David, I'm going to answer probably not directly the question you asked, but I think kind of getting to the point. If you look at incremental margins, it's a very hot topic. It's important, I think, right now considering the volume increase that we have in front of us. Pricing and material cost were a tailwind in 2010 and we're not expecting that much of a tailwind in 2011. However, we had two major headwinds in 2010 that will be much more neutral in 2011. We went from zero incentive compensation to about $800 million incentive compensation between '09 and '10. That was an $800 million drag to incremental margins. Our product mix going from a very favorable mix in 2009 to something more approaching, I think, what you would expect in 2010. That hurt operating profit by more than $1 billion. Between those two items, that more than offset the kind of net benefit of material costs and pricing. And next year, I think it's highly unlikely, unless we far exceed the outlook, that incentive compensation would be a headwind. At the outlook level, it's actually a small tailwind. Sales mix, probably still a little bit negative, but nowhere near as negative as it was in 2010. So we've got some tailwinds in 2010 as well. On the material cost question, sometimes, I think that people believe that 100% of our purchases are steel. In fact, most of our purchases are not that raw, if you will. They're not commodities. The vast majority of what we buy are highly engineered components that have mostly value-added, and we don't see the kind of increases there that you do on commodities. We historically have done a lot of good in cost reduction to the point that you made product redesign. That is definitely a component of the material cost. If all we were looking is commodity, what we're doing was looking at commodities and talking about material cost, it would be an increase. The commodity portion of material cost, I think, unless there's some turnaround in where commodities are today which we don't expect, the commodity portion is going to be up. But other cost reductions, sourcing, product redesign, holding back the commodity increase for as long as we can, that should all help.