Exiting Corona

Based on some reader feedback moderately critical of an earlier Covid 19 post as insufficiently deferential to public health concerns, this blog opens with a discussion of exactly what it intends to do. This is a public policy blog – it constitutes an effort to analyze the optimal public policy response to the events and circumstances our American society faces.  Such an endeavor involves balancing first, second, and third order concerns.  With respect to the coronavirus, first order concerns center on the immediate public health response; second order concerns focus upon mid-term public health strategies and the economic consequences of the short and mid-term public health response; and third order concerns involve the long-term impact of the Covid 19 response on American institutions themselves.  [An obvious example of a third-order consequence is the election of President Donald J. Trump in 2016, an event that almost certainly does not occur without both the Iraq War and the 2008-10 Great Recession.]

In a recent episode of Glenn Loury’s podcast, Steven Teles defined the term “policy analysis” as an attempt “to integrate the insights of a whole bunch of different disciplines whose specialization by definition means they don’t all come to the same agreement.”  It should be noted that the Covid 19 situation presents exactly this source of expert conflict, where the stated goals of many public health experts (restrictive lockdowns designed to crush the spread of Covid 19) come into conflict with those of many economists, who fear that long-term restrictions upon normal economic function may lead to considerable economic pain and an arguably worse long-term outcome for the population than generated by the initial public health impacts.

Turning specifically to the American response to Covid 19 itself, some basic facts and opinions should be detailed before this blog’s conclusions on exiting the current stage of the coronavirus response are shared.

-The science on the impact of the coronavirus is both in its infancy and thus involves a fair degree of uncertainty.  While some might criticize reliance upon scientific studies in such an environment, the most robust public analysis has no alternative but to use the available evidence.

-Covid 19 has been identified as the cause of death for roughly 75,000 to 80,000 Americans in just two months.  There are nearly 1.4M cases of Covid 19 in the United States thus far.  This implies a “Case Fatality Rate” or “CFR” of between 5.5-6.0% among those clinically diagnosed in the United States with this disease.  Notably, there is a fairly wide variation of CFRs amongst nations afflicted with significant numbers of Covid 19 cases.

-There is almost no debate within the scientific community that, due to both testing limitations and the existence of some asymptomatic/less symptomatic infections, the actual fatality rate of Covid 19 is significantly less than the Case Fatality Rate.  The actual fatality rate is termed the “Infection Fatality Rate” or “IFR” (and can be understood as the rate at which those actually infected – rather than infected AND diagnosed – have passed away).

-Based on the limitations in diagnosing all cases of Covid 19, the best means of determining the scope of infections is likely general population testing for antibodies (also known as “serology studies”).  Such testing is not foolproof and has been subject to significant criticism, but thus far no viable alternative to determining the population-wide scope of infection have been presented.

-Epidemiologists have conducted a substantial number of serology studies, displaying a wide variety of results.  One group of researchers (largely out of Stanford University) has conducted at least three such studies – one in Santa Clara County, California; another in Los Angeles County, California; and a third of individuals associated with Major League Baseball across the United States.  The first two studies showed substantial antibody presence relative to diagnosed cases in two California regions; in contrast, the third showed limited antibody presence among people affiliated with MLB.  The state of New York has also conducted seroprevalence testing showing about 15% of New Yorkers have been exposed to the virus (a measure at the time of testing roughly 10 times the numbers of diagnosed cases in New York).

-A number of foreign serology studies have also been conducted, almost all showing a fairly widespread penetration of Covid 19 infection when compared to actual diagnosed cases.  The linked preliminary study from Geneva, Switzerland reveals an antibody prevalence in this city of about 10 times the diagnosed cases in Geneva (similar to New York).

https://www.medrxiv.org/content/10.1101/2020.05.02.20088898v1

-Notably, New York and Switzerland have both been subject to fairly high rates of active Covid 19 infection testing (due to significant outbreaks of the disease in those locations).  By contrast, California has seen a less robust testing regime, likely because of the far lower prevalence of serious disease in the state.  In effect, because New York and Switzerland have conducted more active Covid 19 case testing than California, one would expect California to have a higher rate of undiagnosed cases.

-Based on the known evidence, Covid 19 infections probably exceed the diagnosed rate of Covid 19 cases by at least 10x.  How much higher this number could range is presently an unknown, but it seems unlikely that it would exceed 30x.*  Using these boundaries, that would approximate between 15-40M Americans having thus far been exposed to Covid 19.

-Applying this same 10-30x multiplier to the Case Fatality Rate in the US, the Infection Fatality Rate of Covid 19 in the US would fall between .2% and .6% of cases.  From the evidence we have, New York has presented at the high end of that range, which is consistent with concerns about “hospital overload” resulting in lesser care to too many serious cases of the disease (anecdotal reports from NY also bear this conclusion out).  In effect, this analysis finds that between 2 and 6 of 1,000 people inflicted with Covid 19 die from the disease.

-There has been dramatic geographical variation with respect to cases and deaths from Covid 19 in the United States.  This at least points to the possible impact of both population density and weather, among other factors.  The dense and cooler Northeastern corridor has been hardest hit (centered upon New York City) and several major Midwestern metros (Chicago, Detroit) have witnessed larger concentrations of the disease.  The South and West have been spared to a much greater degree.

-Covid 19 disproportionately impacts the elderly, but is not confined to those of advanced age.  An analysis by the Wall Street Journal estimated that the average fatality from Covid 19 cost its victims 10-15 years of expected lifespan, a number similar to heart disease and, not shockingly, other respiratory ailments.

