Examining the changing dynamics of homicide in Los Angeles: 1990 - present

Homicidal violence - particularly lethal gun violence - has increased across the United States in the last few years after declining for three decades. Although this increase has been well documented, the situational dynamics of this violence have not. Specifically, there is a lack of information and analyses on fundamental issues related to processes, motives, locations, weapons, participants, and other factors that shape this new wave of homicides/gun violence. A clear understanding of the situational drivers of homicide is an essential first step to developing policy and practice to prevent it, but the task also requires detailed data that can only be collected from homicide files (incident-level data).

This project examines the fundamental questions of what situational dynamics are driving this recent increase, how they fit within larger spatial, temporal and socio-demographic trends, and whether there is room for these situational factors to drive violent crime even higher. The project focuses on the city of Los Angeles. To do so, it combines publicly available data with data from Los Angeles Police Department’s “Murder Books” for the entire city for the period of 2010 through 2022, as well as information from 4,000 murder books that covers 1990-2010 for the South Bureau area of Los Angeles stored in Documentum, a database created by the FBI and LAPD. South Bureau historically accounts for a disproportionate number of homicides that occur within the city limits. The value of using incident-level data is well established both within the scholarly treatment of attempts to understand violence, as well in the development of gun violence reduction strategies. For instance, reviewing homicide files was a vital source of information for the development of the “pulling levers” strategy of the Boston Gun Project, which led to multiple replication efforts throughout the country, including in the majority of Project Safe Neighborhoods sites.  

Using a variety of methods from simple statistical analysis to more advanced machine learning techniques, this project uses the rich array of variables culled from the homicide files to improve our understanding of the natural structure and situational dynamics of homicides (typologies), the factors associated with lethality in gun violence, and the natural structure of the processes involved in homicide investigations and its relationship with clearance rates. The project also focuses on gang-related violence and race/ethnicity as key factors intersecting all three goals.

Note: This project contains a research and/or development component, as defined in applicable law," and complies with Part 200 Uniform Requirements - 2 CFR  200.210(a)(14). CA/NCF