Skip to content

Latest commit

 

History

History
72 lines (54 loc) · 6.89 KB

README.md

File metadata and controls

72 lines (54 loc) · 6.89 KB
Raw

Visualizing Police Violence in Louisiana

Git for the Justice Lab: Visualizing Police Violence in Louisiana dashboard. This Git includes documentation, data, and scripts used to build this project, as well as all other work associated with the project.

Mission

We built this dashboard as an interactive tool to empower interested citizens, nonprofit organizations, journalists, attorneys, academics, and even law enforcement agencies to explore and compare statewide policing statistics, uncover crucial insights and identify areas needing greater transparency.

Goals

1. Enhancing Data Accessibility and Stewardship

Our mission is to offer seamless access to crucial policing statistics in Louisiana. Beyond presenting these statistics, we hope to empower users by providing access to the raw data, transforming them into custodians of this valuable information.

2. Amplifying Voices:

This initiative serves as a powerful tool to magnify the experiences of individuals who have faced police violence and their families. Through a combination of data and narratives, we hope to bring attention to and amplify the stories of those affected by violence.

3. Ensuring Accountability:

We are committed to holding the police accountable for their actions. This involves representing instances of violence against the community and ensuring that individuals have transparent access to information about violent departments and officers.

Git Structure

Sources

Data on known killings by police officers were obtained by Mapping Police Violence and include crowdsourced data on killings by police officers that have been reported by the media beginning in 2013. The Mapping Police Violence data are sourced from the three largest crowdsourced databases on police killings in the U.S. (FatalEncounters.org, the U.S. Police Shootings Database, and KilledbyPolice.net). These data have been supplemented by additional research on the victim’s race. More information on these data can be found on the Mapping Police Violence website. The data that we use was downloaded from Mapping Police Violence on January 11th, 2024.

Demographic information was obtained from the 2020 U.S. Census. In the analyses presented here, the Black population includes individuals who identified their race as Black or African American alone or in combination with another race. The white population includes individuals who identified their race as white and their ethnicity as not Latinx.

Data on the number of officer were obtained from the FBI Crime Data Explorer. The Law Enforcement Employees dataset comprises annual data concerning personnel employed by law enforcement agencies, encompassing both officers and civilians. The Law Enforcement Agency names were cleaned using the Law Enforcement Agency Identifiers Crosswalk, United States, 2012 accessed through the University of Michigan on January 11th, 2024.

Data on known misconduct by police officers were obtained by LLEAD through public records requests and include public data collected from a range of law enforcement agencies including police departments, sheriff’s offices, civil service commissions, and more. To learn more about how the data is collected, you can visit their website (llead.co) or their GitHub page. The data that we are using was downloaded from LLEAD on January 11th, 2024.

Data Cleaning

Police Killings

The process of cleaning police killings includes (see the code and analysis here):

  1. Cleaning the Mapping Police Violence data variable names.
  2. Filtering for police killings in Louisiana.
  3. Creating unknown options for the demographic variables.
  4. Extracting different date, year, and month combinations.
  5. Creating age categories.
  6. Cleaning Parish names and removing duplicate victim names.
  7. Retreiving 2020 Census demographic data for Parishes in Louisiana.

Overview

The process of cleaning the overview data includes (see the code and analysis here):

  1. Filtering the FBI Crime Data Explorer Employement data for agencies in Louisiana.
  2. Joining the FBI Crime Data Explorer Employement data with Law Enforcement Agency Identifiers Crosswalk data to yeild more complete names.
  3. Cleaning agency names more thoroughly

Misconduct

The process of cleaning the misconduct data includes (see the code and analysis here):

  1. Creating a full name variable in the LLEAD personnel data.
  2. Joining the LLEAD personnel data with the LLEAD History ID data via their unique identifiers.
  3. Renaming the repeated officer demographic variables.
  4. Joining the above dataframe with the LLEAD agency location data via the agency names.
  5. Categorizing the agencies.
  6. Cleaning the agency names.
  7. Using regex to categorize the dispositions and actions
  8. Using a multi-label classification model to categorize allegations (click here to see how we built this model).

Data

Visualization

The vizualizations for this project were primarily creating using Infogram. The only other visualization used were created through R Shiny and the code is provided in the above.

Product

The dashboard can be found here: https://www.aclujusticelab.org/dashboard/