Okay.
Another witness, John Williams, suggested that we shouldn't try to reconcile budgets with estimates, in the sense that the budget is really a statement of principles, of objectives. It's at a higher level than the estimates are, and we're tying ourselves in knots when we try to tie the two together. I know another congressional system in which they do try to tie them, and I suppose even some Westminster parliaments have tried to do that. The suggestion was just to kind of look beyond it.
Getting back to your point about the incentives, it's the behaviours these engender that are really more important. I like what you had to say about it—let's think about these human beings, whether it's in the government, in the legislative end of it, or in the bureaucracy. What are the motivations that you can kind of instill in people via incentives and then the processes around them?
Can you just talk about this other notion of committees looking at certain spending instead of looking at it in the macro? So instead of looking at a high-level set of estimates, we could focus narrowly on certain programs instead and take more of a sampling approach. Do you think that would be effective? The current process is very high level, and we try to uncover something, but there's really no in-depth analysis in the committees when it comes to the estimates.