What do experts know? This question arises from the reposting the other day of an exchange between Caprice and Dr Sarah Jarvis, wherein Caprice argued for an early lockdown and mask-wearing, only to be opposed by the doctor. In hindsight, it looks as if Caprice might have been more correct. Which poses the question, for economists as much as other scientists: how should we talk to lay-people? Here are some suggestions.

First, know what you know and what you don't. As Charlie Munger said, "it's not a competency if you don't know the edge of it." Economists, for example, do not know how to forecast recessions. We should admit this fact.

In particular, we should know the difference between the map and the terrain. The doctor told Caprice: "Unless you have read every...statistical modelling paper that's come out you cannot argue with me."

One problem here is that you should not close down debate by appealing to expertise or credentials. The only expertise that matters is that you can demonstrate.

Another problem is that it poses the question rightly raised (pdf) by Dani Rodrik: how do we choose which model to apply to which bit of reality? Even today, there is much that's not known about Covid-19. The question of which model to chose to describe its progress was therefore one on which scientists should have been humble - certainly more so than Dr Jarvis. When you don't know what to do, erring on the side of caution - an early lockdown in this case - is not unreasonable.

One example of this problem in economics is that we had a good model of how globalization creates gains from trade and hence potential Pareto improvements. This model, however, over-estimated the ability of resources to shift to new occupations. Instead, as Banerjee and Duflo wrote:

When bad news arrives in the form of greater competition from outside, instead of embracing it and moving resources to their best possible use, there is a tendency to hunker down and hope the problem will go away on its own. Workers are laid off, retiring workers are not replaced, and wages start to drift down. (Good Economics for Hard Times, p65)

Economists therefore under-weighted the prevalence (and anger) of the "left-behind" - which gave us Trump and Brexit. Our talk of GDP growth invited the reply: "That's your bloody GDP. Not ours."

Expertise, especially in the sense of a command of theoretical models, can blind us to ground truth. And lay-people sometimes know this better.

It can also cause us to under-estimate the importance of tacit knowledge. Sure, Caprice has no book-learning of epistemology, but she had a hunch that something big was needed. And tacit knowledge is important. Entrepreneurship rests upon hunches that new processes are possible or that new products are wanted.

This mix of ground truth and gut feel means we should listen to lay people. This is especially true of the totality of them when the conditions are fulfilled for there to be wisdom in crowds. Consumer spending relative to wealth can help predict equity returns, for example, because aggregate spending embodies the dispersed ground truth and gut feels people have about the future. Experts lack such knowledge.

In saying all this, I'm echoing a point made by William Easterly:

The experts are as prone to cognitive biases as the rest of us. Those at the top will be overly confident in their ability to predict the system-wide effects of paternalistic policy-making - and the combination of democratic politics and market economics is precisely the kind of complex and spontaneous order that does not lend itself to expert intuition.

Experts must guard against over-confidence. Just because you know more than the lay person does not mean you know it all, or even that you know enough. They must also be aware of groupthink and deformation professionelle.

So what do experts know that lay people don't?

Connections. That's what. Zeev Kril and David Leiser show that lay people are terrible at making connections and identifying causal links. People didn't blame fiscal austerity for low returns on their savings for example; wrongly believe the effects of "fiscal prudence" are similar to an individual's prudence; and wrongly blame immigrants for a weak jobs market.

It's the job of experts, then, to show that social science is often not common sense: for example, behaviour that's sensible for an individual might be bad for us collectively.

What I'm proposing here is a sort of (loose) division of labour. Experts know explanations and (sometimes) useful aggregate facts. But they often don't know the future or ground truth. We should respect the division of labour. Whether we have the political and media institutions that do so is, however, another matter.