Skip to content

Latest commit

 

History

History
51 lines (40 loc) · 2.49 KB

README.md

File metadata and controls

51 lines (40 loc) · 2.49 KB
Raw

Solidarity and Fairness in Times of Crisis

Repo with data and code for the paper (Cappelen, Falch, Sørensen, and Tungodden). The paper is published (open access) as:

Alexander W. Cappelen, Ranveig Falch, Erik Ø. Sørensen and Bertil Tungodden (2021). Solidarity and fairness in times of crisis. Journal of Economic Behavior & Organization 186: 1-11. https://doi.org/10.1016/j.jebo.2021.03.017

Organization of data processing

There are raw data in data/raw/, one file downloaded from the JHS COVID-19 database, and one Stata-file from our own survey. The Stata file from our own survey has labels provided by Ipsos.

In data/scripts there is one Stata do-file that reads and make some initial transforms of the raw data and saves the processed Ipsos data in data/processed/ipsos_mmnyt_processed_release.dta.

Organization of analysis

There are 3 .Rmd files in the top level directory that prepares the graphs (treatment_graphs.Rmd), the Figure showing our survey in the context of the pandemic (ipsos_and_JHS.Rmd) and the Romano-Wolf adjusted p-values, mht_interactions.Rmd, which calculates the Romano-Wolf adjusted p-values using a bootstrap approach assumes that the computer to run on has a lot of memory and computing cores (it runs fine with 32GB memory and an AMD 16-Core X3950). If run on a more modest computer, the "ncpus" should be set accordingly.

There are 2 .do files in the top level directory that outputs the regression tables for appendix A based on the main data collection and the secondary data collection on mTurk for the revised version of the paper. Referenced tables from the previous appendix C has been incorporated in appendix A.

All graphs are saved in .pdf format to the ./graphs directory and the regression tables, in .tex format, are saved to the ./tables directory.

The descriptive statistics (S1) and the tables with Romano-Wolf adjusted p-values (S18-S23) have been manually entered from numbers in the log files. The code for (and documentation of) the Romano-Wolf procedure was kindly made available by Michael Wolf (Adjust.Rdata and Wolf-ReadMe.txt), at his website.

mTurk supplement

For a revision, we conducted a supplementary data collection on mTurk, the data and analysis is kept apart in separate files, the analysis is in the Stata do-file mturklab_analysis.do outputting tables/mturklab_indic.tex based on data in the data/processed directory.