Replication Data for: "Genocidal Consolidation: Final Solutions to Elite rivalry"

Replication files “Genocidal Consolidation: Final Solutions to Elite rivalry” E van der Maat 16-12-19 The paper contains three analyses. Each analysis has its own replication folder. First analysis (genocidal consolidation onset) — IO_GC_replication_I The first analysis has 3 .R files, one Stata .do file and three data files. To replicate the models of the first analysis, first run: 1) functions.R 2) two-stage probit.R Then run the main latent.R file to replicate the models in the paper; it contains: • replication of Table 4 in the paper • replication of Tables A.4 and A.10 of Appendix C and G Next, run the analysisI.do file; it contains: • replication of the crosstabs in Table 3 (p 30) • replication of the effect estimates (p31) • replication of Table A.2 of Appendix C Second analysis (elite purges) — IO_GC_replication_II The second analysis folder has a single stata .do file and six data files. To replicate the models, run the analysisII.do file; it contains: • replication of Crosstab Figure 3 (p 35) • replication of Table 5 • replication of Table A.5 and A.6 of Appendix D • replication of Table A.11 of Appendix G Third analysis (leader fates) — IO_GC_replication_III The third analysis folder contains a single stata .do file, two data files, and a log file. To replicate the models, run the analysisIII.do file; it contains: • replication of man analysis (Figure 4: leader fates; p 41) • replication of Rosenbaum sensitivity analysis (footnote 136 & 137) • replication of balance checks for various specifications Table A.7 of Appendix E • replication of leader propensity scores and matches (Tables A.8 and A.15—A.25) • replication of alternative specifications (Table A.9 of appendix E) • replication of balance checks for various HI specifications (Table A.12 of Appendix G) • replication of alternative specification with HI (Table A.14 of appendix G) Note that this file may take a very long time to run because of a total 150,000 bootstraps. I’ve added a log for easy reference of outcomes. To check replication results it’s probably easiest to make sure the code works on your machine; then run the file with a log; and check the log when it’s finished running. Good luck!