Fatal Force: Analysis of Police Killings in the US

Project Overview

This project explores a dataset compiled by The Washington Post tracking every fatal shooting in the US by a police officer in the line of duty — from January 1, 2015 onwards. The analysis uncovers patterns across race, gender, age, mental health, geography, and socioeconomic factors.

Datasets Used

File	Description
Deaths_by_Police_US.csv	Core dataset: every fatal police shooting since 2015
Median_Household_Income_2015.csv	US Census median household income by city
Pct_People_Below_Poverty_Level.csv	Poverty rate by US state/city
Pct_Over_25_Completed_High_School.csv	High school graduation rates by state
Share_of_Race_By_City.csv	Racial demographics by city

Census data sourced from factfinder.census.gov

Key Questions & Answers

1. Poverty & Education

   • Q: Which US states have the highest poverty and lowest graduation rates?
   • A: Southern states (e.g., Mississippi, Louisiana, New Mexico) consistently rank highest in poverty. States like California and Texas have wide city-level variation. Low graduation rates correlate strongly with high poverty states.

2. Race

   • Q: What is the racial breakdown of people killed by police?
   • A: White individuals make up the largest absolute count, but Black individuals are killed at a rate 2–3× higher than their share of the population. Hispanic individuals are also disproportionately represented.

3. Gender

   • Q: How many more men are killed compared to women?
   • A: Men account for approximately 95%+ of all police killings. Women represent a very small minority of fatalities.

4. Age

   • Q: What percentage of victims were under 25?
   • A: A significant portion of victims were young adults. The mean age is roughly 36–37, with a notable share of killings involving people under 25, highlighting youth vulnerability.

5. Manner of Death

   • Q: Is there a difference between men and women in manner of death?
   • A: "Shot" is the dominant manner of death for both genders. "Shot and tasered" cases also appear. Average ages differ slightly by manner and gender, but the overall patterns are similar.

6. Armed Status

   • Q: What percentage of people killed were armed vs. unarmed?
   • A: Approximately 93–95% of victims were armed (most commonly with guns). Around 5–7% were entirely unarmed — still hundreds of people over the dataset's timeframe.

7. Mental Illness

   • Q: What percentage of victims showed signs of mental illness?
   • A: Roughly 25% of people killed by police showed signs of mental illness, underscoring the need for better crisis intervention training.

8. Geography

   • Q: Which cities and states see the most police killings?
   • A: Los Angeles, Phoenix, and Houston rank among the top cities. California, Texas, and Florida lead at the state level — partly due to population size, but rates remain high even when adjusted.

9. Race by City

   • Q: What is the rate of death by race in the top 10 cities?
   • A: In cities like Los Angeles, Hispanic victims form the largest share. In cities like Chicago or Houston, Black victims are disproportionately represented. The racial breakdown varies significantly by city demographics.

10. Trend Over Time

    • Q: Is the number of police killings increasing or decreasing?
    • A: The data shows a relatively stable or slightly increasing trend since 2015, with roughly 900–1,000 killings per year — no significant downward trend despite public attention and policy discussions.

Libraries & Tools

• pandas — data loading, cleaning, and aggregation
• matplotlib — bar charts, histograms, line graphs
• seaborn — grouped bar/box plots
• plotly — interactive choropleth map
• numpy — numerical operations

Setup

pip install pandas matplotlib seaborn plotly numpy

Then run the notebook Fatal_Force__start_.ipynb cell by cell in Jupyter or Google Colab.

Note: Run %pip install --upgrade plotly in Colab before starting.

Key Findings Summary

• Racial disparity: Black and Hispanic individuals are killed at disproportionately higher rates relative to their population share.
• Gender: Men account for the overwhelming majority of police killings.
• Age: A significant portion of victims were under 25 years old.
• Armed status: While most victims were armed (often with guns), a notable percentage were entirely unarmed.
• Mental health: A meaningful share of victims showed signs of mental illness, highlighting gaps in crisis intervention.
• Geography: Police killings cluster in specific cities and correlate with high-poverty, low-graduation-rate states.
• Trend: The data reveals whether killings have increased, decreased, or remained steady since 2015.

Epilogue

After completing your own analysis, compare your findings with The Washington Post's professional investigation: https://www.washingtonpost.com/graphics/investigations/police-shootings-database/

Data Source

The Washington Post — Fatal Force Database
Tracking details include: race, age, gender, armed status, mental health indicators, and location.
Supplemental data gathered from law enforcement websites, local news reports, social media, and independent databases ("Killed by Police", "Fatal Encounters").