Hello everyone,
I am using the -xtabond2- command. I am using 70 countries over a period of 14 years. I am struggling with two things.
1. Does my main independent variable need to be endogenous? If I can assume based on economic reasoning this to be strictly exogenous and use it under instrumental variable system, should there be any issue?
2. I am struggling to run estimations across sub samples. For instance, in my list I have OECD, developed, developing countries. I want to check the results across each groups. Can you help me with a sample code (of any one), how to incorporate the "if" option in xtabond2 command.
Kind regards