The following report was provided by a third party verifier, who conducted the verification on behalf of the AEA Data Editor.
Edits to run_all.do
a) Change “
file name.do” to “file_name.do”b) Change “
rdabc_analysis.do” to “rd_abc_analysis.do”Edits to
rd_abc_analysis.doa) Change
${prog_dir}sumstats.doto“{prog_dir}/sumstats.do”(add quotation marks and forward slash “/” (all platforms)
Note by DATA EDITOR: please use forward slash throughout and consistently, this works on any platform in Stata
Tables
Table 1: In the directory for logs, I did not find the file “
sumstats_all.log” as it states in the readme file. I foundsumstats_rdabc.log,sumstats_rd.logandsumstats_rkd.log. I was able to find all values for the variables listed under the RD Sample column in Table 1 though the section order of the output in the log file is somewhat different.Table 2: See attached file “
t2_covariates.xlsx” and the areas that are highlighted in yellow that are different from the manuscript. The differences are not numerous and are minor and do not change the interpretation. (Image provided)Table 3: All numbers from “
t3_cross_section.xlsx” match manuscript.Table 4: See attached file “
t4_heterogeneity.xlsx” and the areas that are highlighted in yellow. Differences mostly in the point estimates at the 3rd decimal place with some differences in standard errors as well.Table 5: See attached file “
t5_otherdata.xlsx” and the areas that are highlighted in yellow. Minor differences driven by sample size changes.Table 6: All numbers in “
table6.xlsx” match manuscript.Table A.1: All numbers in “
tablea1.xlsx” match manuscript.Table A.2: All numbers in “
tablea2.xlsx” match manuscript.Table A.4: All numbers needed to match the manuscript were found in
sumstats_rd.log,sumstats_rdabc.logandsumstats_rkd.log.
NOTE from AEA Data Editor: This is normally a failure - we expect to have clearly marked output enter a table.
Figures: I reviewed all the generated PDF files that contained all figures and they visually matched to the manuscript.