The Decline of Experts and Faith in Expertise

Tom Nichols wrote an essay in the March/April 2017 issue of Foreign Affairs entitled “How America Lost its Faith in Expertise — and Why That’s a Giant Problem.”

Nichols decries Americans’ rising skepticism of experts’ positions, and the public’s substitution of relativism (everyone’s opinion has equal weight) for the measured hierarchy of expertise (experts’ conclusions matter because they’re based on verifiable data and scholarship).

Nichols paints the notion that “my ill-informed view is as worthy as your informed view” as extremely dangerous to a society as complex as ours — a danger that makes common sense: skepticism must be grounded and informed by repeatable experimental data, careful, conscious awareness of the limitations of certitude, and so on.

Evidence must be repeatable by 3rd parties with no financial or career stakes in the outcome.

But Nichols misses the key dynamics in why experts and expertise are in poor repute.

1. Systems and problems are now so complex that in any assessment, expert or amateur, the results are defined by the model, parameters and inputs selected by the assessment.

In other words, structure the model in a certain way, and you get the results you sought.

There are inherent limits to models of complex, inter-connected systems with a multitude of feedbacks. Consider the complexity of the modern global economy and the relative simplicity of the models being used to explain and predict economic behaviors and trends. No wonder the experts have gotten it so wrong.

2. Experts are often wrong, for a variety of reasons. One is the inadequacy or distortions resulting from their simplistic models that make a variety of implicit but poorly defined assumptions — in effect, what to include, what not to include, what to overweight, what to underweight.

Look at the field of nutrition. A vast array of studies on nutrition, diet and health reach apparently contradictory conclusions, and the consensus of experts on the ideal diet have changed considerably over the past decade. What was once “good” (margarine) and “bad” (butter) have reversed places.

Is it any wonder that supposedly “factual” results generated by experts has lost traction?

3. many of the credentialed experts work for private enterprises that benefit financially from the results of “science” (in quotes for the reasons noted above — models, inputs and statistical analysis can all be adjusted to get the desired results).

In other words, who’s writing the checks? Consider the large number of STEM employees who toil for Big Tobacco or other cartels profiting from products that have negative impacts on human health.

4. The line between public relations and science have been blurred to the benefit of various special interests. Confirmations of previous studies are hyped, conflicting data is underplayed or ignored, and the limits of certainty are left unsaid.

5. Many of our systems’ complexities are partially the result of the intrinsic complexities of human behavior — behavior that doesn’t model as precisely as the physical world.

Consider the expert-economists’ view that “the wealth effect” generates positive growth in the economy. In effect, they are attempting to quantify and manage “animal spirits” — a complex mix of emotions spurred and reinforced by internal perceptions and herd behavior.

Can this even be considered “science”, given the inherent impossibility of quantifying internal states and herd/crowd influences? Yes, we can ask people their views in surveys, but does their stated view correspond with their internal state? If so, for how long? Might we get entirely different answers next week? How can we tell if the person is giving us the answer he/she reckons we want to hear?

6. Experts generally have no skin in the game — they can issue assessments and conclusions, and the accuracy of their conclusions doesn’t impact their pay, security or standing, unless it is provably fabricated. There’s no personal price with espousing a contingent conclusion as “fact” and “scientifically proven,” when as if nutrition is the equivalent of physics.

7. Probabilities, contingencies and uncertainties are stripped out of expert assessments to make the biggest public-relation/profitability impact. If you actually dig down and study the statistical foundation of Phase III drug trials, for example, you often find serious limits on the applicability of the conclusions: the sample size is small, only a few of the trial patients benefited significantly from the medication, and so on.

Yet the medication is issued as “safe”, “reliable” and “helpful” — all claims based on very sketchy statistical foundations.

8. The public has difficulty sorting “soft science” such as economics and sociology from “hard science” (physics, chemistry) and both of those from highly contingent science such as nutrition.

Part of this difficulty can be laid at the feet of those attempting to financially benefit from supposedly scientific studies, and partly at the feet of our educational complex, which typically does a poor job of explaining the limits of double-blind studies, self-selected surveys, small sample sizes and the inner workings of probability/certainty.

All of which is to say that to some degree, the rising skepticism of experts and expertise is self-inflicted by polarized institutions and for-profit lobbying by those who benefit from the public “buying into” their interpretation or massaging of expert assessments.

Consider the range of experts’ views on foreign policy. Articles arguing for diametrically opposed policies have equal weightings of credentialed experts. The common-sense approach is to discount both claims of expertise and demand something more than mere credentials as the reason we’re supposed to accept a conclusion as being worth more than a randomized opinion.

Experts appear to be complaining that their word is no longer enough. It isn’t, and that is a positive development: show us the science, be clear about the limits, contingencies and uncertainties, and accept that human bias is often driven by self-interest that must be identified: cui bono, to whose benefit?

If you found value in this content, please join me in seeking solutions by becoming a $1/month patron of my work via