Adam Fleisher says it was the impact of the reserve clause, which ties young players to their teams at below-market salary structures. More at the Sports Economist.

I think that Fleisher's argument can, to some extent, be reconciled within Moneyball itself. As I read it, he seems to suggest that the marginal value of moneyball simply may be higher when it comes to picking amateur players at the beginning of their careers, rather than free agent professionals. This is a point that Lewis stresses though--that the A's also developed better mechanisms for evaluating amateur players, such as taking college players. Note too that drafting college players reinforces Fleisher's argument: because college players are older and more mature, they are likely to have a greater number of productive years under the reserve clause umbrella than younger players.

Regardless, it is an interesting gloss on Moneyball.