i want to investigate the relationship between unemployment and corruption.
i used cross sectional data, with one time range for a depend variables and with another time range for independent and control variables. To avoid reverse feedback.
In the first regression I used unemployment as independent variable, to see whether large unemployment results in large corruption levels.
in the second regression, i used corruption as independent variable ot see, whether high level of corruption results in low unemployment.
IN both regression I used robust Standard errors
Reg1 Corruption (y1)=constant+1Unemployment+2Unemploymentsquared+c ontrols
Unemployment (y2)= 1Corruption+2corrultionsquared+controls
I wanted to use sureg ( seemingly unrelated regression) to check whether the errors are correlated between this two models. However, I'm not sure if it will work, I'm not so familiar with this method. I can't find enough info about sureg.
1)Can I use it also if i have different independent variables in the equations?
2) How can I use it with robust standard errors?
3)Can I use sureg if I have squared variables?
4)As I understand I should not use all controls for both equations?
5)If after sureg Breusch- Pagan test will be not rejected, how i can interpret the results?
Have a good day, thanks in advance!
i used cross sectional data, with one time range for a depend variables and with another time range for independent and control variables. To avoid reverse feedback.
In the first regression I used unemployment as independent variable, to see whether large unemployment results in large corruption levels.
in the second regression, i used corruption as independent variable ot see, whether high level of corruption results in low unemployment.
IN both regression I used robust Standard errors
Reg1 Corruption (y1)=constant+1Unemployment+2Unemploymentsquared+c ontrols
Unemployment (y2)= 1Corruption+2corrultionsquared+controls
I wanted to use sureg ( seemingly unrelated regression) to check whether the errors are correlated between this two models. However, I'm not sure if it will work, I'm not so familiar with this method. I can't find enough info about sureg.
1)Can I use it also if i have different independent variables in the equations?
2) How can I use it with robust standard errors?
3)Can I use sureg if I have squared variables?
4)As I understand I should not use all controls for both equations?
5)If after sureg Breusch- Pagan test will be not rejected, how i can interpret the results?
Have a good day, thanks in advance!
