These Fed economists observe that an increase in unemployment can be broken down into two factors: a rise in inflows into unemployment (due to companies increasing layoffs) and a lengthening of the duration of unemployment spells (due to companies suppressing hiring).
The following link is a chart of how these two factors have played out in each recession since 1969:
http://economistsview.typepad.com/.a/6a00d83451b33869e2011278d9b39028a4-popup
In modern recessions, it's the lengthening duration of unemployment spells, not the rise in inflows (layoffs), that have been the dominating factor. Recessions have usually started with a brief, modest burst of inflows (layoffs), which quickly subsides. What then drives the sustained rise in unemployment is a long, plateauing increase in unemployment duration due to lingering hiring freezes, which takes until the end of the recession to reverse. (I think this is related to the shift out of manufacturing; both hiring and firing are less volatile over the course of the business cycle nowadays because inventory adjustments are less of a factor.)
This recession is different. The rise in inflows has been unparalleled. Not only has the rate of inflow already risen by more than any other recession, but it's been on an unbroken rising streak for a much longer period of time, and there's no sign of a reduction yet. And given that we still have a service sector economy, there's no reason to think that duration will subside any more quickly then in the last three recessions. Also, with household balance sheets what they are, there will be less of the kind of pent-up demand that could drive hiring - unless consumer borrowing patterns go back to the crazy years, which seems unlikely.
The CBO unemployment projections may be subject to rapid obsolescence.
SA