-That said, the hardest hit cohort is clearly those in elder care.  It is probably reasonable to estimate that about 40% of the Covid 19 fatalities in the United States are amongst persons housed in long-term care prior to this outbreak.  This number does vary widely by state, with some states (including relatively hard-hit Massachusetts) exceeding 50% and others (such as New York, where overrun hospitals may have exacerbated the death rate among previously healthier individuals) presenting a less concentrated impact among the most disease-vulnerable populations.

-Nationally, Covid 19 cases and deaths have been declining in recent weeks (cases more than deaths, which tend to be a lagging indicator), and probably more importantly, the rate of positive Covid 19 infection tests has dropped significantly (i.e., a lower percentage of people being tested have turned up positive).  In many locations, Covid 19-related hospitalizations are also on the decline.  While the disease is far from eradicated, there are few signs of accelerating spread, and the risk to the medical system being overwhelmed has at least been contained.

-The hit to the United States’ economy from Covid 19 and the resulting lockdowns have been profound.  First quarter US GDP fell nearly 5%, with a massive plunge expected for Q2.  Unemployment is at levels not seen in generations, though many of the presently unemployed anticipate being rehired relatively quickly as American returns to work (meaning the level of long-term unemployment caused by Covid 19 may not be well reflected by the catastrophic short-term unemployment numbers).  That said, United States’ equity markets have rebounded considerably from previous lows, apparently reflecting a sentiment on Wall Street that quick federal action to provide substantial monetary stimulus may ameliorate the worst economic consequences of the shutdown.

So, you may ask, where do all these facts and conclusions leave us when determining the best Covid 19 policy as of today?  First, a summary of conclusions from the information above is in order.  In effect, Covid 19 has had significant, albeit mostly localized/regionalized, impacts to public health.  It is likely that low tens of millions of Americans have been exposed.  Of course, this conversely means several hundred million Americans have not been exposed and are at legitimate risk of contracting the disease should a dramatic resurgence take place (and the likelihood of this occurring is, frankly, an unknown).  The risk of death for an inflicted person is relatively low, albeit considerably higher than most common diseases found in the United States, such as the flu, colds, and bacterial infections.  The worst health outcomes are, as is typical for most ailments, concentrated amongst the elderly population.  Short-term economic damage has been substantial, but long-term economic forecasts do not yet reflect overwhelming pessimism regarding a recovery.

Premised on the above analysis, this blogger takes the clear position that, as a general proposition, most regions of the United States should be reopening (and frankly, should have been doing so for the last two to three weeks).  Reopening, of course, does not mean that “everyone and everything returns to normal tomorrow.”  Instead, a state of “reasoned normalcy” should be encouraged in most locations throughout the country.  The prongs of reasoned normalcy – and the related discussions about how to engage in reasoned normalcy while working to limit unnecessary Covid 19 spread (and attendant fatalities) – are identified as follows.

1.      States and regions within states that have not suffered from significant per capita outbreaks must initiate the process of reopening immediately, should they not already be doing so.  The large mass of the United States, including many areas where the political leadership has been resistant to reopen (including the West Coast and the non-urban portions of the Midwest) have no excuse not to initiate a re-start … yesterday.

2.     Reopening does not constitute an immediate return to the way things were in early March.  Physical distancing, such as six feet limitations, should be strongly encouraged amongst individuals, and mandated at places such as restaurants where there is not a free flow of people.  In that vein, limiting capacity at retail stores and restaurants is a sensible solution (at the present time, 50% capacity for such businesses seems to make a degree of sense, with upward movement of these numbers if initial reopening measures are deemed to have succeeded).

3.     The wearing of masks in most interior spaces of commerce should be encouraged and, in certain cases, mandated.  While this blogger is sympathetic to the resistance to mask-wearing, there is enough evidence that masks help to limit the spread of disease that wearing them is not an unreasonable ask.  That said, there are businesses, such as restaurants, where masks are not practical, and mask-related restrictions cannot be mandated in connection with such spaces.

4.     Information about individual and familial risk must be shared with the public in a meaningful way.  The average person should not have to dig deeply to find a fair assessment of his or her own risk.  Unfortunately, that has become the norm, and it results from the media having become active participants in every debate rather than mere purveyors of facts.  With respect to Covid 19, a large mass of the media has settled on the unworkable and likely disastrous “lockdown forever” position, an opinion that likely won’t be eviscerated until their employment prospects are in jeopardy.

5.     Americans over 65 should be encouraged by federal, state, and local authorities to continue limiting their unnecessary contact with others.  This is the segment of the population clearly at the most risk, and efforts to contain the worst case scenarios of the disease must be focused on those most vulnerable to it.  Saving lives does not mean restricting everyone in search of an amorphous “common good,” it means identifying the best means of limiting troubling outcomes for the most individuals.

6.     Nursing homes must be locked down, with every precaution taken to protect their residents.  The death toll in the US has been heavily concentrated here, and this outcome was entirely predictable.  The lack of focus on this weak point in early March is easily the biggest policy failure of the Covid 19 pandemic, and one where early proactive measures would have likely had the most dramatic impact.  This blogger would suggest that governors consider training and using National Guard soldiers housed on-site at elder care facilities to perform some of the basic care tasks necessary for their residents, which would limit the “in and out” of personnel at these locations.

7.     Because multi-generational households (especially those in the New York City area) appear to have seen mass transmission from less impacted younger generations to more vulnerable older ones, free government-provided housing, particularly at currently unused hotels, for essential workers living with more vulnerable, older relatives, should be offered.  Such a policy would have the ancillary benefit of assisting the rather beleaguered hotel industry.

In sum, lockdowns are not a maintainable strategy, and reasoned normalcy is achievable.  The United States just has to be willing to take the risk … and the precautions … to achieve it.

* The 30x number was selected as the upper bound in the 10-30x estimate because that range has consistently shown up as a rough lower bound of some of the seroprevalence studies estimating a far greater rate of infection to diagnosed active caseload.