Hello everyone. I hope you are doing well.
I am using a panel data (cross-sectional) from year 2007-2019 and trying to assess a policy effect (secondary policies which is 11) using difference-in-difference (DID) controlled by state-quarter clusters (approx.450 clusters). The outcome variable is emergency visits (count) and the main independent is 11 secondary policies. The covariates are state, year-quarterly, population estimates(state), unemployment rate(state), income ratio (state), gender, insurance type, and other 3 health policies. The main analysis is using a negative binomial regression:
*for policy1-policy11 is actually described separately but I expressed it here as pocily1-policy 11 to save space (e.g., policy1 policy 2....policy11)
I didn't have an issue with applying the above code for age-group analysis. But when I tried to do sub-analysis dividing them by rurality, which has 5 categories, it is not concaving (?).
I wanted to check if first the above nbreg command appropriate in my research question and second, if I want to solve the convergence issue, where should I start and what should I consider? Please let me know if other information is necessary. Thank you.
I am using a panel data (cross-sectional) from year 2007-2019 and trying to assess a policy effect (secondary policies which is 11) using difference-in-difference (DID) controlled by state-quarter clusters (approx.450 clusters). The outcome variable is emergency visits (count) and the main independent is 11 secondary policies. The covariates are state, year-quarterly, population estimates(state), unemployment rate(state), income ratio (state), gender, insurance type, and other 3 health policies. The main analysis is using a negative binomial regression:
*for policy1-policy11 is actually described separately but I expressed it here as pocily1-policy 11 to save space (e.g., policy1 policy 2....policy11)
egen panel = group(state_time)
nbreg edvisits treatment_policy1-treatment policy11 time_policy1-time_policy11 did_policy1-didpolicy11 state time population unemployment insurance female insurance other_policy 1 other_policy2, cluster(panel) irr
nbreg edvisits treatment_policy1-treatment policy11 time_policy1-time_policy11 did_policy1-didpolicy11 state time population unemployment insurance female insurance other_policy 1 other_policy2, cluster(panel) irr
I wanted to check if first the above nbreg command appropriate in my research question and second, if I want to solve the convergence issue, where should I start and what should I consider? Please let me know if other information is necessary. Thank you.
