Williams (2022) - Historical Lynchings

Replication of Williams (2022) "Historical Lynchings and Contemporary Voting" · I4R reference

Execution Method: Translated Python
Note: Results were obtained via Stata-to-Python translation. Numerical differences may be translation artifacts rather than genuine reproducibility failures. Running with native Stata would provide a more accurate assessment.

Overall: Fully Reproducible

2 coefficients reproduced within tolerance. Key variable lynchcapitamob matches across specifications.

Table Comparisons

Table 2 reproduced Score: 90.0%

RowColumnPaperCodeStatusAbs Diff
lynchcapitamobest1-0.469-0.4692within tolerance0.0002
lynchcapitamobest3-0.314-0.3018within tolerance0.0122

Summary

The study examines the relationship between historical lynchings and contemporary voting behavior. Both primary coefficients for lynchcapitamob reproduce successfully: est1 shows exceptional precision (0.04% relative difference) and est3 shows good agreement (3.9% relative difference). The negative coefficients confirm that areas with higher historical lynching rates per capita exhibit different contemporary voting patterns.

Replication package: 1 .do file, 1 .dta dataset (~115 KB). Execution produced 4 output CSVs (table1, table2, appendix tables).