Some voices call for slowing down our response. They are wrong.
I lean conservative/libertarian. I’m not sold on climate change, I support the 2nd Amendment, and I believe that much of the time government solutions are worse that the problems they try to solve. I base these opinions on data.
Ever morning I pull up RealClearPolitics.com. Today (3/21), articles from The Federalist and American Greatness argue that we are sacrificing too much wealth on something we don’t know enough about. Sometimes, though, even chicken little is right and the sky is actually falling.
I am not equipped to debate Britain’s Imperial College report, but I can walk an average reader through some observations in the data that don’t need fancy modeling software, but can be checked on the calculator on your phone. Looking infection and death rates reveals two basic alternatives.
The South Korea/Germany model. South Korea instituted aggressive testing and tracking early and has probably tested everyone that gets the sniffles. As of this writing on the evening of 3/21, they have 8799 cases and 102 deaths for a 1.16% mortality rate. Calculating this way is a trailing indicator, because we don’t know how many of the still open 6K cases will die, but it is enough to know that it is a low estimate.
Germany has 22,364 cases with 84 deaths, for a mortality rate of .38%. In Germany’s case, there is widespread testing that probably also captures the majority of cases. Their number is likely lower because their cases are newer and the current infected population skews younger. Germany’s mortality rate will likely climb to be more like South Korea.
Both of these countries have advanced medical systems with hospital capacity keeping up with hospitalizations, so there is no added mortality from lack of capacity. Let’s take a guess and say that when receiving a high standard of care, mortality is about 1% or even an absolute best case of .5%.
The China/Italy model. What happens to mortality when the medical system is overwhelmed? In Italy, there are 53,578 cases with 4,825 fatalities, or 9%. Unlike, Germany or South Korea, though, there are large number of people infected but not tested. Patients don’t get tested unless they go tot he hospital, and people don’t go to the hospital unless they are critical.
Hubei province is a better case study, with 71,859 cases and 3,259 deaths for a mortality of 4.5%. It’s very likely that there were asymptomatic and lightly symptomatic cases that were never tested, but then again I don’t trust the reported death rate either.
Let’s take a wild guess on the positive side, and say that mortality in an overwhelmed medical system is 3%. This allows for another 36K cases in Hubei and 107K cases in Italy. There are probably more non-diagnosed in Italy, but remember that calculating this is also a trailing indicator; many of the recently infected will still be hospitalized.
Now to the specific arguments:
- The stealth hypothesis
Julie Kelly at American Greatness argues that COVID-19 has been circulating widely since the first of the year, and that in January and February it was classified as influenza-like illness (ILI). She cites around 100K per week visiting a health care provider for ILI and testing negative for the flu, plus however may didn’t go see their doctor.
This doesn’t seem to work. Even if you assume an absolute best case of .5% mortality, 8–10 times worse than the flue, 400K infections would have led to 2K excess deaths. More likely, 400K infections would have overwhelmed a hospital somewhere.
Yes, there were probably COVID-19 cases in the US in January and February that were below the radar. But not hundreds of thousands.
2. Not that many people are dying.
Joy Pullman at the Federalist argues, in part, that, the United States is exceptional in its mortality rate. She points to an article by the Richard Epstein at the Hoover Institution written on 3/16, where Epstein points to our low death rates compared to other countries. Pullman also shows an Excel chart made by her husband that shows America with a much lower fatality rate at this point in our epidemic, more closely tracking Germany or South Korea rather than Italy or Spain.
As already discussed, the absolute best case for a modern, non-saturated medical system like the Germany/South Korea model or the cruise ship experience is .5%. The real number as the pandemic advances is more likely to be 1.0% or higher. If 100K become infected in New York/New Jersey or another state, it’s hard to see how the local medical system would not collapse without intervention. Then we will be looking at 3% mortality, and we will then start tracking Italy and Spain. If we have multiple large epicenters, we will not be able to supplement out of national assets.
3. We just don’t have enough information for the level of economic pain
Ultimately, both articles can be summed up by a quote from Pullman:
“It seems a fool’s errand to pre-emptively and indefinitely risk everyone’s livelihoods without hard information about what is happening and a risk assessment that includes the serious dangers of killing the U.S. economy, not what computers project will happen with lots of missing, unreliable, and rapidly changing information.”
Today, this is demonstrably untrue. We don’t truly know what the worst case scenario might be, but we do know that the best case scenario leads to .5 to 1% mortality based on modern countries with low infection rates where it has already happened. Without social distancing, closures, and lockdowns, would there be 1 million infections? 10 million infections? 100 million infections? No one will ever know.
What we do know is that uncontrolled spread shortly leads to medical system collapse and at least 3% mortality. We don’t need to know the exact shape of the curve. We only need to look abroad to see what’s already happening.