Front Page Magazine has an excellent article (see related comments at Samizdata; via Instapundit) by Hussein Shirazi. Some people advocate the immediate withdrawal of US and Allied troops from Iraq (because, according to them, this would stop the insurgency). Shirazi analyses the past accuracy of some of these pundits' predictions - and the picture is not pretty. More intellectual capital should be invested in this kind of enterprise. The public should know to what extent the people who loudly advocate a certain policy have been utterly wrong in the past. This is not to say that a mistake should invalidate pundits' opinions and analyses, but the extent of trust placed in them should, at least in part, be proportional to their general accuracy. Alas, this is often not the case at all. A heightened focus on this aspect of the public discourse would lubricate the market of ideas and increase its efficiency.