SUR models

[Olive Bat]

Hi.. i am trying to run "sureg" for my analysis .. as i have variables which are correlated .. can someone plz guide me if my command is correct..d1........dn are my dependent variables.. and x1- consists of my independent variable..my independent variable is same for all sets..
sureg(d1 x1)(d2 x1) (d3 x1) (d4 x1) (d5 x1),small dfk
.. plz someone verify if my command is correct..

[Olive Bat]

can someone please tell if Seemingly unrelated regression solves the problem of endogenity or not? Please reply

[Sven-Kristjan Bormann]

The short answer is: No, SURE does not deal with endogeneity but rather accounts for the fact that the same shocks might influence different equations.

The assumption of the model is that error terms ε_ir are independent across time, but may have cross-equation contemporaneous correlations.
Source: https://en.wikipedia.org/wiki/Seemin...ions#The_model

My understanding of endogeneity is as a correlation between the error term and the independent variable or as reverse causality (left hand side variable influences right hand side variable).

See also https://en.wikipedia.org/wiki/Seemin...valence_to_OLS regarding your first question.
If you want to have a more detailed answer, then please a provide a more detailed description of what you want to do and maybe an example of your dataset.

[Olive Bat]

is this a correct command for SUR on STATA
sureg(d1 x1)(d2 x1) (d3 x1) (d4 x1) (d5 x1),small dfk
my explanatory variables are same for all dependent variables..

[Sven-Kristjan Bormann]

It should be correct. I don't know if SUR is the correct model for you. Have you tried out the command?

[Jeff Wooldridge]

In this setup you will just get OLS estimates for each equation. The only reason to use SUR is to test cross equation restrictions.