You've really hit the nail on the head on the legacy costs problem with your question about what happens as the amount of production shrinks.
Legacy costs are composed mostly of two items: first, the pension costs; and second, the health care costs related to the retired worker and ascribed to the portion of an active worker's career that is already over. In essence, they are a form of compensation for work that's already been done. The two main ones are pensions and health care.
Pensions should not be a legacy cost. That's the whole idea of pension funding. A regulation associated with pension funding is that you pay for it as you go along. The problem is that the regulation or the structure has not been successful--because of the volatility in financial markets, because of the dramatic changes in life expectancies and actuarial assumptions, and because, in the case of General Motors, of a weakness in pension regulation at the provincial level; they have access to a loophole. So there is a pension legacy cost, even though there shouldn't be.
For health benefits, there is no pre-funding mechanism. We have been exploring with the companies the possibility of a pre-funding mechanism for retiree health benefits. It would be akin to a pension plan.
So those are the sources of the costs. As production declines and current employment declines, then the legacy costs measured not in dollars but measured per hour of work--or measured per vehicle, as you've suggested--do balloon out of control. We have no control over that. We can't control the fact that they're laying off our people and not producing as many vehicles. That makes the costs per unit look higher.
We have taken many initiatives in this contract that will substantially reduce GM's legacy costs on both the pension side and the retiree health benefits side. I can't put a precise number on it, since it's all dependent on actuarial estimations, but there is a huge legacy cost reduction coming out of this contract.