On Friday, Roger Koppl of Syracuse University presented his book Expert Failure at the Invisible Hand Seminar co-sponsored by George Mason University and the Institute for Humane Studies.
The thrust of Dr. Koppl’s book is that, similar to market and government failure, experts can fail in their task as well. Experts can give poor advice, he influenced by known or unknown biases, and react to incentives just like the rest of us. An expert, for the purposes of this discussion, is anyone paid for their opinion (this is how Koppl defines expert in his book).
There is another aspect of expert failure which plays into the problem and that is a failure to communicate.* Even if an expert does everything perfectly, if s/he cannot adequately communicate their message to the decision-maker, then expert failure can still result.
This expert failure to communicate is a major issue in law & economics. Economists are trained in statistics and empirical methods. Lawyers and judges may not and juries almost certainly are not. In his 2009 textbook, Basic Concepts of Probability and Statistics in the Law, Columbia law professor Michael Finkelstein gives two examples were ambiguous wording by experts and misinterpretation of probabilities by juries led to false convictions (see pages 3-5). These failures can lead to pro-prosecution bias in juries (or, alternatively, pro-defendant bias if other statistical fallacies are made). In the two legal cases above, the convictions were overturned on appeal, but one has to worry about the error rate given the already confusing nature of statistical wording and hypothesis testing.
In Expert Failure, Koppl discusses increased competition among experts as a means for solving expert failure. A good expert can explain his argument in common-language and not necessarily hide behind jargon. Competition between experts may indeed reduce this expert failure to communicate, but it ultimately rests on the head of the decision maker. And indeed the more complicated a topic is, the more the reliance on expert opinion is needed, which increases the likelihood of expert failure to communicate.
Ultimately, I think this all adds up to a rather strong presumption of liberty. The more control experts have over our lives, the more likely it is for expert failure to occur. Indeed, some of these highly complex functions, like contract issues or tort issues, may be best left to arbitrators rather than judges. The more complex an issue is, the less one would want to centralize it.
*Full disclosure: this point may come up in the book. I have not read it, having ordered it shortly after the seminar