Dear all,
I was wondering on whether it is possible to use (and whether it makes sense to do in this way) the DFL (DiNardo, Fortin and Lemieux) methods for deriving the counterfactual wage in my sample population.
In particular, I would like to estimate the predicted wage for each worker that would have prevailed if he had been started to work in his own region rather than in the region of current job, given the regional characteristics. Usually, DFL is used for binary outcome (e.g, Union=0,1; Out_Region=0,1) but in my case I would predict the wage considering a categorical/multilevel outcome (regions).
I am interesting in the counterfactual since I want to derive a "Regional premium" as the difference between the current wage in the region of workplace and the region of current residence.
Is there a way to do this with the DFL method?
Thank you in advance
Kind regards
I was wondering on whether it is possible to use (and whether it makes sense to do in this way) the DFL (DiNardo, Fortin and Lemieux) methods for deriving the counterfactual wage in my sample population.
In particular, I would like to estimate the predicted wage for each worker that would have prevailed if he had been started to work in his own region rather than in the region of current job, given the regional characteristics. Usually, DFL is used for binary outcome (e.g, Union=0,1; Out_Region=0,1) but in my case I would predict the wage considering a categorical/multilevel outcome (regions).
I am interesting in the counterfactual since I want to derive a "Regional premium" as the difference between the current wage in the region of workplace and the region of current residence.
Is there a way to do this with the DFL method?
Thank you in advance
Kind regards
Comment