COVID-19: The Dark Side Of “What Gets Measured Gets Managed”

Normally I’m extolling the virtues of measurement.  Most notably, I advocate that firms measure their R&D productivity.  I believe failing to measure it in the past, is the reason R&D productivity has declined 65%, which in turn is the reason GDP growth is stagnant.  Today however, I want to talk about the dark side of measurement, to address an issue that is more urgent, and that also has implications for how much GDP growth revival is required.  

Every morning for the past month, we’ve been able to read how many COVID-19 deaths have occurred through the day before.  This morning it was 23,608 in the U.S.  That sounds like a big number, but it is actually meaningless without context, so let’s provide that.  Here’s the problem—the most recent data I could find on the CDC website for US mortality was from 2018.

On average 233,000 people die per month in the US.  Here are the top causes.

            Heart disease              54,000

            Cancer                         50,000

            Accidents                    14,200

            Respiratory illness      13,300

            Stroke                          12,200

            Alzheimers                  10,100

            Diabetes                        7,000

            Flu                                 4,600

You should be skeptical of the analysis I’m about to do, because I am comparing high resolution data on COVID deaths, with old as well as low resolution data on other deaths, but that’s what we have to work with.  With that as caveat, let’s allocate all COVID deaths to a single month.  This is conservative since they’ve been occurring since February, but it seems reasonable since they haven’t yet peaked. 

Generous case for bearing COVID costs: it causes additional death

Let’s first examine COVID in the most generous light—that COVID deaths are on-top of these other deaths. If we compare the 23,608 COVID deaths to the rates above, we see COVID deaths represent less than 10% of all U.S. deaths per month.  Moreover, COVID is responsible for only 40% as many deaths as heart disease in an average month.   Shouldn’t the people with heart disease and cancer feel it is unfair COVID is getting more attention than their disease? 

This isn’t just an abstract question—health care providers across the country are turning away patients to accommodate COVID.  As a consequence, their revenues have plummeted—the Mayo Clinic reports theirs have been slashed by 50%.  This in turn has meant furloughing healthcare workers and/or cutting their pay.  

If it’s reasonable to spend $2 trillion in federal relief, in addition to sacrificing 10% of GDP, to address COVID, shouldn’t people with heart disease expect us to spend $5 trillion and 25% of GDP on heart disease?   The answer of course is no—we can’t afford it.  Taking it to the logical extreme, we certainly can’t afford 100% of GDP for all causes of death combined.  

A plausible rebuttal is that without intervention, COVID deaths would exceed heart deaths.  The problem is lack of data to support or refute that argument.  Despite the fact we have daily data on COVID deaths, we have no current data on any other causes of death.  Relatedly, we have no data on the comorbidities in the COVID deaths. Thus, we have no sense of whether COVID has actually increased the death rate, or if it merely replaced one cause of death for another.

Less generous case for bearing COVID costs: it substitutes for other causes of death

Lets’ now examine a less generous case for COVID—that many of the 23,608 deaths attributed to it, would otherwise have been attributed to heart disease, cancer, etc.   Given who is most at risk for COVID, this seems reasonable. If death from COVID is merely replacing death from other causes, but preserving overall mortality, then its threat doesn’t justify imprisoning the population, jeopardizing businesses, and burdening the next generation with a $79 billion higher annual budget deficit from the interest on $2 Trillion.   

We are in a classic dilemma of managing what gets measured, while ignoring what doesn’t. This particular dilemma comes at unprecedented cost, however—regardless of whether COVID increases the death rate or merely substitutes for other causes of death. Citizens and policymakers should demand that CDC collect data on all deaths (or at least the comorbidities for the COVID deaths) in the same manner they collect data for COVID deaths.  Policymakers should not make any decisions on lockdowns or further stimulus until they have that data.