Sports events and violence in large cities in the United States

Matt Ashby and Alina Ristea
UCL Department of Security and Crime Science

Sports venues are important features of a city

Fans are heterogeneous

Ristea et al. (2018)

Struse and Montolio (2014)

Kurland & Johnson 2019

Klick & MacDonald 2021

Fabel & Rainer 2021

Menaker et al. 2021

Hypothesis

• Daily assault frequency around venues will be different on days with sports events

Crime data for 10 major US cities, 2010–2019

Event data for six sports, 2010–2019

	Austin	Chicago	Detroit	Kansas City	Los Angeles	New York	San Francisco	Seattle	St Louis	Tucson
College football
Major League Baseball
Major League Soccer
National Basketball Association
National Football League
National Hockey League

Model

Estimate relationship between daily count of assaults and:

• lagged crime count (fixed effect)
• minutes of daylight (centred, random effect)
• sporting event (random)
• day of the week (fixed)
• first day of month (fixed)
• New Year’s Day/4 July/Halloween/Thanksgiving (fixed)

Multi-level Bayesian negative-binomial model

Fixed effects: normal prior

Random effects: Student \(t\) prior

• American football and baseball games associated with higher violence
• Ice hockey sometimes associated with higher violence
• Soccer and basketball not associated with higher violence

Different effects for …

• different sports
• different cities
• downtown vs out-of-town stadia
• weekdays/weekends

These slides:
bit.ly/esc-stadia-2024

Questions:
matthew.ashby@ucl.ac.uk or a.ristea@ucl.ac.uk

Pictures: Ben Vardi, Bradley Johnson, Muhammad Ravel