Persistent high unemployment has fostered the employment persistency hypothesis according to which employment changes are driven by unanticipated shocks whereas anticipated shocks that potentially could change employment are absorbed by wage changes. Empirical tests of the persistency hypothesis fail to distinguish between the properties of shocks and endogenous propagation mechanism causing persistency. This paper develops a new test strategy by explicitly distinguishing between these two factors. The methodology is applied to the manufacturing sector in Denmark, and some support in favor of an endogenous propagation mechanism causing employment persistency is found.