January 17, 2007

Report: Three of every two government statistics are flawed

Okay, I'm joking. The real headline from the Guardian newspaper is One in five Home Office statistics are unreliable, says department head.  Maria Farrell at Crooked Timber makes the point that non-neutral claims of facts degrade public discourse. But I wonder whether even a putatively  independent body can create trustworthy facts when they're subject to subtle pressures (budgetary, etc.). Those who look for such independence are correct to criticize obvious warping of data, but those who think that nominal independence is great (as opposed to better) have never read political theory from the iron triangle forward.

Even at the level of shaping a study, negotiation can often decide what is studied. The National Reading Panel is a case in point. The report trumpets the extensive public hearings that shaped the priorities of the panel. There are two conclusions one can draw from that fact. Either the panel had decided in advance what would be studied, and the hearings were a sham, or the panel was sincere in letting the public input shape the substudies. If the first is true, the definition of research was political by exclusion. If the second is true, the definition of research was political by inclusion.

(This conclusion about the negotiability of research is true whether or not you agree with the NRP conclusions.)

Listen to this article
Posted in Education policy on January 17, 2007 8:44 AM |