Dear usesrs
I am running an instrumental variable random effects model. In my panel data I have two types of variables: first, some who are measured once per individual, like gender, race, ... and second: some who are measured at several timestamps like, income, drug use, ...
The equation is specified as: y[it] = a + B0X[it] + B1Y[i] + u[i] + e[it]
To avoid endogeneity I have chosen an instrumental variable. So I am unsure, what the first stage equation does look like in stata, as the instrument is measured only once per individual.
Does the equation looks like: z[it] = L0 + L1X[i] + L2Y[it] + e
Or does it look like: z[i] = L0 + L1X[i] + e
In the stata output the G2SLS has 23,870 number of observations and 7,528 Number of groups. In the first stage the number of observations is 23,870. So there are clearly no groups in the first stage.
I am running an instrumental variable random effects model. In my panel data I have two types of variables: first, some who are measured once per individual, like gender, race, ... and second: some who are measured at several timestamps like, income, drug use, ...
The equation is specified as: y[it] = a + B0X[it] + B1Y[i] + u[i] + e[it]
To avoid endogeneity I have chosen an instrumental variable. So I am unsure, what the first stage equation does look like in stata, as the instrument is measured only once per individual.
Does the equation looks like: z[it] = L0 + L1X[i] + L2Y[it] + e
Or does it look like: z[i] = L0 + L1X[i] + e
In the stata output the G2SLS has 23,870 number of observations and 7,528 Number of groups. In the first stage the number of observations is 23,870. So there are clearly no groups in the first stage.
