I need to estimate a panel probit model using probability weights
(pweights) -- xtprobit allows the use of pweights only with the use of the
population-averaged model. I dont know much about the model and its
difference from the random effects probit -- can anyone suggest any
references / explanations? Thanks!
I have an endogenous variable, also 0/1 in the probit model -- with
cross-section data the endogeneity can be handled adequately by bivariate
probit. Any ideas on how to correct for this in panel data? I'd be very
grateful if anyone has a program that implements this correction in the
panel and is willing to share it.