This repository provides replication files for:

COVID-19 and pro-sociality: How do donors respond to local pandemic severity, increased salience, and media coverage?

Maja Adena (WZB) and Julian Harke (WZB)

Description of files

Experiment - oTree code

For our sessions we use oTree. The implementation of our experiment can be found in otree.

The required Python packages can be installed via¹:

pip install -r requirements.txt

Note: oTree 3.4.0 requires Python 3.7.x - 3.8.x.

The experiment can be started via:

otree runprodserver

Visit localhost:8000 to start the experiment.

Analysis - Stata

For the analysis we use Stata.

Data

The folder stata_files/data contains two datasets:

• corona_donation_experiment_anonymized.dta: This Stata file contains the anonymized survey and experimental data.
• UTLAdata_lnskew.dta: This Stata file contains region data at UTLA level.

The description of the variables can be found in 3_labels.do.

Do-files

The folder stata_files/do_files contains the code to replicate our results.

The required Stata packages can be installed via:

ssc install estout
ssc install erepost
ssc install reghdfe

To replicate our analysis run 0_MAIN.do².

Replication output

Results will be stored in the replication output folder as follow:

• prepared_panel_data.dta: This Stata file contains all data from merging corona_donation_experiment_anonymized.dta with UTLAdata_lnskew.dta as well as edited data (for code see dofiles 1_merging.do and 2_variables.do).

• Tables_descriptives.rtf and Tables_descriptives.tex: These files contain tables of descriptive output for the paper.

• Tables_regressions.rtf and Tables_regressions.tex: These files contain regression tables presented in the paper and regression tables for the supporting hypotheses as described in Online Appendix, section D.

Sources of data

Show data sources

Source	License and Terms of Use
nomis	Attribution required, Open Government Licence v3.0
COVID-19 UK Data	Attribution required, Open Government Licence v3.0

Footnotes

1. We recommend installing the required Python packages inside an virtual environment. ↩

2. The do-files 5-rtf_regressions.do and 5-tex_regressions.do create identical output but in either rtf or TeX format. The same applies to do-files 6-rtf_descriptivestats.do and 6-tex_descriptivestats.do. ↩