I am new to the estimation of Stochastic Frontier Analysis. My issue is as follows I have a panel data set for 53 districts with 46 observations. My dependent variable is output and there are seven independent variables (Labour fertilizer etc). I need to run a stochastic frontier model to calculate total factor productivity using a translog function. My question is that I need to represent the separate districts using intervention analysis (dummy variables which take the value 1 for a district and 0 otherwise). However the number of dummies is a lot and leads to 53 variables in the model. Is generating dummies the correct way to estimate the model or is there any alternative?
-
Login or Register
- Log in with
