Dear Stata Users,

I need to apply instrumental variable (IV) cross sectional (country level) regression in the following setting:

What is given:
(DEP) continuous dependent variable
(CAT) categorical independent variable/regressor that needs to be instrumented (like level of democracy = 0, 1, 2)
(IV) instrumental variable candidates that are continuous variables in all cases. (like percentage of people in the population with some appropriate characteristics)
Number of observations = around 45
My questions:

Is it fine/statistically OK to run this STATA command: ivregress 2sls DEP ( CAT = IV ) , robust first
  • or I need to do some transformations or run other models in STATA due to the specifics of the regressor (categorical) and IV (continuous)?
  • do you know statistical literature where exactly this case is covered?
Best regards,
Iva