Dear Stata community,
I am working on the effect of land tenure security on farm investment in Ghana. My tenure security variable is binary (I.e., 1 if secure and 0 if otherwise) and potentially endogenous in the farm investment equation. The farm investment is consist of long term investment in land improvement technologies (1 farmer invest in long-term land improving techniques and 0 if otherwise) and short-term land improving techniques (1 if farmer invests in short-term land improving techniques). The land improving techniques are also potentially interrelated. Thus, the problem is to control for the potential endogeneity of land tenure security, and estimate is impact on an interdependent land improving techniques. My question is: is there any Stata model to jointly control for endogeneity and the interrelated techniques while estimating the impact of tenure security on farm investment. I am aware that the biprobit and its extension (I.e., multivariate probit ) control for potential interdependencies between choices. But am not sure if such models can control for endogeneity. Moreover, it seems plug-in of the first stage residuals from probit model of land tenure security into biprobit or multivariate probit of farm investments will be inappropriate according to Woolridge (2002).
Thank you.
Abdallah Abdul-Hanan.
I am working on the effect of land tenure security on farm investment in Ghana. My tenure security variable is binary (I.e., 1 if secure and 0 if otherwise) and potentially endogenous in the farm investment equation. The farm investment is consist of long term investment in land improvement technologies (1 farmer invest in long-term land improving techniques and 0 if otherwise) and short-term land improving techniques (1 if farmer invests in short-term land improving techniques). The land improving techniques are also potentially interrelated. Thus, the problem is to control for the potential endogeneity of land tenure security, and estimate is impact on an interdependent land improving techniques. My question is: is there any Stata model to jointly control for endogeneity and the interrelated techniques while estimating the impact of tenure security on farm investment. I am aware that the biprobit and its extension (I.e., multivariate probit ) control for potential interdependencies between choices. But am not sure if such models can control for endogeneity. Moreover, it seems plug-in of the first stage residuals from probit model of land tenure security into biprobit or multivariate probit of farm investments will be inappropriate according to Woolridge (2002).
Thank you.
Abdallah Abdul-Hanan.
Comment