Some Sobering Math

I've had a number of people - including clients - argue that, if the government had not shut down the economy, resulting in what will probably be peak unemployment of around 25% for May, there would be no economy, because huge numbers of people would be dead, and thus couldn't work or consume.

Clients in particular have made that argument when I explain that they should not include the impact of credit losses, reduced earnings on loans and investments due to lower rates, reduced loan demand, reduced credit and debit card interchange income, etc., when considering pandemic risk. The reason is twofold.

The first reason is that I already have my clients assessing those risks, and we don't want to double-count the residual risk (risk after mitigation). The pandemic risk should be assessed in terms of business continuity: what would the impact have been had we not spun up all the responses to the pandemic and the resulting jurisdictional guidance as quickly and effectively as we did? And how well did those responses mitigate that impact?

The second reason - and forgive this economic purist for making such distinctions - is that the pandemic did not cause the recession. The pandemic evoked a government response, and the government response caused the recession.

Now, I'm politically agnostic about that when it comes to my clients. I'm not going to share my thoughts with them on whether that response was appropriate. There will be plenty of time for recriminations and Monday-morning quarterbacking when we actually know something about this virus: how easily it spreads, the number of actual cases, the number of actual deaths based on COVID as the direct and primary cause of death, etc. Some of that we'll never know. So I don't offer clients an opinion regarding whether the government response was right, wrong or indifferent.

I won't do it here, either. My purpose is solely to apply real math to the most severe worst-case projection, and demonstrate what, even in that worst-case scenario, the economic impact might have been had the government not responded as it did.

Also, let me say that, if the death toll were as high as the numbers I use below, it would have been a heartbreaking human tragedy. And when I say later in this post that the number of deaths among those under the age of 16 is insignificant as a data point, I'm not saying that any of those deaths were insignificant to the parents, grandparents and siblings of the succumbed. I'm only talking economic impact, because that's the argument I'm countering. Let someone else decide whether the human loss outweighs the economic impact. (But hint: let them consider suicides, reduction in life expectancy, alcoholism and drug addiction, and other human costs related to an economic collapse, especially among the most vulnerable participants in the labor force, in balancing that scale.)

Okay, let's start by looking at data regarding COVID deaths by age cohort. My source is worldometers.info, which uses data from the CDC, WHO, Johns Hopkins, and other sources. According to their data, about 72% of COVID deaths are in the age 65 and older cohort, with 28% among ages younger than 65.

Now, let's look at the U.S. civilian labor force by age cohort. Only 10% of the labor force is comprised of individuals aged 65 and older. My source here is the Bureau of Labor Statistics (BLS), and I'm using pre-pandemic data.

So, let's combine the data, by multiplying the deaths by age to workers by age. That means that 72% of the 10% of U.S. workers over the age of 65 have died, or about 7% of the total labor force. And, if we assume that all of those under 65 who died were employed (an important assumption - I'll explain momentarily), we can count the whole 28%. That gives us a total of 7% of the labor force, plus 28%, for a total of 35%.

However - and pay close attention here - we can't assume that 35% of the labor force would have been wiped out. Why?

Because the 35% is based on the total number of deaths, not people. It's not a per capita number. So it can't be applied to the entire labor force, any more than it can be applied to the entire population. If it could be, then we would indeed have 35% of the labor force, or about 55 million, deaths in the U.S. And if it were applied to the entire U.S. population, we'd have about 114 million deaths. Not even the most dire models projected numbers anywhere close to that, and the actual data is running about 0.1% - 0.2% of those numbers.

Further, while I'm going to assume, as a worst case, that all individuals under the age of 65 who died were employed, that's not the case. Some people are fortunate enough to retire before that age. Some were among the 3.2% unemployed before the government shutdown ensued. Also, some COVID deaths occurred among individuals under the age of 16, which is the low-range cutoff for BLS data on the labor force. However, those numbers are insignificant as data points. Just know that the 35% is skewed to the high side for those reasons. (It's skewed even further by the cause of death methodology employed by the CDC, but that's another discussion for another day.)

Okay, still with me? The Imperial College of London put out the first COVID model, and it has been thoroughly debunked as hot garbage. A dumpster fire. Worse than the fatally flawed IHME model, which I have debunked just as thoroughly in this blog. Suffice it to say that the Imperial College's model creator has resigned his position as a government advisor.

So why use data from the Imperial College model? First, it was the model used by the British and U.S. governments to initiate the lockdowns that have so severely affected those two countries' economies, at least the service and factory sectors thereof. And second, it projected the most dire scenario regarding total deaths if we did nothing, so it provides the ultimate worst-case data, ridiculously extreme as that data is.

The Imperial College model projected 2.2 million deaths in the U.S., a number that you've heard bandied about quite a bit, assuming you've been awake the past three months. So let's apply our combined deaths-and-employment percentage to that.

Had 2.2 million Americans died of COVID, the math indicates that 35% of them (on the high side) were participants in the civilian labor force. And 35% of 2.2 million is 770,000. Divide that by the total civilian labor force pre-pandemic of about 157 million, and you get about 0.49%.

In other words, had the government done nothing in terms of shutting down the nation's economy, had the model been accurate (and it wasn't), and assuming that all COVID deaths under the age of 65 were labor force participants (they weren't), and assuming that all reported COVID deaths actually resulted from COVID as primary cause of death (and, according to the CDC's own website, they didn't) -

The unemployment rate would have increased by 0.49%. So today, it would be about 3.7%.

Let that sink in.