now taking any and all measures to contain it. And the only way to successfully do so is to have radical transparency: massive testing, widespread dissemination of the data, etc.
5) Another reason the coronavirus might stop spreading so quickly is that, just like the normal influenza virus, it may be sensitive to ultraviolet light and heat, so that it will “burn off” as humidity increases and temperatures rise.
According to a study by the University of Maryland School of Medicine:
“…all cities experiencing significant outbreaks of COVID-19 have very similar winter climates with an average temperature of 41 to 52 degrees Fahrenheit, an average humidity level of 47 to 79 percent with a narrow east-west distribution along the same 30-50 N” latitude.”
The study includes this map showing “average temperature data from March 2019 to April 2019 to predict the at risk zone for community transmission of COVID-19. The zone at risk for significant community spread in the near-term include land areas within the green bands, outlined in dark black”:
This map would certainly help explain why eight of the 10 most populous countries in the world have been spared (at least so far), including Russia, India, Brazil, and Mexico.