C. Regression-Adjusted ModelsPanels A and B of Table 2 present basic regression estimates using the BLS ES-202 longitudinal sample of fast-food restaurants. The models presented in this table essentially parallel the main specifications in Card and Krueger (1994). The dependent variable in the first two columns is the change in the number of employees, while the dependent variable in the last two columns is the proportionate change in the number of employees. Following Card and Krueger (1994), the denominator of the proportionate change is the average of first- and second-period employment. Employment changes are measured between FebruaryMarch 1992 and November-December 1992. Columns (1) and (3) include as the only regressor a dummy variable indicating whether the restaurant is located in New Jersey or eastern Pennsylvania. These estimates correspond to the difference-in-differencesestimates that can be derived from row 1 of Table 1. The models in columns (2) and (4) add a set of additional control variables: dummy variables for the identity of the restaurant chain, and a dummy variable indicating whether the reporting unit was a subunit of a multiunit employer
đang được dịch, vui lòng đợi..
