Hello everyone,
I am not very experienced and I need a bit help.
I am doing cross sectional analysis an I am using one time range for my dependen variable and another time range for the independen variable and for controls to avoid reverse feddback from my dependent variable. This method is very often used.
I am investigating the dynamic raletionship between unemployment and corruption. To see how large unemployment affects corruption and whether higher level of corruption results in a lower unemployment.
To analyse the relationship I estimate one model using unemployment as a dependent variable and corruption as an indepemdent, and in the second model using corruption is the dependent variable and unemployment as the independent.
I got an advice to use the respective past values of the variables as instrument. But I have never done it before and I do not know how to perform this. If you are familiar with that could you please send me an example?
Thanks a lot in advance!
I am not very experienced and I need a bit help.
I am doing cross sectional analysis an I am using one time range for my dependen variable and another time range for the independen variable and for controls to avoid reverse feddback from my dependent variable. This method is very often used.
I am investigating the dynamic raletionship between unemployment and corruption. To see how large unemployment affects corruption and whether higher level of corruption results in a lower unemployment.
To analyse the relationship I estimate one model using unemployment as a dependent variable and corruption as an indepemdent, and in the second model using corruption is the dependent variable and unemployment as the independent.
I got an advice to use the respective past values of the variables as instrument. But I have never done it before and I do not know how to perform this. If you are familiar with that could you please send me an example?
Thanks a lot in advance!
