Unemployment and Excess Capacity

The excess capacity series is defined as 100 - capacity utilization rate.
The unemployment series is the civilian unemployment rate.

The excess capacity series (red line) peaked in June of this year, and has been moving downward ever since. If the pattern in the two most recent recessions holds, those in 1990-91 and 2001, the peak in the unemployment rate will come between 16 and 19 months after the peak in excess capacity, i.e. around a year from today (though prior to 1990 the peaks were coincident).

The most recent data on the unemployment rate showed a downward tick from 10.2 percent to 10.0 percent, so perhaps unemployment has already peaked and the lag will be shorter this time. But perhaps not. As an inspection of the unemployment series in the graph shows, the unemployment rate bounces around even when it is trending upward or downward. So it's hard to tell from one month's data whether the downward tick in the unemployment rate is temporary and unemployment still has a ways to go before peaking (as in the last two recessions), or a sign that a turning point has been reached and things are getting better (which would represent a reversion to the more coincident movement in the two series observed before 1990).

Note, however, that in the 2001 recession, unemployment fell briefly just after excess capacity peaked, but then resumed its upward movement for several more months before reaching a turning point 19 months after the turning point in excess capacity. Thus, while the recent downward tick in the unemployment rate is good news, certainly better than an uptick, we should be prepared for the possibility that the pattern in the last recession might repeat itself and unemployment will head back upward for several more months before it reaches its peak. I hope that doesn't happen, the sooner unemployment returns to normal the better, but we need to be better prepared than we are for the very real possibility that unemployment will continue to trend upward. I'd like to see more done on both the monetary and fiscal policy fronts as a preemptive measure, we can always ease off if things turn our better than expected, but at the very least we need to resist calls from the deficit and inflation hawks to begin pulling back and continue the programs that are already in place.

[Question: What happened from mid 1997 through the beginning of 1999 that caused the two series to move in opposite directions and separate?]

Update: Paul Krugman follows up here.