Learning to kill: Why a small handful of counties generates the bulk of US death sentencess


Frank R. Baumgartner, University of North Carolina, Chapel Hill
Janet M. Box-Steffensmeier, Ohio State University
Benjamin W. Campbell, Ohio State University
Christian Caron, University of North Carolina, Chapel Hill
Hailey Sherman, University of North Carolina, Chapel Hill

PLoS-ONE 15 (10): e0240401

This page contains replication materials and Supplemental Information for our article. These are also available at the journal website, here. This page gives more intuitive and useful file names. The jounal web page lists the same files only with names S1 File, S2 File, etc.

To replicate the analysis, download each of the following files to a directory. Within that directory make a folder called "Output" as well. This will be where Table 2 is replicated.

Alternative ZINB models and constructing the homicide database. This is the Appendix to the paper, a word document. (S1 File on the PLoS website).

Main Stata do file containing commands to generate the analysis in the article other than those associated with Figure 1 and Table 2. (S7 File on the PLoS website).

Main Stata dta file with county-year level data on all death sentences and covariates (S6 File on the PLoS website)

Stata dta file with data on cumulative death sentences by county. (S8 File on the PLoS website).

Stata dta file with Actual, Predicted, and Simulated Counts. This is a Stata dta file needed to replicate Figure 2 in the paper. (S2 File on the PLoS website)

R command file to generate Figure 1. (S10 File on the PLoS website).

RDS file needed to recreate Figure 1. (S3 File on the PLoS website).

Bootstrap function called by Fig1.R in generating Figure 1. (S4 File on the PLoS website).

RDS file with data associated with Table 2. (S5 File on the PLoS website).

R command file to recreate Table 2 (S9 File on the PLoS website).

(Last updated, January 15, 2021 ).