There Are No Experts In The Category For Which We Require Expertise.
Late last week during a press conference, Governor Andrew Cuomo of New York said that he did not trust the scientific experts employed to advise his office about the coronavirus pandemic.
Without meaning to exonerate Governor Cuomo for his mishandling of several aspects of his state’s pandemic response – much of it, of course, informed by the advice of the same experts he no longer trusts – his volte face is entirely appropriate. We should be more skeptical of what science can do for policy.
The pandemic has revealed a shortcoming in our ever more technocratic approach to addressing social problems: there are more and different kinds of problems than there are kinds of experts. The division of scientific expertise according to discipline, field, and sub-field does not encompass, much less map onto, all of the problems for which we might want experts.
The best policy approach to the pandemic would not have prioritized health to the exclusion of economic, psychological, and sociological considerations. The best policy would have minimized human suffering, all things considered. Unfortunately, though there are epidemiological experts and economic experts, and psychological experts and sociological experts, there are no scientists whose expertise encompasses epidemiology, economics, psychology, and sociology.
The problem we raise is not among those frequently pushed against government-by-experts. In his influential paper “The Use of Knowledge in Society,” F. A. Hayek argued that the technocratic approach to policy making was doomed to fail because scientific knowledge was only a small part of the knowledge required to make technocracy effective. No matter the scientific wisdom of policy experts, they could never provide the local knowledge, the “knowledge of the particular circumstances of time and place,” required to control society to the degree necessitated by the technocratic approach.
It is, moreover, true that scientific consensus is often fleeting and regularly overturned, and that, in any case, consensus is neither unanimity nor a marker of infallibility. But the problem that we raise would remain a problem even if scientists were unanimous and infallible in their respective fields, and omnipotent about particular circumstances of time and place.
Scientists learn their trades through specialized study of specific phenomena, which they investigate in isolation under carefully controlled conditions, holding other factors constant. The best scientists in their fields are experts about a very limited range of phenomena that are assumed not to interact or integrate with any other kinds of phenomena. Outside of the graduate seminar and theoretical model, beyond the laboratory and controlled trial, however, other things are not constant.
When the phenomena of multiple scientific fields interact, such as when it is necessary to trade off the health costs of a virus against the economic and other costs of a lockdown, policymakers can turn to experts about isolated phenomena. But there are no experts about the interaction of different kinds of phenomena or about the proper weighting of some against others. Policymakers can ask epidemiologists to weigh in on epidemiology, infectious disease specialists to weigh in on infectious disease, and economists to weigh in on economics. But there are no experts about how these subjects interact or how to balance them.
Covid-19 is foremost a public-health problem, and policymakers have rightfully turned to physicians, epidemiologists, medical researchers, and other public-health experts for advice about how to respond to the pandemic. But just because Covid-19 is primarily a public health problem doesn’t mean that it’s only a public health problem. It has been obvious from the beginning of the pandemic that policies aimed at preventing the suffering associated with being infected with the virus would lead to knock-on consequences concerning the economy, society, and human psychology, and their secondary effects on human health.
When making policy about complex phenomena, politicians can only guess which experts to attend to, and which to ignore. There is no reason to assume that such guesses hit their marks more often than not or, indeed, that they tend to improve the relevant circumstances rather than make them worse. Policymakers do not have access to interdisciplinary experts about complex, multi-causal phenomena. Such experts do not exist.
Perhaps unfortunately, policymakers have apparently turned less regularly to the economists, psychologists, political scientists, and sociologists who might have provided advice about the side effects of policies meant to limit deaths and other forms of physical suffering from the virus. Policy responses to the pandemic have tended to treat one form of human suffering, the suffering directly due to the effects of viral infection, as more important than all other forms of suffering, such as economic, social, and psychological harms, and the further physical suffering that these harms cause.
As Hayek also had occasion to note in his 1974 Nobel Prize Lecture, playfully titled “The Pretence of Knowledge,” it is a basic but frequently made error to think that the complexity of modern society requires a scientific approach to policy making. As phenomena get more complex, the number and kind of relevant causal factors multiplies. As the number and kinds of causal factors multiply, as their interrelations proliferate and grow more intricate, our scientific understanding of them and, thus, the prospects for effective technocratic policy making necessarily diminishes commensurately. In a world where human knowledge is limited, complexity undermines – not buttresses – arguments for technocracy.
The discovery, development, and distribution of vaccines has revealed another shortcoming of the technocratic approach: at best, expertise extends as far as the design of a policy; science has nothing to say about how a policy is best implemented and administered. As a consequence, even policies based on unanimous and infallible scientific principles (which, again, do not exist) will falter in their application to real-world circumstances. “There’s many a slip ‘twixt the cup and the lip.”
When relevant phenomena are complex, political decisions can be no more technocratic, no more justified by the best scientific knowledge than were the decisions of kings and queens in a pre-scientific age. One possible answer is to reform science to better understand complex phenomena, the interplay of causal factors heretofore studied exclusively under ceteris paribus clauses. If they are not to give up on technocracy, then politicians need experts about complex phenomena capable of pronouncing scientifically on the pronouncements of experts about simpler phenomena.
Of course, this only pushes the problem back a level and sets the defender of technocracy on a potentially infinite regress: when the complex phenomena that these higher-level experts understand also interact, experts at yet a higher level will be required to make technocracy effective.
The other possible answer is to give up on technocracy and return to democracy, individual rights, and popular sovereignty.