How deep and long will the recession be?

At this point, most people accept that the U.S. economy is in a deep and long recession, but how deep and how long is total speculation at this stage. If there were no additional fiscal stimulus, the economy might begin to inch back upwards in six months to a year. With substantial fiscal stimulus the economy might begin to bounce back in three or four months.

One difficulty is that a stimulus-fueled recovery is not an indication that the economy is back to health. The real question is when the economy will be healthy on its own without significant stimulus. Otherwise, the removal of stimulus could simply result in a double-dip recession.

So, maybe the proper question is that how long the recession will last, but how long it will take before the underlying economy shows enough strength that most people would agree that the underlying recession is really over. That could be nine months, a year, eighteen months, or even two years.

Another difficulty is that virtually all government spending is effectively fiscal stimulus, so how do we really say that any particular dollar of government spending is really targeted at fighing a recession rather than a "normal" investment in a stable economy. For example, one element of fiscal stimulus is bound to be transportation construction spending for bridges and highway and the like. The Democrats would argue that we should have been doing that spending all along. So, maybe once we start that additional stimulus, it stays in place indefinitely, raising the baseline economy. In fact, one could argue that both the current and preceding recessions were due to the same cause, insufficient investment in government infrastructure.

I would also caution that military spending is also a fiscal stimulus, so any decline in military spending due to a withdrawal from Iraq (that doesn't immediately simply shift over to Afghanistan) will result in a decline in GDP unless replaced with other fiscal "investment." One unknown is the impact of returning a significant number of guard and reserve soldiers to civilian life, where their jobs are as uncertain as anybody else's.

As far as my headline question, I really do not know, but I suspect that the recession will "destroy" about 5% of GDP and that it will take another one to three months after a modest to moderate level of fiscal stimulus (starting in February?) before the underlying economy starts trending up from that "step down" in GDP. But, it is also possible that employment will continue to trend down and lag the overall GDP recovery by another six months after that.

Note that technically a recession ends as soon as the up-trend begins, which would be the month that a significant level of fiscal stimulus hits the streets. Recessions (ala NBER) are measured "peak to trough" and recoveries are measured "trough to peak." In other words, even if there is a 5% "step down" of GDP, we do not need to build back any of that 5% decline before a recovery can technically begin. This is a part of the reason why a jobs recovery typically will not kick in right away.

-- Jack Krupansky