It came principally from the fact that we have been gauging our construction forecast from an expectation that builders would try to adjust production to reduce the months’ supply of unsold homes down toward a more normal level over the next couple of years. Between the March forecast and this forecast we were surprised (1) by the weakness in the level of home sales and (2) by the very substantial further increase in the months’ supply of unsold homes. So to get the inventory of unsold homes back into a more normal relationship—and even then we are not getting back to where we were in the March forecast in terms of months’ supply of unsold homes—we saw a bigger production adjustment as being necessary. Now, the awkwardness in relation to your question about the fundamentals is that we actually think that the overhang of unsold homes is a fundamental, even though it is not really in our models of residential construction. Those models do not have a special component for the inventory of unsold homes, in part because that factor is much more episodic than it is a normal sort of right-hand-side variable. So we do view it as fundamental. We think that what we learned was that a bigger problem is out there, a bigger overhang, and we are having it work off. Therefore, it takes longer to work that off with deeper production cuts.