Dear all,
I am struggling with the following puzzle. I have an (unbalanced) panel and want to estimate the treatment effect Di,t (binary) on a continuous outcome variable Yi,t conditional on a number of controls Xit. I have good reasons to think that treatment is endogenous (selection linked to earlier realisations of Yi,t ). However, I also believe that I have potential instruments Zi,t (continuous) for Di,t . In a cross-sectional setup I would follow Wooldridge (2010) which more or less goes as follows:
Thanks!
--
Reference:
Wooldridge, J. M. (2010). Econometric analysis of cross section and panel data. MIT press.
I am struggling with the following puzzle. I have an (unbalanced) panel and want to estimate the treatment effect Di,t (binary) on a continuous outcome variable Yi,t conditional on a number of controls Xit. I have good reasons to think that treatment is endogenous (selection linked to earlier realisations of Yi,t ). However, I also believe that I have potential instruments Zi,t (continuous) for Di,t . In a cross-sectional setup I would follow Wooldridge (2010) which more or less goes as follows:
- Probit model: Di as a function of Zi and Xi.
- Obtain fitted values, D_hati
- 2SLS model: Yi as a function of D_hati and Xi
Thanks!
--
Reference:
Wooldridge, J. M. (2010). Econometric analysis of cross section and panel data. MIT press.
Comment