Study the Efficacy of Consultants' Predictive Power

I've been saying for years that "if you can't predict the future then you don't understand the present". It's basically a layman's version of The Scientific Method (with big props to Karl Popper). If you don't understand the present then your only participation in the public debate should be to listen/read/gather information until you can improve your understanding to a level where you become productive to the debate. In any public policy debate there are invariably myriad researchers weighing in with not only analysis of what was or is but also with what they think will be (predictions). Predictions have the benefit of being "falsifiable". Take the 2009 Healthcare Reform Debate for example - there are numerous predictions on which the public is basing their decisions. Let me quickly mention here my personal pet peeve of debaters giving equal (or greater) weight to predictions that support their side as they do to the objectively demonstrable facts which refute their side. This blog post is meant to start a "meme" that encourages professors/researchers/PhD candidates to do a thorough study that measures the track record of consultants' predictions. This group of consultants to be studied can and should include: 1) Think Tanks (e,g, RAND, CATO, Heritage, Center for American Progress etc.) 2) Universities (e.g. Michigan, Berkeley, the Ivy League etc.)  and definitely don't forget the for-hire consultancies... 3) Management Consultants (e.g. McKinsey, Booz Allen Hamilton, Pricewaterhouse etc.). If one group, entity or person is particularly ineffective at predicting (no better than chance) or even worse if one's predictions are inversely correlated to reality then that would be useful to know for the public debate. CONCLUSION: Someone / some group should start tracking the predictive power of the players in policy debates. That way we can put more faith in those with accurate predictions and less in those who make bad calls. NOTE - Brill's Content, a great magazine around the turn of the millennium, used to have a cool feature where they tracked pundits' predictions and, as I recall, the right wing pundits had substantially worse scores than the liberals (I'd appreciate it if someone found proof that I am wrong or right about that). I'd like to see that tracking revitalized going forward somewhere online (maybe mediamatters.org politfact.com or factcheck.com?). Let me know if you find such a resource.