Forum: help

RE: micEcon and micEconCES for panel data [ Reply ]
By: Arne Henningsen on 2022-03-09 20:45

I am happy to hear that you find my/our R packages useful :-)

There are a lot of challenges with estimating a CES function with panel-data methods, for instance because the CES is non-linear in parameters, the assumptions for random-effects panel-data estimators are usually largely violated, and fixed-effects panel data estimators generally give inconsistent estimates when applied to non-linear models.

You might find the following book chapter (a revised version of a working paper about the micEconCES package that you may know) useful: https://doi.org/10.4337/9781788976480.00030

You could also take a look at my "Introduction to Econometric Production Analysis with R" (for other functional forms than CES that are more suitable for panel data), which you can download for free (if you set the price to zero) or buy (at any price you find reasonable) at: https://leanpub.com/ProdEconR

As a Translog function is linear in parameters, you can estimate it as a fixed-effect panel data model, e.g., with the translogEst() function of the "micEcon" package or 'manually' with the "plm" package.

micEcon and micEconCES for panel data [ Reply ] By: Tallys Feldens on 2022-03-08 20:39	[forum:49363]
Dear developers, Thank you all for this wonderful package. I'd like to ask how should I run translogEst and cesEst for panel data. Besides setting my data as pdataframe, there is something else I should do? Thanks again.