Daniel Drezner has a piece in Slate looking at the charge that the Bush administration has strongly favored cronies and campaign contributors in awarding reconstruction contracts for Iraq and Afghanistan. The conclusion:

Running the numbers . . . there is indeed a positive correlation between contributions and contracts. [But] the correlation coefficient turns out to be 0.192 and not statistically significant.

Update (1026): Dan responds to some queries on the piece on his blog, including all sorts of wonky stats geekery.