Data Source
| Below is a list of all datasets used in the replication files. Please see each dataset’s codebook for more information about that file. data/38651-0001-Data.rds This is a subset of LEMAS data that is used to compare race between LEMAS and L2. United States Department of Justice. Office of Justice Programs. Bureau of Justice Statistics. Law Enforcement Management and Administrative Statistics (LEMAS), 2020. Inter-university Consortium for Political and Social Research [distributor], 2023-03-07. https://doi.org/10.3886/ICPSR38651.v1
data/aa_combined_with_crime.rds Data used to perform balance tests in Appendix J. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/all_census_tracts_agg_in_l2.rds L2 data for all civilians aggregated to the Census tract level. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/chicago/chicago_voters_party_by_beat.rds Civilian party and demographics information at the district-beat-sector level. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/chicago/overview_data_main.csv Total number of outcomes for each officer group in our data. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/chicago/per100shifts_arrests_main.csv Descriptive statistics on the number of arrests made by each officer group on each civilian group. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/chicago/per100shifts_force_main.csv Descriptive statistics on the number of force events made by each officer group on each civilian group. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/chicago/per100shifts_stops_main.csv Descriptive statistics on the number of stops made by each officer group on each civilian group. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/chicago/units_and_l2.rds Officer-level information in Chicago, including the officer’s race, party, district, and the number of L2 civilians in that district. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/data_with_lemas_for_replication_archive.rds Officer-level race and agency data to use when comparing L2 race vs LEMAS race. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/houston/analyses_results_main_20240216_padj Regression results in Houston that analyze the enforcement decisions of officer groups within each MDSB. Weights based on the within-MDSB prevalence of each group are used to obtain unbiased estimates of the average treatment effect for additional details on estimation). Standard errors are clustered by officer. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/houston/houston_data_for_beat_analysis Civilian voter information in Houston with latitude and longitude for their address, and their identified party and race. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/houston/houston_officers_deidentified Deidentified data on Houston officers with only their L2 probability match, unit assignment, race, and party. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/houston/overview_data_main Total number of outcomes for each officer group in our data. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/houston/per100shifts_arrests_main Descriptive statistics on the number of arrests made by each officer group on each civilian group. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/houston/per100shifts_force_main Descriptive statistics on the number of force events made by each officer group on each civilian group. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/houston/per100shifts_stops_main Descriptive statistics on the number of stops made by each officer group on each civilian group. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/lear_geocoded.rds This file has the coordinates of agency headquarters based on their addresses in the LEAR dataset and is used for mapping. United States Department of Justice. Office of Justice Programs. Bureau of Justice Statistics. Law Enforcement Agency Roster (LEAR), 2016. Inter-university Consortium for Political and Social Research [distributor], 2017-04-05. https://doi.org/10.3886/ICPSR36697.v1
data/means_table_results_for_tie_fighters.rds Mean values of our full data (i.e. combining all 99 agencies) for the share of officers and share of civilians (“Hypothetical Representative Officer %”) for variables. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/officer_data_leoka_lemas_rows_deidentified.rds This is our officer data from the 99 agencies we collected data from, with one row for every officer. Officers with an L2 match of at least 0.90 have non-NA values for L2 variables. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/officer_data_leoka_lemas_rows_deidentified_age_true.rds & data/officer_data_leoka_lemas_rows_deidentified_dob_true.rds & data/officer_data_leoka_lemas_rows_deidentified_probability_0.95.rds & data/officer_data_leoka_lemas_rows_deidentified_race_adjusted.rds These are different versions of the data/officer_data_leoka_lemas_rows_deidentified.rds file and are used in robustness checks. In data/officer_data_leoka_lemas_rows_deidentified_age_true.rds we include only agencies that have the officer’s age in the roster and include age while matching. In data/officer_data_leoka_lemas_rows_deidentified_dob_true.rds we include only agencies that have the officer’s date of birth in the roster and include date of birth while matching. data/officer_data_leoka_lemas_rows_deidentified_probability_0.95.rds uses 0.95 as our matching criteria rather than 0.90. data/officer_data_leoka_lemas_rows_deidentified_race_adjusted.rds debiases race data that affects officers in agencies that do not report to LEMAS. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/PD.sample.LEAR.xls LEAR data showing data on all agencies in that dataset. We use it to identify the top 100 agencies (of which we collected data for 99), the state they are in, and the type of agency (e.g. municipal, sheriff). United States Department of Justice. Office of Justice Programs. Bureau of Justice Statistics. Law Enforcement Agency Roster (LEAR), 2016. Inter-university Consortium for Political and Social Research [distributor], 2017-04-05. https://doi.org/10.3886/ICPSR36697.v1
data/pew/ Each file contains SHAP values for machine learning models’ prediction of policing attitudes. There are 21 files for each of the policing attitude outcomes. Each cell is the feature's additive contribution to the model's prediction for that row. The following explains the categories for each feature, but their raw value is not what is presented in the files. Raw data comes from: “The American Trends Panel Survey, Wave 20.”. https://www.pewresearch.org/ politics/dataset/american-trends-panel-wave-20/.
data/sensitivity_means_table_results_for_tie_fighters.rds Mean values of our full data (i.e. combining all 99 agencies) for the share of officers and share of civilians (“Hypothetical Representative Officer %”) for variables. For officers it shows the lower and upper bound value generated by the “worst case” and “best case” measurements described in the paper. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/sensitivity_tests.rds Agency-level information about officers in the agency and civilians in that agency’s jurisdiction. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse.
data/shapefiles/houston_zones.shp These are the police beat shapefiles for the Houston Police Department. City of Houston GIS Data Hub. https://cohgis-mycity.opendata.arcgis.com/datasets/63096b9e650b48e2ac5d29b3f771f37d_5/explore?location=29.839729%2C-95.387800%2C10.76
data/shapefiles/geo_export_3cda220f-dc4d-49bd-afa2-14e36d0eb08f.shp These are the police district shapefiles for the Chicago Police Department. Chicago Data Portal. https://data.cityofchicago.org/Public-Safety/Boundaries-Police-Districts-current-/fthy-xz3r
data/social_explorer_tracts.rds Census data from Social Explorer for tract-level information about residents. U.S. Census Bureau. American Community Survey (2015-2019 5-Year). Prepared by Social Explorer.
Officer_tract_data_for_tract_table.rds Officer-level data without identifying officer information but with information about civilians in the officer’s home Census tract, and information about civilians in the jurisdiction of the agency the officer works for. Ba, Bocar; Ge, Haosen; Kaplan, Jacob; Knox, Dean; Komisarchik, Mayya; Lanzalotto, Gregory; Mariman, Rei; Mummolo, Jonathan; Rivera, Roman; Torres, Michelle, 2024, "Replication Data for: Political Diversity in U.S. Police Agencies", https://doi.org/10.7910/DVN/CZOPH3, Harvard Dataverse. |