The Collapse of Scientific Authority (cont.) -- Confirmation Bias

Megan McArdle explains why attempts to support biases with scientific "evidence" undermines scientific authority. She writes:

This is a post about how much time we all seem to spend pretending that our views on things like homosexuality, property rights, abortion, income redistribution, stay-at-home mothers, environmental issues, and fast food can be Scientifically Proven Using the Latest Techniques! And how that erodes the trust of those we argue with.
....

I think one of the biggest problems facing economists, and to some degree other social scientists, is the feeling that if you're just a little bit willing to fudge facts, you could do a great deal of good. If you'd torture the numbers just a little--not even torture, really, just waterboarding and a few stress positions--you could convince people to do what you know, deep in your heart, is the right thing....

The subtler version of this is confirmation bias: to a libertarian analyst, papers showing that taxation causes people to stop working make perfect intuitive sense, while papers suggesting that stiff environmental legislation saves lives and money set off a pulse-racing, heart-pounding determination to discover just where the author went wrong....

Conversely we tend to grant instant credibility to studies that confirm, rather than challenge, our biases.

This is not a recent problem. When I was a working social scientist we used to talk about "massaging" data, not "torturing" them, but that was a kinder gentler day. But the problem seems to have gotten worse lately. Confirmation bias was rampant back then. Today, though, it is pervasive and not just in journalism.

It used to be assumed that within the general public and in journalism, bias would cancel out. Conservatives would rigorously examine and contest findings that supported a liberal position and vice versa. The truth would out in the end. But that has not seemed to have happened.

We further assumed the existence in each field of a "community of competence" -- a body of credentialed experts who could be trusted, through the process of peer review, to render objective, disinterested judgment on the validity of any study. That, as I have argued in numerous posts, has proven to be a chimera.

What has developed instead is the corruption of the scientific community as researchers promiscuously fudge data and respected journals publish fatally flawed studies. Science has been subordinated to the demands of ideological, political, professional, and marketplace forces. The scientific establishment been revealed as just another collection of interest groups..., one among many. The pronouncements of scientific authorities are met with healthy skepticism rather than naive acceptance.

And that is right and proper. The dream of a pure sience, invested with methodologies and review processes that would insulate it from human error, may have gone a'glimmering, but with it has gone the danger of technocratic tyranny. Scientific opinion today must compete with, and be modified by, political, economic, social, moral, and other perspectives. And in a democracy that is, to paraphrase a famous criminal mastermind, "a good thing."