Hello everyone,
I am currently doing my thesis on the subject : Efficiency of government in R&D spendings and I am really a new user of Stata !
In order to do my analysis, I chose to follow a production frontier model based on battle and coelli 95. The model links the dependent variable BERD (business R&D expenditure) with BERDBYGOV (subsidises), GOVERD (spending of the government in public research institution), and HERD (R&D spend in the higher education system).
What I would like to do is to run my frontier regression by making the conditional mean of the inefficiency term u (mean of the truncated-normal distribution) depends on a linear function of other covariates such as the complexity of the government (size), the difficulty of the rules, and the delay it takes to implement a procedure.
I have done this : sfpanel lnberdgdp lnberdgovgdp lngovgdp lnherdgdp, model(bc95) emean( complexity delay rules)
I don't know if it is correct to do this... Can someone help me on this ?
Moreover, in a next step what I really would like to do if the making the conditional mean of the inefficiency term u (mean of the truncated-normal distribution) depends on a linear function of other covariates which would represents the continent in which the country is. As a result, i would like to be able to see if when the country belongs to Africa or Asia, the inefficiency is bigger or smaller than a reference group which could be for instance Europe....
But I have no idea on how i can do this...
Thank you so much for your consideration
Alexandra
I am currently doing my thesis on the subject : Efficiency of government in R&D spendings and I am really a new user of Stata !
In order to do my analysis, I chose to follow a production frontier model based on battle and coelli 95. The model links the dependent variable BERD (business R&D expenditure) with BERDBYGOV (subsidises), GOVERD (spending of the government in public research institution), and HERD (R&D spend in the higher education system).
What I would like to do is to run my frontier regression by making the conditional mean of the inefficiency term u (mean of the truncated-normal distribution) depends on a linear function of other covariates such as the complexity of the government (size), the difficulty of the rules, and the delay it takes to implement a procedure.
I have done this : sfpanel lnberdgdp lnberdgovgdp lngovgdp lnherdgdp, model(bc95) emean( complexity delay rules)
I don't know if it is correct to do this... Can someone help me on this ?
Moreover, in a next step what I really would like to do if the making the conditional mean of the inefficiency term u (mean of the truncated-normal distribution) depends on a linear function of other covariates which would represents the continent in which the country is. As a result, i would like to be able to see if when the country belongs to Africa or Asia, the inefficiency is bigger or smaller than a reference group which could be for instance Europe....
But I have no idea on how i can do this...
Thank you so much for your consideration
Alexandra
