Chaotic Planning Solutions in the Textbook Model of Labor Market Search and Matching

This paper demonstrates that cyclical and chaotic planning solutions are possible in the standard textbook model of search and matching in labor markets. More specifically, it takes a discretetime adaptation of the continuous-time matching economy described in Pissarides (1990, 2001), and computes the solution to the dynamic planning problem.The solution is shown to be completely characterized by a first-order, non-linear map with a unique stationary solution.Additionally, the existence of a large number of periodic and even aperiodic non-stationary solutions is shown.Even when the well-known Li-Yorke and three-period cycle conditions for chaos are violated, we are able to verify the new Mitra (2001) su.cient condition for topological chaos.The implication is that even in a simple economy characterized by search and matching frictions, an omniscient social planner may have to contend with a fairly robust and bewildering variety of possible dynamic paths.