Dear users,
I appreciate your help combining a sample selection model with an endogenous treatment model. My data consists of the top 100 outsourcing contracts implemented between 1996 and 2005, and I am trying analyze the impact of the contract type - fixed price or variable price - on the operational efficiency of the firm that has engaged in the outsourcing contract.
Thus, two selection mechanisms are at work. The first selection bias results from observability of sample variables only for the firms that outsource. The second selection rule arises from endogenous contract type included on the right hand side of the operational efficiency equation. Thus, the following simultaneous three equation model is built up:
(1) Efficiency = f(covariates, contract) estimates the impact of endogenous contract type on efficiency
(2) Contract = f(covariates) estimates the drivers of contract choice
(3) Selection = f(covariates) estimates the observation mechanism or decision to outsource
Efficiency and Contract are observed only for firms that outsource, i.e. Selection=1.
I don't quite know what STATA functions I would use to extend Heckman's two step estimator of the univariate selection model to this bivariate selection rule. How would I obtain the inverses of Mill's ratio for equations (2) and (3) that may be used in (1)? Also, how would I proceed with full information maximum likelihood (FIML) estimation in this case?
Thanks in advance for your time with this.
Regards,
Deepa
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/