Welfare decomposition: technical efficiency about cp2015 HOT 4 CLOSED

Ah yes, thanks. They take estimates for the change in trade costs from a gravity estimation

from cp2015.

Hi @zauster ,

Thanks for the question. This derivation is not presented in any paper. I've just followed the derivation presented by Caliendo and Parro (2015) in Appendix B. The technical efficiency term is not presented in the original formula (Equation 16) because they assume that there is no variation in the $d_{ni}^j$ variable. If we assume a change in this variable, we have to modify the Equation B2 to
$$\mathop{d}\ln P_n = \sum_{j = 1}^J \alpha_n^j \sum_{i = 1}^N \pi_{ni}^j \left(\mathop{d}\ln c_i^j + \mathop{d} \ln d_{ni}^j + \mathop{d} \ln {\tilde{\tau}_{ni}^j} \right).$$

Then, follow the derivation presented in Appendix B to derive a new version of Equation 16 with this additional term.

Please let me know if you have any questions about my explanation.

Best regards,

Paulo

from cp2015.

Thanks, that explanation makes it perfectly clear.

Have you already used this technical efficiency term in your work? Because my next question would be how do you get/estimate a value for the change in trade costs (a ratified free trade agreement, abolished non-tariff measures)?

from cp2015.

Hi @zauster ,

I think this paper conducts an experiment similar to the one you are interested in: https://www.ifo.de/DocDL/wp-2016-219-aichele-felbemayr-heiland-welfare-ttip.pdf

It is possible to set the relative change of the variable $d_{ni}^j$ after you have estimated the change in trade costs.

from cp2015.