On April 1, 1992, New Jersey’s minimum wage was increased from$4.25 to $5.05 per hour, while the minimum wage in Pennsylvaniastayed at $4.25 per hour. Energetic students collected data on 410fast food restaurants in New Jersey (the treatment group) andeastern Pennsylvania (the control group). The “before” period isFebruary 1992, and the “after” period is November 1992. Using thesedata, we will estimate the effect of the “treatment,” raising theNew Jersey minimum wage on employment at fast food restaurants inNew Jersey (i.e., H_0:?=0 versus H_A:?<0). It is easier and moregeneral to use the regression format to compute thedifferences-in-differences estimate using sample means. Let y=FTEemployment , the treatment variable is the indicator variable NJ=1if observation is from New Jersey, and zero if from Pennsylvania.The time indicator is D=1 if the observation is from November andzero if it is from February. (a.)Write out the regression equation.(b)Report the least squares estimates . (c)At the ?=.05 level ofsignificance the regression region for the left tail test in abovehypotheses is t?-1.645, what is your conclusion? (d)As withrandomized control (quasi) experiments it is interesting to see therobustness of the result from (c). Please, add indicator variablesfor fast food chain and whether the restaurant was company-ownedrather than franchise-owned. These changes alter the DID estimator?(e)Please, add indicator variables for geographical regions withinthe survey area. These changes alter the DID estimator?
