Roughly speaking, if we had been held back at the June level, we would estimate that housing starts probably would have been about 100,000 units weaker than they currently are. Running that through the model suggests roughly 0.3 percent on the level of GDP. Whether all that would have played out by now is not entirely clear. In terms of broader economic consequences, one thing that we showed in an alternative simulation in the Greenbook was, if this unusually low term premium turns quickly and we get a rise of another 50 basis points there accompanied by some weakness in the stock market and some endogenous response on credit spreads, the effect would be pretty powerful. It results in a forecast of an unemployment rate rising to 5½ percent, even with your easing of monetary policy, according to the estimated Taylor rule, down to 4 percent for the fed funds rate. I think that those kinds of movements in long-term interest rates work powerful effects on our overall economic outlook.