How Long Will It Take Unemployment to Go Down to Something More Like Full Employment?

Jimmy was asking about this after class last week.

It also came up in a Brad Schiller piece in the February 9th issue of The Wall Street Journal.

And, I’ve mentioned in class that it is either a 1) stunning political blunder, or 2) magical thinking, that led Democrats to suggest that they could reduce the unemployment rate relatively quickly.

If it’s a political blunder, it comes from not listening to economists. No matter how bad they though Bush and the Republican were, and how good Obama and the Democrats are, it’s a fact that the unemployment rate doesn’t drop as fast as it rises.

If it’s magical thinking, it’s supported by a willful ignorance of the history and data. All they had to do was look at some graphs.

Here’s Schiller:

… We would need monthly job gains of 460,000 to achieve full employment in time for the 2012 presidential elections.

We created that many jobs one time in the last four years (May 2010). That fact should be scary for Democrats.

The White House keeps hoping for monthly job gains of 250,000. But even gains of that magnitude—more than double the average gain last year— would not get America back to full employment until 2018.

I crunched some numbers for you guys to get a sense of the scale of the problem. They are on the G drive too.

I downloaded the unemployment rate from 1948 to 2011. Then I took 2 month intervals and classified them as either the unemployment rate going up or down. Then I pulled out the change in the unemployment rate, and divided by two.†

What I found was that the average rising month went up by about the same as two average dropping months (0.16% to -0.09%).

Now, basic probability tells us that even if the unemployment rate went up steeply in this recession, that this is water under the bridge. There is no reason to expect it to come down steeply in the recovery, and in fact we ought to expect it to drop at its historically average rate: about 11 months to drop 1%.

The unemployment rate peaked at 10.1% in October 2009, so a rough estimate would be about 56 months to drop to 5% (that’s summer of 2014), and 67 months to drop to 4% (that’s summer of 2015).

Of course, the great problem in making this kind of forecast is what happens if there’s another recession before we get to those levels? There’s no reason why there can’t be (this was the problem in the 1970’s and early 80’s). Getting to summer of 2014 without a recession would make this expansion one of the longest in peacetime (although we’ve been having long expansions for 50 years now, so that wouldn’t be surprising).

A second problem is our experience with structural unemployment in the 1980’s. There, the unemployment rate peaked at a higher level, and dropped dramatically, before slowing down quite a bit: 10% to 7% looked easy, but 7% to 5% took a lot longer. This is usually attributed to that range being composed of hard cases: structurally unemployed people who became convinced, only slowly, that their old job and lifestyle were not coming back. To the extent to which we have a lot of structural unemployment in this recession, we may be faced with the same problem in a few years.

Jimmy was also curious about how the policies of extending unemployment benefits would affect this. I don’t think anyone really knows for sure, but no one suspects that it would speed up the process of getting the unemployed back into jobs.

† The reason for doing it 2 months at a time is that there are some months of no change, that are hard to classify as up or down – and there are enough of them to be problematic. Using the average over 2 months got rid of most of those.