Aaron Ginn wrote a long, charts-and-statistics-filled blog post at Medium arguing that the available public health data shows that COVID-19 is less easily transmitted, less fatal, and more likely to fade away with the hot weather than the conventional wisdom would have you believe. Medium deleted the post, which is now hosted at ZeroHedge. Deleting arguments like Ginn’s is a bad and dangerous way to handle the still-roiling debate over what governments and society should do in order to react to the disease.
Whether or not you agree with Ginn’s arguments for, say, reopening schools, people like Ginn and Justin Hart are asking some important and detailed questions about what we know about the progress of the coronavirus. I am squarely in the camp that thinks swift and aggressive public steps have been a sensible and necessary response to a serious public health threat. We can come back more easily from the economic damage of an overreaction than from letting the virus run wild to see if it’s really as bad as we think. But it is unhelpful and hazardous to ignore the real, human costs of protracted lockdowns, which will require increasily strong justifications the longer they drag on. As Congressman Chip Roy explains, those lockdowns will at some point become unsustainable, and the debate over the conditions needed to end them will be important ones, turning on a mix of scientific data and our political and philosophical instincts about risk and liberty (more on which below from Theodore Kupfer).
How do we make those decisions? Stanford epidemiology professor John P.A. Ioannidis argues that the data we have so far should be taken with huge grains of salt for being incomplete and inconsistently collected:
Projecting the Diamond Princess mortality rate onto the age structure of the U.S. population, the death rate among people infected with Covid-19 would be 0.125%. But since this estimate is based on extremely thin data — there were just seven deaths among the 700 infected passengers and crew — the real death rate could stretch from five times lower (0.025%) to five times higher (0.625%)…reasonable estimates for the case fatality ratio in the general U.S. population vary from 0.05% to 1%. That huge range markedly affects how severe the pandemic is and what should be done. A population-wide case fatality rate of 0.05% is lower than seasonal influenza… The most valuable piece of information for answering those questions would be to know the current prevalence of the infection in a random sample of a population and to repeat this exercise at regular time intervals to estimate he incidence of new infections. Sadly, that’s information we don’t have.
David Katz of Yale University’s Yale-Griffin Prevention Research Center (a CDC-funded public health research institution), argued in the New York Times for a similar caution in how far we go based on limited data:
The data from South Korea, where tracking the coronavirus has been by far the best to date, indicate that as much as 99 percent of active cases in the general population are “mild” and do not require specific medical treatment. The small percentage of cases that do require such services are highly concentrated among those age 60 and older, and further so the older people are…These conclusions are corroborated by the data from Wuhan, China, which show a higher death rate, but an almost identical distribution. The higher death rate in China may be real, but is perhaps a result of less widespread testing…We have, to date, fewer than 200 deaths from the coronavirus in the United States — a small data set from which to draw big conclusions. Still, it is entirely aligned with the data from other countries. The deaths have been mainly clustered among the elderly, those with significant chronic illnesses such as diabetes and heart disease, and those in both groups.
Ginn and Hart, unlike Ioannidis and Katz, are data guys without a medical background. Those of us who are amateurs at medicine in general and epidemiology in particular should have some humility in working through competing claims on these points. Generally applicable statistical principles are important, but so is knowing how they apply to a specific area. I’ve spent enough time over the years working with baseball statistics, public opinion polls, and analysis of stock price movements to recognize that the general concepts that are common across disciplines also need to deal with the specific characteristics of the real-world things being measured by the data.
But stifling those points of view and hand-waving them with “shut up, I’m an expert” is not a healthy way for democracies to make important decisions. Expertise is valuable at a time like this, but it is only effective so long as experts are able to marshal their expertise to explain to the public what we know about the answers to many of the questions Ginn’s piece raised: What are the real transmission rates and transmission mechanisms? How likely, really, is infection within a household, or from generally just being around asymptomatic people? How long does the virus survive in infectious form on surfaces? What is the likelihood that the virus can survive and spread in hotter climates, and what can we learn about that from where it has not spread so far? What are the real mortality rates compared to the baseline mortality rate of the ordinary flu?
Too many experts have an insular tendency to assume they will be believed just from credentials, because their friends believe them that way. And too many hear clever amateurs say “I’m asking questions” and immediately treat them like Holocaust deniers or moon-landing skeptics, when we are in fact dealing with a data set that is evolving daily and full of uncertainties. Interrogating the elite, expert narratives is important. That is doubly true when you remember that most of the population is getting its information from ignorant second-hand sources that are just as likely to be too alarmist as too dismissive.
University of Washington biology professor Carl Bergstrom took a more detailed crack on Twitter at debunking specific problems with Ginn’s analysis. More engagement of that sort with the details is better. The process of getting to the truth and having the truth accepted by the people – who, in America, are still sovereign – works better with an adversarial airing of differences rather than just trying to drive dissenting voices off their platforms. The consolidating elite and institutional consensus may look solid now, as it appears to span politicians and doctors, Republicans and Democrats, Americans and foreign governments. But elite groupthink has been wrong before. It is better now than later to ensure that it can justify itself to a world in which trust in elites and institutions has fallen precipitously in the past two decades.