[R] IV estimation

[kebrab67]

Hello, 
I have a set of 100 variables with 1560 observations. I did an O.L.S
regression of three of these variables on a fourth. But there are problems
of endogeneity... So I look in my dataset for instruments to do an IV. I
can't find a good instrument because their correlation with my endogeneous
variables are too low. But I see that when I create a combined variable
composed of 12 variables of the dataset my correlation is much stronger.
Can I use such a combined variable to estimate my equation with 2SLS, or
GMM.

--
View this message in context: http://r.789695.n4.nabble.com/IV-estimation-tp4599559.html
Sent from the R help mailing list archive at Nabble.com.