Weekly Standard: Kickin' Back with Tax Payer Money : NPR
The work done by most of my colleagues did bear on issues of wider relevance and not all of it was so ideologically compromised as to be useless. But the readership of academic journals is tiny, and most of this work had no impact beyond a small circle of interested academics — for understandable reasons. Philip Tetlock, a research psychologist at Berkeley, tested the accuracy of 82,361 predictions made by 284 experts including psychologists, economists, political scientists, and area and foreign policy specialists, 96 percent with post-graduate training. He found that their prognostications did not beat chance. The increasingly ideological nature of social science will not improve this record.
