The Cyclicality of Search Intensity in a Competitive Search Model

Reasonably calibrated versions of the Diamond-Mortensen-Pissarides search and matching model of unemployment underpredict, by a wide margin, the volatility of vacancies, unemployment, and the vacancies-unemployment ratio - variables at the heart of this model. These shortcomings motivate two modifications to the Diamond-Mortensen-Pissarides model. First, wages are determined via competitive search (wage posting by firms along with directed search on the part of workers) rather than the usual Nash bargaining. This change is motivated by the fact that most unemployment variation in the U.S. is due to non-college educated individuals, and that wages of newly-hired non-college educated workers are predominantly set by wage posting. Second, workers are permitted to take direct action to affect the outcome of their labor market search through search effort. With these modifications in place, the benchmark model captures 70% of the standard deviation of unemployment and the vacancies-unemployment ratio, and almost 80% of the volatility of vacancies. A recalibration of the model that targets the variability of the vacancies-unemployment ratio results in reasonable parameters, and can account for almost all of the cyclical variability in unemployment and vacancies.