Lost Opportunities: Work during High School, Establishment Closures and the Impact on Career Prospec

Since average monthly earnings is driving the reduction in the estimate after the inclusion of the control variables, I narrow my comparison group to students in the same earnings quartile within a given class and the same in-school job industry. [...] The covariates that matter for the magnitude of the estimated effect of a closure are the control for the characteristics of the closing establishments, mainly average monthly earnings and size of the establishment (measured as the number of employees). [...] While the introduction of the individual background characteristics matters little for the size of the estimate, we see that the reduction in the magnitude is indeed driven by the earnings and size controls, indicating that selection on observable individual characteristics is of little concern in 17 practice. [...] In line with the literature, the magnitude of the effect is larger for students in the bottom grade quartile (relative to students in the middle of the grade distribution) and during times of high unemployment. [...] Table 8 shows the results from a pooled regression with separate fixed effects and covariates for students working in industries related and unrelated to the vocational track they are enrolled in.14 The effect of a closure on stable employment is negative in both cases, but roughly three times larger if the industry of the in-school job was related to the vocational track, indicating that the majo.