If you can keep your head when all about you are losing theirs, you’re not paying attention to the experts. Epidemiologists tell us that if we do not hide in our houses with the door securely locked, hundreds of thousands will surely perish. Economists tell us that if we do not return immediately to work, civilisation will collapse. Good luck figuring out which expert has the better advice. Is it any wonder a harried Michael Gove blurted out, “I think the people in this country have had enough of experts from organisations with acronyms saying they know what is best and getting it consistently wrong.”

Expert fear-mongering did not begin with the pandemic or Project Fear. In 1922, John Maynard Keynes warned that “squalor follows” if we do not make the economist “king.” Daniel Defoe complained of the “calculators” and “quack-conjurers” whose fear-mongering “kept up their trade” in London’s plague year of 1665. He shrewdly observed, “And had the people not been kept in fright about that, the wizards would presently have been rendered useless, and their craft had been at an end.”

Defoe complained of quacks and wizards, whereas today’s epidemiologists and economists have rigorous scientific training, mathematical models, advanced statistics, and careful evidence all going for them. True. But today’s scientists are still people. And that means they respond to incentives just like everyone else. The issue is not lying and cheating. Sure, some modern experts are quack-conjurers who lie and cheat. Let’s not mistake a white lab coat for a golden halo. But that’s not the main thing. Even when the experts are trying to be sober, scientific, and scrupulously neutral, they will feel certain pressures.

Think if it were you. You’re an epidemiologist and the prime minister calls to ask you how many will die if we don’t have a lockdown. What do you tell him?  You can’t just look up the number. The pandemic is only now taking off and your knowledge of it is correspondingly sketchy. It’s hard to say. Every number is a guess. If you give the prime minister a low number, there will be no lockdown. What if he accepts your low numb

er and we have no lockdown?  Maybe everything will be fine. But maybe there will be many more deaths than you predicted. You will get blamed. People will shame you as a bad scientist.  And, because you are a good and decent person, you will feel guilty. Blame, shame, and guilt. This is a bad outcome.

If you give him a high number, there will be lockdown. No one will ever be able to say that your estimate was too high, because your estimate assumed no lockdown. Even if a lot of people die during the lockdown you can say, “See? Think how much worse it would have been without the lockdown.” Thus, if you give the prime minister a high number, you will get credit for saving lives. You will be able to take pride in your sterling reputation as a scientist. And you won’t have to feel guilty about lost lives. Praise, pride, and innocence. This is a good outcome. The logic of the situation is clear. You have every incentive to predict doom and gloom if no lockdown is ordered.

It may be that the famous epidemiologist Neil Ferguson, who, until recently, was an important member of SAGE, has felt such pressures in his career. At one point in the pandemic he told a columnist for the New York Times that 1.1 million deaths was the “best case” for the US.  In 2001 he blasted as “unjustifiably optimistic” a study suggesting that mad cow disease deaths “may peak at 100 cases per year in Britain and kill no more than a few thousand people in coming decades.” Rejecting this relatively optimistic view, he said deaths are in the long-term likely to be much higher at something only slightly less than 136,000.  The true number as of June 2014 seems to have been 177.  In 2005, he was alarmed by bird flu (H5N1). “Around 40 million people died in 1918 Spanish flu outbreak,” he told the Guardian. “There are six times more people on the planet now so you could scale it up to around 200 million people probably.” That’s a lot more than the World Health Organisation’s estimate for cumulative worldwide deaths, 2003-2020 of, ahem, 455.

Before we allow ourselves to feel too smug about Ferguson’s seeming propensity to overestimate epidemiological dangers, ask yourself once again what you would have done if the prime minister had asked you for such an estimate. If it is true that Ferguson has systematically erred on the side of pessimism, we can likely blame it on his incentives. Experts are people, and they respond to incentives.

We should value expertise, but fear expert power.

We need experts and expertise. I want the help of doctors when making decisions about my health, of educators when making decisions about my children’s schooling, and so on. And when the pandemic swept in, our politicians needed to consult the experts. Time will tell whether we needed the relatively strong lockdown we got, whether the Swedish model was better (maybe, maybe not), and so on. I would not pretend to judge at this relatively early moment. Whatever the best path might have been, we should beware experts with lousy incentives. If experts are people who respond to their incentives just like non-experts, then we may need to rethink how we use experts in the political system.

Today we have the “rule of experts.” Monopoly experts have the power to choose for you in one field after another, including child protective services, economic policy, and pandemic response. But if you give some humans the monopoly power to choose for other humans, you have created some dangerous incentives. The rule of experts gives you the highest chance of expert failure. We should value expertise, but fear expert power. Whenever possible, then, we should do away with the rule of experts by empowering the people. Let each person choose for themself, and let the experts compete with each other to provide advice. That’s a call for ramping down the power of government bureaucrats and ramping up personal freedom. But you can push that idea only so far with pandemic policy.

Whatever the best policy might have been, at least some restrictions were clearly needed. In the moment of danger, governments cannot avoid turning to experts to help them craft policy. When confronting a pandemic, then, is there nothing a government can do but listen to the epidemiological experts and obey their wizardly words? There may be a few things governments can do to limit “expert failure” in moments of crisis.

Governments should recognise that their experts are, all of them, giving a partial perspective. Apparently, British and American policy was driven primarily by a report whose lead author was Neil Ferguson. That report seems to have considered only one danger: Covid. The one-sided analysis of that report may have left governments in the US and UK insensitive to the possibility that that lockdown itself might create its own fatalities, which might even end up larger than the number of Covid deaths. As economists never tire of reminding us, we are always facing tradeoffs and must adjust along all margins.

Governments should also be more diligent in the pursuit of competing opinions. In his essay, “What is science?” Richard Feynman remarked “Science is the belief in the ignorance of experts.” A government that respects science should be sceptical of experts and, perhaps, more diligently seek out multiple viewpoints. In other words, when governments cannot leave the matters in the hands of the people, it should do what it can to simulate a competitive market for expert advice. A simulation is not the real thing, and we may grimly expect that in future crises governments will again fall victim to expert failure. But a greater effort to engage diversity of expert opinion within and across areas of expertise and a livelier scientific scepticism toward experts and their expertise may at least make expert failure less frequent and less severe.

Roger Koppl is Professor of Finance in the Whitman School of Management of Syracuse University and a faculty fellow in the University’s Forensic and National Security Sciences Institute.  His most recent book is Expert Failure.