I'm estimating the following fe 2sls model using xtivreg
I'm trying to test the predicted values of ppe for strict exogeneity using Wooldridge (2010) well-known test. Can I use the lead value of the instrument to predict lead ppe in a model where I also use the instrument in period t to predict ppe in period t. I think this is ok because even though you are using the same set of data, different versions of the instrument in period t and t+1 are technically different instruments.
would the following equation be correct?
Code:
xtivreg Mathprofrate_z (ppe = decreaseenroll) enrollment enrollsquared debtratio teachersalrealinfout per_freeclean per_min PER_LEP EMdisratecomp i.Year if Year >=2006, fe vce(cluster District) first
would the following equation be correct?
Code:
xtivreg Mathprofrate_z (ppe F.ppe = decreaseenroll F.decreaseenroll) enrollment enrollsquared debtratio teachersalrealinfout per_freeclean per_min PER_LEP EMdisratecomp i.Year if Year >=2006, fe vce(cluster District) first
