How to make a pie: reproducible research for empirical economics and econometrics

Abstract

Empirical economics and econometrics (EEE) research now relies primarily on the application of code to data sets. Handling the workflow that links data sets, programs, results, and finally manuscript(s) is essential if one wishes to reproduce results. Herein, we highlight the importance of “reproducible research” in EEE and propose three simple principles to follow: organize your work, code for others, and automate as much as you can. The first principle, “organize your work”, deals with the overall organization of files and the documentation of a research workflow. “Code for others” emphasizes that we should take care in how we write code that has to be read by others or later by our future self. Finally, “automate as much as you can” is a proposal to avoid any manual treatment and to automate most, if not all, of the steps used in a research process to reduce errors and increase reproducibility. As software is not always the problem and will never be the solution, we illustrate these principles with good habits and tools, with a particular focus on their implementation in most popular software and languages in applied economics.

Replaces

Valérie Orozco, Christophe Bontemps, Elise Maigné, V. Piguet, A. Hofstetter, Anne Lacroix, F. Levert, and J.M Rousselle, “How To Make A Pie: Reproducible Research for Empirical Economics & Econometrics”, TSE Working Paper, n. 18-933, July 2018.

Reference

Valérie Orozco, Christophe Bontemps, Elise Maigné, V. Piguet, A. Hofstetter, Anne Lacroix, F. Levert, and J.M Rousselle, “How to make a pie: reproducible research for empirical economics and econometrics”, Journal of Economic Surveys, vol. 34, n. 5, December 2020, pp. 1134–1169.