re: st: Panel data with simultaneous equations

[Christopher Baum <kit.baum@bc.edu>]

<>
I am working on an application to model the impact of social media on overall online media consumption. 

To that end, i have a panel of users.
I am looking at the following specification :

social_{i,t} =  f_1( nonsocial_{i,t}, x_{i,t})
nonsocial_{i,t} =  f_2( social_{i,t}, x_{i,t})

where :
social_{i,t} i the social media of subject ``i" at time ``t"
nonsocial_{i,t} i the non-social media of subject ``i" at time ``t"
x_{i,t} are the  co-variates of subject ``i" at time ``t".  i could also have time fixed co-variates. 


I am looking at a linear model, where both f_1, f_2 are linear. 

Is there a way i can estimate such a model using stata? 


Yes, of course, as long as you have some plausible exclusion restrictions. If the X variables are the same in each equation, neither
equation is identified.

You could use official xtivreg or Schaffer-Stillman xtivreg2, from SSC, to estimate this model using fixed effects in the absence
of time-invariant covariates.

Kit

Kit Baum   |   Boston College Economics and DIW Berlin   |   http://ideas.repec.org/e/pba1.html
An Introduction to Stata Programming   |   http://www.stata-press.com/books/isp.html
An Introduction to Modern Econometrics Using Stata   |   http://www.stata-press.com/books/imeus.html


*
*   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/