Part 2: Should Experts Agree?
In my last EconLog essay, I examined the impossible tradeoff between expert fields. Each area of expertise is a separate silo, I argued, and we don’t have any good way to strike a balance among these competing perspectives. That’s why the politics of pandemic policy are so fraught. In this essay, I want to discuss a different source of expert failure: uniformity of opinion within areas of expertise. With the speed, scale, and scope of this pandemic we need to break away from rigid orthodoxies and move adaptively to new models. But such flexibility is thwarted by the slotting of experts into professions that often enjoy governmental support. It takes diversity of viewpoints and meaningful discussion to reach the new insights we need to fight this thing. The enforced orthodoxy of professional associations frustrates such free-wheeling inquiry.
Professions such as medicine, law, and pharmacy serve to keep outsiders out and insiders in. The expertists template of certification, professional education, and continuing education has been applied across the globe and across the disciplines. Take medicine. In the US, you cannot generally sit for a medical license unless you graduated from an accredited medical school. And guess who does the accrediting? Strictly speaking it’s the Liaison Committee on Medical Education (LCME). But they are “sponsored by the Association of American Medical Colleges (AAMC) and the American Medical Association (AMA).” Crudely put, the doctors decide who can join their club, and the AMA is their membership committee. Just consider the incentives that creates. Most obviously, doctors have an incentive to restrict entry. Outsiders must be kept out. That’s bad for you and me. Healthcare costs more, and we get less of it. Less obvious is the incentive toward uniformity of opinion and the way that sets us up for gloom and doom.
There is an incentive toward uniformity of opinion because doctors and other professionals make their money from the officially recognized knowledge of their profession. If there is too much disagreement among the experts in a field, we might start to have doubts. Yes, doctor, but why should I believe you? Uniformity of opinion becomes an imperative. Doctors and other experts are kept away from supposed quackery, “by a whole body of professional knowledge that offers them ‘scientific proof’ of the folly and even wickedness of deviance.” Insiders must be kept in. Think “epistemic Berlin wall.”
Such enforced orthodoxy creates an incentive to forecast doom for anyone foolish enough to flout expert advice. “Thus the general population is intimidated by images of the physical doom that follows” from going against the advice of the doctor or the economist.
All of this creates a dubious dynamic in times of crisis. We have come to expect that experts don’t disagree. We are correspondingly quick to follow their advice even perhaps without much critical attention. The Imperial College model of the pandemic might be an example. This is the model that predicted 2.2 million deaths in the US. It seems to have been important in driving at least some governors in the US to put their states in lockdown. I would not pretend to judge whether the Imperial model is good or bad, or whether those governors made good or bad choices. But it is noteworthy how surprised many people were when the competing Oxford model came out. This surprise seems to have inspired news stories explaining that a model is, well, a model and not the real thing. The projected death toll of 2.2 million assumed “absence of any control measures or spontaneous changes in individual behavior.” And yet this vital qualification fell from public view as any number of second-hand experts predicted unqualified doom. We have even seen fights on whether subsequent remarks by the report’s principal author were or were not a retraction. (Spoiler alert: Not really.)
Whatever the merits of the Imperial College model or the Oxford model, whatever the merits of the many lockdowns we have seen, we should not like enforced orthodoxy or the incentive it creates to forecast doom for anyone foolish enough to flout the experts. Going forward we might seek out a better system of expert influence on social policy. I will share some thoughts on that issue in my next essay on EconLog, the last of this short series.
Roger Koppl is Professor of Finance in the Whitman School of Management of Syracuse University and Associate Director of Whitman’s Institute for an Entrepreneurial Society (IES).
READER COMMENTS
Phil H
Mar 31 2020 at 10:50pm
Can this model distinguish between actual expertise and unmotivated exclusivity?
I suggest that there is a difference between the Alchemist’s Guild and the GMC (Britain’s highest medical authority). Both exclude outsiders; but one produces real knowledge that benefit outsiders; and the other doesn’t.
I can’t see anything in Koppl’s model of expertise that recognises this difference. So I can’t see any value in this model.
Michael
Apr 1 2020 at 7:12am
I think one of the failures we have seen in the response to this crisis has been, essentially, the presentation of a false choice: save lives (that would otherwise be lost to COVID-19) or save the economy (that would otherwise be ‘killed’ by the shelter-in-place orders implemented to save lives).
In reality, we don’t that kind of clean choice – if we reopened the economy on Easter Sunday as some have suggested, it would ultimately be devastated by the rising toll of infections, hospitalizations, and deaths.
robc
Apr 1 2020 at 7:55am
Are you sure? If the Oxford model is remotely correct, the majority of Americans have already been infected.
The hospitalizations and deaths occurring now are the results of early March infections.
What we need is massive random antibody testing, in order to know for sure. Although I guess 1000 random Americans would give a SD about 3% (stats people, is my math right?) that might be good enough.
Alan Goldhammer
Apr 1 2020 at 9:38am
Massive testing is indeed what is required and it only prevented by a failure of imagination. Paul Romer was on Noah Feldman’s ‘Deep Background’ podcast that dropped yesterday (you can listen to it at the website). Romer makes the persuasive argument (one that I agree with as one who has been reading far too many pre-prints these last two weeks) that we should be screening everyone in the US at 14 day intervals. Those who present with illness go home and self-quarantine for two weeks; those who do not continue working.
