README Document for Reproducing Results in

"Prices, profits, proxies, and production"

Victor H. Aguiar [email protected]

Nail Kashaev [email protected]

Roy Allen [email protected]

Data Sources

The paper uses one data set from

• Epple, Dennis, Brett Gordon, and Holger Sieg (2010). A new approach to estimating the production function for housing, American Economic Review, 100 (3), 905–24. https://doi.org/10.1257/aer.100.3.905

and Julia code from

• Kim, Youngseok, Peter Carbonetto, Matthew Stephens, and Mihai Anitescu (2020). A fast algorithm for maximum likelihood estimation of mixture proportions using sequential quadratic programming, Journal of Computational and Graphical Statistics, 29 (2), 261–273. https://doi.org/10.1080/10618600.2019.1689985

Software

A version 1.6.4 of the Julia programming language was used in coding the analysis files. For details about how to install Julia on different platforms and make Julia programs executable from the command line see https://julialang.org/downloads/platform/. After installation of Julia 1.6.4. run using Pkg and Pkg.instantiate() in the Julia terminal after setting the replication folder as the main one.

Some simulations use KNITRO 12.3.

Hardware

The code was run on Mac mini (M1, 2020) with 16 Gb of RAM

Content

• Application -- the folder contains the analysis files to replicate the results in Appendix C.

• Simulations -- the folder contains the analysis files to replicate the results in Online Appendix A.

• Manifest.toml and Project.toml -- toml files with all necessary Julia packages.

Below I describe the content of every folder.

Application

data

• data_cleaned_8.csv -- processed data from Epple et al. (2010)

• Pittsburgh_post1995.txt -- original data from Epple et al. (2010)

• US Zip Codes from 2013 Government Data.txt -- data used in construction of the zip-code using geographical coordinates of houses

results

This folder contains estimated output levels and output prices.

• output_level_8_4.csv -- estimates of output level

• output_level_8_4.csv -- estimates of output price

tables and graphs

This folder contains all tables and figures from Appendix C.

• Fig_x.pdf -- Figure x from Appendix C (x in {6,7,8})

• Tablex.pdf -- Table x from Appendix C (x in {1,2})

root files

• application_functions.jl -- the functions used in application_main.jl

• application_main.jl -- this code generates the output levels and output prices (output_level_8_4.csv and output_price_8_4.csv)

• mixSQP.jl -- the procedure for estimation of finite mixtures described in Kim et al. (2020)

• preparing_data_elbow.jl -- this code cleanes the data from Epple et al. (2010) and constructs markets (data_cleaned_8.csv)

• tablesandgraphs.jl -- this code generates all tables and figures in tables and graphs folder

Simulations

results

This folder constains simulation results used in construction of table in Online Appendix A

• thetap_DGP_n_epanechnikov_4.csv -- simulation results for estimated \pi and \rho for sample size n with the epanechnikov kernel (n in {500,1000,1500})

tables and graphs

This folder constains all table in Online Appendix A

• Table_param_epanechnikov_4.csv -- Table for estimated param with the epanechnikov kernel (param in {\pi, \rho})

root files

• mixSQP.jl -- the procedure for estimation of finite mixtures described in Kim et al. (2020)

• simulation_functions.jl -- the functions used in simulation_main.jl

• simulation_main.jl -- this code generates all simulation outputs in results folder

• simulation_all.sh -- this script runs simulation_main.jl with different input parameters (e.g., kernel, sample size)

• tablesandgraphs.jl -- this code generates all tables and figures in tables and graphs folder