My mistakes
I have spent an awful lot of time working on issues in this pandemic to have not been one of the “always rights.”
Welcome to Plugging the Gap (my email newsletter about Covid-19 and its economics). In case you don’t know me, I’m an economist and professor at the University of Toronto. I have written lots of books including, most recently, on Covid-19. You can follow me on Twitter (@joshgans) or subscribe to this email newsletter here. (I am also part of the CDL Rapid Screening Consortium. The views expressed here are my own and should not be taken as representing organisations I work for.)
In the New York Times, Ezra Klein pointed out that, economist, Alex Tabarrok, has been consistently right throughout this pandemic:
He called for vastly more spending to build vaccine manufacturing capacity, for giving half-doses of Moderna’s vaccine and delaying second doses of Pfizer’s, for using the Oxford-AstraZeneca vaccine, for the Food and Drug Administration to authorize rapid at-home tests and for accelerating research through human challenge trials. The through line of Tabarrok’s critique is that regulators and politicians have been too cautious, too reluctant to upend old institutions and protocols, so fearful of the consequences of change that they’ve permitted calamities through inaction.
He’s not the only one with a good wrap. Earlier, the New York Times said the same about Zeynep Tufecki.
In recent years, many public voices have gotten the big things wrong — election forecasts, the effects of digital media on American politics, the risk of a pandemic. Dr. Tufekci, a 40-something who speaks a mile a minute with a light Turkish accent,has none of the trappings of the celebrity academic or the professional pundit. But long before she became perhaps the only good amateur epidemiologist, she had quietly made a habit of being right on the big things.
And that was back in August and she has only been righter since then.
This got me thinking. I have spent an awful lot of time working on issues in this pandemic to have not been one of the “always rights.” It is time to account for what I have got wrong and what lessons can be drawn from it.
First, let me list the issues on which I have been wrong:
February 2020: I argued to my colleagues who were thinking of wearing masks on a plane that this was unnecessary because that is what the public health people were saying. That, obviously turned out to be completely wrong. I ended up writing a chapter about that particular bit of misinformation in The Pandemic Information Gap.
March 2020: I argued that we need the armed forces to be mobilised to build hospitals to handle the massive overflow. This was the post that got me started on this journey and I followed it up with other pieces. But, as it turned out, in most places that wasn’t needed. Why? The epidemiological models that I was relying on to make this assessment were incorrect. As it turned out, people would voluntarily engage in social distancing and other behaviour and the worst outcomes could be avoided. I ended up writing a research paper about how behaviour modifies those models; a summary of which is here.
May 2020: I argued that we are going to have a terrible problem rationing a vaccine. We have a supply shortage and we have a global issue regarding rich and poor countries. But I thought we would have a big issue within countries in terms of frustration at being able to get the vaccine. We have some of that but it wasn’t the calamity that I had in my mind. Instead, for the most part, people are patient. Part of this is that I thought that the vaccine would confer immediate rights that would lead to two classes of people. That didn’t happen because the vaccine had unknown properties. But it still could happen but, by that time, the shortages will not be as much of an issue. That is why I put this in my “I was wrong” bracket.
November 2020: I was the first person to raise the question of whether we needed two doses of the vaccine. But, being a responsible person, I consulted with experts in immunisation and argued that we likely needed two doses. I then followed up on that in December but made a maths mistake. As it turned out, that going to a “first doses first” strategy was the right decision based on knowledge and the balance of risks. We delayed doing that a couple of months.
This list may not seem like much. I have certainly said more that was right than was wrong — especially when seeing pandemics as an information problem. Presumably, you are reading this because I am “mostly right.” But those mistakes stick with me.
There is a common thread in three of these cases. For each one, I relied on the scientific experts all the way through to determining the prediction or policy outcome. Those assessments were not wrong per se but they were incomplete. For instance, cloth masks weren’t as protective as N95s but that was different from not being useful at all. The pandemic did look bad in March 2020 and we didn’t have great information as to how a population would react. But that didn’t mean we couldn’t take it into account more explicitly and marry epidemiology and economics more concretely rather than siloing each. And in first doses first, I had asked the right question but the standard answer from the experts was incorrect and based on its own risk calculations that themselves were incorrect. There was more room to move there.
The broad issue here is not that the scientific experts were incorrect. They weren’t on the key baseline facts. The issue was that those facts get translated into a decision using a risk calculus. Those decisions/predictions/recommendations are made with the expert’s risk calculus but what we need to do is expose that calculus and question it. This needs to be done at every stage. This is perhaps clearest with first doses first where an extremely risk-averse approach to “what we know” led to the incorrect judgment.
The one that is “on me” is the vaccine rationing issue. It seemed to be a no-brainer that it would be good to be vaccinated and freeing. But there was enough uncertainty regarding sterilising immunity that two classes didn’t really emerge. Also, there was no much pull for them to emerge while many people were not vaccinated. I didn’t think hard enough there.
It is a tough bar to get things right all of the time — so kudos to those who do. But the same bar is tough for scientists as well as economists. The right strategy is to give people a break — there is currently some big blame game going around on experts that isn’t helpful — but instead ask everyone to look at their predictions and reflect on where they went wrong. That is the purpose of my hand wringing approach here. Hopefully, that will lead to getting more right.