Friday, August 6, 2010

Wilkinson and Picket in Full Retreat

Having suffered a number of devastating blows in their public debate with us, Wilkinson and Pickett now appear to be in full panic mode. They have made a bizarre declaration on their homepage demanding that "all future debate should take place in peer-reviewed publications"!

This is yet another abuse of the trust that the general public places in
science.

* The very book we are debating, “The Spirit Level”, was not peer-reviewed.

* The principal conclusion of the book, namely that weak correlations between health and inequality constitute “evidence” of a causal link, was never peer reviewed.

* When Wilkinson and Picket lied about the work of Nobel prize winner James Heckman, that was not peer-reviewed. Recall that Heckman, upon learning how Wilkinson and Picket cited his article, called it a "misrepresentation of my work". Lies like that would not make it through peer review in serious journals.

* When I pointed out that there was no statistically significant relationship between life expectancy and inequality, in any sample of countries, with any measure of inequality, from either the U.N or the OECD, and forced Wilkinson and Picket to admit in the Wall street Journal that the strongest claim in their tour was simply wrong, there was no need for peer review.

*It is simply not an appropriate or standard code of conduct in academia to demand that a policy debate be peer-reviewed. Arguments should be evaluated on their own merits. Someone purporting to be a scientist ought to understand this.

When Nobel prize winning scientists in serious fields, such as (say) Milton Friedman or Paul Krugman engaged in a policy debate on issues of relevance to the broader public, they never claimed that the debate should only be peer-reviewed. Have you ever previously encountered such a claim? I certainly haven't.

But these two Nottingham and York professors, from the soft field of social medicine, have the audacity to attempt to try to silence the critics of their shoddy work by demanding peer-review in a public debate?

Why would I need peer-review to point out - for example - that more serious work by more senior scholars in far more prestigious journals comes to the opposite conclusion of Wilkinson and Picket? Economists already know this. They can read the original work themselves. Economists and anyone with elementary statistical training knows also that "association" (the criteria used by W&P in their survey article work) is not the same as proving causation (the claim W&P make in their book).

Of course everyone knows why they are making this comical demand. It takes time and effort to publish peer-reviewed articles and few serious social scientists have any incentive to point out the obvious flaws in Wilkinson and Picket's book. It is simply a lot of extra work which yields few rewards.

People who are winning debates rarely try to end them.

Wilkinson and Picket are hoping that we will give up debating their lies, stop pointing to their cherry picking of data and misrepresentations of the state of science. However, unsophisticated non-scientists such as Swedish Social Democratic party leader Mona Sahlin can perhaps be fooled about what is the norm in science and believe that our critique should be discounted because Wilkinson and Picket say so. After all, they are professors.

My response to Wilkinson and Picket demand that "all future debate should take place in peer-reviewed publications" is simply this: It is not for you to decide how I choose to point out the distortions in your book. But if you promise to limit your self-promoting book-sale tour aimed towards the general public to "peer-review", I will limit my critique of your campaign to academic journals. If you continue to make false claims in public, I will continue to expose your false claims in public.

Sounds fair, right?

Also, here is a challenge. Manage to publish in any serious (say, top 50) economics journal your claim that inequality causes mortality. If you can do that we will attemp to respond in a peer-reviewed publication.

No comments:

Post a Comment