• Stata code for Ackerberg-Caves-Frazer (2006) estimation: acf_bootstrap_code.zip

• Sample dataset to run this code on: sample.zip

Notes

This code implements a value-added production function, estimated per the approach in Akerberg-Caves-Frazer's 2006 working paper. Please check Dan Ackerberg's website for more recent versions of the paper, where the approach may have been revised.

Unzipping the code will yield two files: acf_solution.do & bootstrap_code.do. Modify bootstrap_code.do to point to the relevant data file and also to acf_solution.do (you would need to suitably change directory/data names). You can change number of bootstrap replications from the default (100) by modifying the code.

The bootstrap samples do "block bootstrapping" i.e. entire timeseries for each plantid is sampled in blocks, which allows the errors to be correlated within plants.

Users would need to define variables consistent with the variable names used here or suitably alter the code. lnva is log value added, lnb is log (blue collar employment), lnw is log(white collar employment) and lnm is log(real material cost).

You would need to download and install amoeba.ado