st: xtivreg2

[natasha agarwal <agarwana2@googlemail.com>]

Dear Statalist,

I have a question relating to the use of xtivreg2.

I am estimating a cobb-douglas production function where I am regressing

Y = L+K+CF+FDI+CF*FDI+year dummies+error

I believe that cash flow (CF)and output (Y) in the above regression
have reverse causality, i.e. CF determines Y but Y could also
determine CF.

So I thought instrumenting CF would be the best idea to break the
reverse causality.

To do so I first differenced the above equation and then used lags
levels of CF of year 2 and year 3 as the instrument.

However, I don't know how would one do the instrumenting when the
endogenous variable is interacted with the exogenous variable??

I would be grateful if anyone could help me on this.

Many thanks
Natasha
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/