He puzzles over them here:
"There are good arguments that wages are sticky for (many of) the employed. Observed wage changes cluster in funny ways, indicating an unwillingness of the boss to change the nominal wage at all, and employers testify to morale problems from wage cuts (see Alan Blinder’s work). In terms of the financial crisis, Keynesian theory explains the initial lay-offs fairly well, but it — at least the sticky nominal wage version — has a tougher time explaining unemployment persistence at such a high level.
Why don’t the unemployed lower their wages to find a job? The more tragic you think unemployment is, the greater the puzzle here, and yet the people who stress the tragedy are often least likely to admit the positive puzzle (and vice versa)."
He proceeds to give lots of very good reasons why nominal wages for the unemployed are sticky. But I think it's all a whole lot easier to understand if you realize that:
1. Labor isn't bought on a spot market. Labor is more like an investment good than a consumer good, and
2. Unemployment and labor surplus are not the same thing. You can have lots of unemployment in a clearing labor market.
I'm sure wages are a little sticky for all classes of workers, but that's not the major problem we're facing.
Liquidity Traps, Local and Global (Somewhat Wonkish)
19 minutes ago