Dr. Patrick Irungu, an AGRODEP member since 2011, recently received a STATA question from one of his MSc students at the University of Nairobi, Kenya. The student is using STATA to run a profit function under the seemingly unrelated regression (SUR) model, but has run into a problem where STATA is dropping some of the constraints. The student describes the problem as follows:
"I am working on "Demand Analysis for Antibiotics in Poultry Production" for my MSc thesis work. I intend to estimate a normalized restricted profit function in a translog form and input share equations simultaneously in STATA using a Seemingly Unrelated Regression (SUR) model. The use of profit function approach in estimating input demand requires the profit function to satisfy some conditions which are set as constraints during the estimation. After the estimation of the simultaneous equations, an individual and a global test for the conditions (constraints) are conducted to find out if the results are consistent with the maintained hypothesis.
I have already estimated the profit and the input share equations in STATA. However, after conducting the test for the constraints, some of them are being dropped such that there are no results for the particular constraints that have been dropped. This means one can't tell whether the particular conditions hold or not."
Have you run into a similar problem using STATA? If you have advice for Dr. Irungu and his student on how to fix this issue, please contact Dr. Irungu directly and also share your insight on the blog!