Of course this is a massive undertaking as maybe 25 million people will need to be screened daily. the cost is not insignificant but likely to be far less than the shutdown of the economy and the huge stimulus that has already been approved. That number might be decreased based on what an appropriate statistical sampling resulting in a reduction of R0 < 1. The lack of a rapid serological test is really hurting this effort. Some of the Chinese data that I’ve seen show that a well designed serological test is as good as the RT-PCR gene test. Those who have had mild SARS-CoV-2 and now show no clinical symptoms will have a positive antibody test and not need to be tested further.
Outside the box thinking has been largely absent from this whole discussion.
robc
Apr 1 2020 at 10:30am
Watched an interview with Thomas Massie this morning, he said basically the same thing. For less than 1% of the cost of the stimulus package, we could test everyone in the US.
But as a first round, a random test of 10k or so people to see who has already had it would be super useful. If it is 10%, then do what you suggest. If it is 70%, send everyone back to work because we have already blown thru the numbers the 30% may want to continue to isolate for a few weeks, especially if they are older or etc.
robc
Apr 1 2020 at 10:33am
I don’t know if you saw it, but on the previous thread, I suggested Sweden, Iceland, and Belarus as the control countries you asked about. Apparently they aren’t doing any kind of massive shutdowns.
Michael
Apr 1 2020 at 11:01am
You are correct that massive testing is needed – not just test everyone once but test everyone repeatedly. And sufficient testing alone isn’t enough – also needed is a capacity to respond to local outbreaks as they are detected. We need to build up all of this in order to reopen as much as we can while minimzing the risk of major outbreaks.
And I was unclear in my previous statement. There wasn’t and isn’t a clean tradeoff between lives and the economy as implied by all of the “is the cure worse then the disease?” talk from last week.
But there are absolutely plenty of tradeoffs here. Exactly what those tradeoffs are depends on our testing, tracking, quarantining capacity and effectiveness.
Alan Goldhammer
Apr 1 2020 at 12:42pm
Adding to my earlier comments. In my Maryland county we house not only the National Institutes of Health but also many biotech companies all of whom have capacity, knowledge and reagents to run RT-PCR testing. Resources such as these could be repurposed by setting up a short list of requirements so that the testing capacity could be improved many fold. There are other regions of the country that could do the same thing.
We are able to manufacture home pregnancy test kits and some of that technology can be extrapolated here to come up with a blood test.
The inability to be creative has resulted in this SNAFU (I wanted to use a more coarse term but that would have been flagged 🙂 )
robc
Apr 1 2020 at 12:52pm
Adding to my earlier comment, that is also what Massie said in the interview I saw, he wanted a Manhattan project type commitment to testing and fighting the virus.
Tests, masks, cures, vaccines, whatever, the funds in the stimulus package would have been far better spent on that. Probably at a fraction of the cost.
robc
Apr 1 2020 at 12:49pm
But isn’t that what I do every time I commute to work? I am trading off a risk of dying in a car crash vs my personal economy.
Sure, we probably don’t know enough to make the calculation accurately with the virus, but that is nothing new.
Michael
Apr 1 2020 at 10:05pm
A risky probabilistic game like Russian roulette provides a clean tradeoff between one’s life vs one’s wealth.
In real life, we rarely have that kind of certainty.
During the “should we reopen everything by Easter” debate, we could be reasonably certain that either choice would lead to significant loss of life and significant devastation to the economy, the magnitudes of which were heavily dependent on one;s starting assumptions. That’s not how most people framed the decision, however.
Michael Rulle
Apr 2 2020 at 9:30am
Re: the essay itself by Roger Koppl
The existence of licensing requirements across the board, from hair stylists to doctors and almost anything one can imagine, has been a feature of our country for a long time. It probably started out innocently enough to be check on those prone to take advantage of others. Now it has become the method by which various groups systematically take advantage of everyone—-similar to a guild system. Politicians and Governments benefit from these laws by fees and contributions. Same with Professors, as I am sure Prof Koppl would agree—I assume.
There are obviously many theoretical solutions to these problems, but when the state——and most of the people——support these systems or are indifferent—it’s not changing. Personally, as a matter of habit, I do not believe anything from studies where the self interest of the producer of such studies is clear—-or where the statistical complexity required is so great that it would almost be idiotic to believe it.
To use a too simple example——-everyone is impressed with physics. But their science is easy by comparison, for example, to climate science or economics—-the latter two being very opaque to science’s desire to understand. Einstein was a nobody—a literal clerk stuck in a ridiculous bureaucratic Swiss patent office. But within 13 years a solar eclipse in South Africa was able to “confirm” a prediction. Anybody can produce science if the science is good—-no one prevents it from happening—-as much as guilds do hold the paying jobs.
As it relates to this virus—-as Julian Simon often remarked——-sometimes you don’t need science or statistics to see the obvious. A Tornado does not require genius to know you have to get out of the way.
We all have likely had many “what if’s” about how to respond to this virus—-and who knows about 2million versus 200k etc——and even who knows if we were ignorant of this existing at all whether we would even perceive the outcomes as different than normal—-after all, the current deaths are still a rounding error. I do worry about this greatly—-with no exit strategy other than a vaccine. YET—-once we started to count—-it was impossible to not do what we are currently doing.
Comments are closed.