Replication policy

It is an integral and essential part of scientific investigation that results are easily replicated. We will only publish papers if they are accompanied by a complete replication package.

Such a package consists of:

  • a readme file describing the contents of the package and the version of the software it was written for and the respective versions of extra packages
  • the data used in the analysis, including a complete, transparent and precise documentation describing all variables. If the data is provided in a proprietary format, such as in STATA format, then an ASCII file should also be provided
  • all code necessary to re-produce the tables and figures in the paper as well as supplements and appendices, including the code used to run simulation/Monte Carlo studies (the authors are requested to set a seed so that the exact numbers that are reported can be obtained); if the code uses additional packages that are not part of the standard distribution, then these packages should be included so that the code still runs after many years.

Ideally, but not necessarily, it should include a makefile, so that all results can be replicated from the provided data and code using the make utility. This package will be posted alongside the article as Supporting Information.

Submission indicates that you license users to download, copy, and modify this material. Users of the material must acknowledge all authors as the original creators and The Econometrics Journal as the original publishers. The editorial board of The Econometrics Journal reserves the right to refuse publication of those papers whose authors do not comply with these requirements.

The editorial board of The Econometrics Journal recognises that some issues relating to data and programs may arise. Requests for exemptions from providing materials or restricting their usage must be stated clearly at first submission. The editorial board will then decide whether the submission can be considered.

The Econometrics Journal discourages papers that rely on completely proprietary data. As with other practical difficulties, authors who feel that their paper might constitute an exception should forward the case to the editorial office. Papers that are based on data that are only available at some cost (either monetary or because access to the data is limited, e.g., at a National Statistics office) and can therefore not be included in the package may in exceptional cases be considered. However, the authors will then still be required to provide all programs and instructions a researcher would require to replicate their results once the researcher has access to the data. The source of the data must be described in enough detail so that researchers can apply to obtain access to them. The authors will also be required to supply code that simulates data, as much as possible mimicking the actual data that were used in the application, as well as a file running the analysis on the simulated data. Alternatively, synthetic data can be provided.