We study inferences about the dynamics of labor adjustment obtained by the “gap methodology” of Caballero and Engel  and Caballero, Engel and Haltiwanger . In that approach, the policy function for employment growth is assumed to depend on an unobservable gap between the target and current levels of employment. Using time series observations, these studies reject the partial adjustment model and find that aggregate employment dynamics depend on the cross-sectional distribution of employment gaps. Thus, nonlinear adjustment at the plant level appears to have aggregate implications. We argue that this conclusion is not justified: these findings of nonlinearities in time series data may reflect mismeasurement of the gaps rather than the aggregation of plant-level nonlinearities.