Recession in 2006/7?
First, Prof. Hamilton does is go through the various attempts to try and model the dynamics of unemployment.1 He notes that all attempts have pretty much failed to come up with a reliable analysis that provides much in the way of prediction.
But he does not that there is an effect that economists call “duration dependence” which means that the probability of an contraction ending is a function of the length of the contraction. There is a similar effect for expansions, but only after an expansion has been underway for a while. So what does all that mean? Well the currect expansion has been underway for 14 quarters and that means that the probability of the expansion ending is going to start increasing the longer the expansion continues. Add on the fact that oil prices are rising, the real estate market is look pretty unstable, and that the Fed is slowly, but surely raising interest rates and the idea of a recession in the next 18 months to 2 years isn’t that hard to believe.
Anyhow I recommend reading Prof. Hamilton’s entire post, as well as Barry Ritholtz’.
1Unemployment is only one of the time series that the NBER looks at in making a determination as to when a contraction starts or ends.