Some Thoughts on Unemployment

I just finished reading This Time is Different, it's a great but very dry read about the history of financial crises. The authors put together what is probably the largest amount of data ever compiled on economic statistics over the last 1000 years in an attempt to perform some analytics and draw some conclusions about what economies look like in the run up to a crises (useful in helping to avoid them) and what kind of severity and duration can be expected after they occur.

I lack the time and, quite frankly, the ability to relate even a fraction of what the book has to say. I can, however, summarize quite nicely. The power of delusion is so nearly absolute that with almost no exception the general public will deny all the signs pointing to a crises and will be thoroughly convinced that "this time really is different" all the way up until the crises finally hits. At that point the leaders will then usually throw up their hands and remark how it was impossible to predict. Finally, the return to "normalcy" is a lot harder and takes a lot longer than anyone likes to admit, especially when a banking crises is involved.

But my point today is not to review or rehash the book. I just wanted to use one piece of it as a stepping off point for some comments about unemployment. In an analysis of post-1900 banking crises which have occurred (of which there are 14), the authors found that on average the unemployment rises (as measured from trough to peak) by 7 percent and takes an average of 4.8 years to do so. Interestingly enough, on average the unemployment rates in emerging economies tend to rise less and last shorter than those in developed country. It's speculated (though not proven) that this has a lot to do with far more wage flexibility in the emerging economies.

So I thought that I would use this to see how our current situation stacks up to these historical averages. Here's a graph of unemployment since 2000 (which comes directly from the Bureau of Labor and Statistics):


Unemployment basically started increasing around June of 2007 (when it jumped from 4.4 to 4.6%), although it was actually fairly flat until 2008. The peak to this point (and hopefully THE peak) occurred in October 2009 at 10.1%. So all in all, we've got an increase of 5.7% (about 20% less than the average) in ~29 months (almost exactly half the average).

There's a couple conclusions to draw from this. First, compared to historical average we've actually done quite well. Certainly a double-dip recession is possible, but we've now had a year of decreasing unemployment (albeit a painfully slow decrease) so I don't think I'm jumping the gun too much. Of course, the big question now is whether or not this is because of the stimulus or whether this was going to transpire anyway and the stimulus just represented futile paddling against the current.

Speaking of the stimulus, it was pretty comically inept for Obama and his advisers to ever make the statement that with the stimulus they hoped unemployment would peak at 8%. It's usually not politically smart to ever tie yourself to any one number but if you're going to do it the 8% number was an incredibly bad one to use. Of course, by the same token it's also just as ridiculous to say that the stimulus was a failure because it didn't keep unemployment under 8%. Now the argument about whether or not the stimulus worked is for another time. I'm just saying you can't criticize a car because it can't take you to the moon, even if the car salesman told you it could.

As to where we go from here, prognosticator John says that we'll finally get unemployment down below 9% by mid-2011, hopefully below 8% by end of 2011, and below 7% by mid-2012 right when the election starts really ramping up. And I really think that 7% is the key number. If it gets below that, I think Obama is reelected easily. If it doesn't, he's got big problems.