We propose a new business cycle theory. Firms need to randomize over firing or keeping workers who have performed poorly in the past, in order to give them an ex-ante incentive to exert effort. Firms have an incentive to coordinate the outcome of their randomizations, as coordination allows them to load the firing probability on states of the world in which it is costlier for workers to become unemployed and, hence, allows them to reduce overall agency costs. In the unique equilibrium, firms use a sunspot to coordinate the randomization outcomes and the economy experiences endogenous and stochastic aggregate fluctuations.
In other words, by coordinating with each other, if only implicitly, employers make the firing threat more fearful. You don’t have to interpret this paper literally as an entire explanation for cyclical unemployment, only that it may have something to do with the story.
And here is his about to appear JPE piece with Greg Kaplan (pdf):
We propose a novel theory of self-fulfilling unemployment fluctuations. When a firm increases its workforce, it increases the demand facing other firms—as employed workers spend more than unemployed workers—and decreases the extent of competition facing other firms—as employed workers have less time to search for low prices than unemployed workers. In turn, the increase in demand and the decline in competition induces other firms to hire more labor in order to scale-up their presence in the product market. The feedback between employment and product market conditions generates multiple equilibria—and the possibility of self-fulfilling fluctuations—if the differences in the shopping behavior of employed and unemployed workers are large enough. Empirical evidence on spending, shopping and prices paid suggests that this is the case.
In general, not enough popular macro discourse asks the question of how much of the cycle results from self-fulfilling prophecies. Furthermore what does that imply for policy? yes, “confidence” can be important, but confidence in what exactly?