For this purpose, we run a model of the outcome variable on a constant, the treatment dummy, year dummies, and interactions between the treatment and year dummy variables including only pre-intervention periods.
To test for differences in the change in reporting rate in spring 2007 among areas, we included an interaction term between area and the dummy year variable described above.
Controls include gender, marital status, age dummies, education dummies, a full set of age-education interactions, region dummies, year dummies, log trade volumes, and urban versus rural area dummy.
b Each equation also contains age, age squared, skill controls (share of native population with degree, with completed school, still in education), region dummies, year dummies and 2-digit occupation dummies.
For this experiment, we use our preferred model with state and year fixed effects and six dummies for years before/after abolition.
The dummy for year before repeal is positive, suggesting that couple formation increased in transition states several months before CLM was repealed.
Regression controls include dummies for year age 18, survey year, age, birth state, sex, race, and ethnicity.
Year dummies capture the year-specific mean shifts.
OLS regressions include dummies for year of entry, state of residence at entry and year of the survey; IV regressions include dummies for year of predicted entry, state of residence at age 14 and year of the survey.
Regressions include country time quarter dummies and year dummies.
Moreover, we add industry dummies and year dummies to control for the industrial fixed effect and dynamic changes in the macroeconomic environment.
