I'm not a historian—I try to be an economist—but I think it goes as far back as something like 1968, when it was brought in. What would be the context back in the late sixties, and why did they decide that this system had gotten to such a state where if they didn't look at estimates, they would say, let's just approve them? Obviously it was a sad state back then, which I think speaks to Mr. Brian Jean's comments that this is an ongoing process that we need to continue to improve.
On February 29th, 2012. See this statement in context.