An Empirical Study of the Adverse Selection Model of Strikes.

Problems with existing empirical studies of the asymmetric information theory of strikes are (1) industry-level data are used as proxies for firm-level variables, (2) studies using firm-level data potentially suffer from simultaneity bias, and (3) one of the key predictions of the theory is not tested. The present paper addresses these issues, while estimating a probit model of strike incidence using data comprising 2,459 contracts from Canadian industry for 1964-85 that are merged with firm-specific data. Under a number of alternative specifications, results support the main predictions of the asymmetric information theory of strikes.