Pandemic Restrictions and Spatiotemporal Crime Patterns in New York, São Paulo, and Stockholm

• Author: Tulio Kahn,Anders Östlund,Vania Ceccato,Christopher Herrmann
• Date: 01 February 2022
• DOI: 10.1177/10439862211038471
• Published date: 01 February 2022
• Subject Matter: Articles

https://doi.org/10.1177/10439862211038471

Journal of Contemporary Criminal Justice

2022, Vol. 38(1) 120 –149

© The Author(s) 2021

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/10439862211038471

journals.sagepub.com/home/ccj

Article

Pandemic Restrictions

and Spatiotemporal Crime

Patterns in New York, São

Paulo, and Stockholm

Vania Ceccato1, Tulio Kahn2, Christopher Herrmann3,

and Anders Östlund4

Abstract

Studies are showing evidence of the effect of changes in routine activities due to

the 2020 COVID-19 pandemic on crime levels in many cities worldwide. This study

evaluates the potential impact of the COVID-19 pandemic on temporal and spatial

patterns of crime in three major cities under very different national contexts.

Each of the three countries and cities experienced different levels of pandemic

restrictions and societal closure. The cities of New York (the United States),

São Paulo (Brazil), and Stockholm (Sweden) were selected as cases. Temporal

quantitative methods, spatial statistics techniques, and Geographical Information

System (GIS) underlie the methodology used in this study. Findings show that there

is a statistically significant break in the trend in crime levels after the stay-at-home

orders were implemented in New York City, São Paulo, and Stockholm in the first

months of 2020; the only exception was for murder. Such an impact varies by crime

type and city context, but increases again after a few months, indicating how fast

crime and criminals adapt. Residential burglary decreased, whereas nonresidential

burglary increased overall. Changes in the levels and geography of vehicle thefts

were observed, with an overall increase of significant cold spots but, in several

cases, also solidification of existing crime concentrations in known crime attractors

and in some deprived areas.

1KTH Royal Institute of Technology, Stockholm, Sweden

2Fundação Espaço Democrático, São Paulo, Brazil

3John Jay College of Criminal Justice, New York City, USA

4Swedish Police Authority, Stockholm, Sweden

Corresponding Author:

Vania Ceccato, Department of Urban Planning and Environment, KTH Royal Institute of Technology,

Teknikringen 10 A, 100 44 Stockholm, Sweden.

Email: vania.ceccato@abe.kth.se

1038471CCJXXX10.1177/10439862211038471Journal of Contemporary Criminal JusticeCeccato et al.

research-article2021

Ceccato et al. 121

Keywords

COVID-19, routine activity, spatial–temporal analysis, police statistics, hotspots,

ARIMA, differential local Moran statistic

Introduction

Urban life has changed considerably since the start of the COVID-19 pandemic in

early 2020. People now spend less time out of their homes and offices, many employ-

ees work from home, children are learning from home, and mask wearing and social

distancing have become the new social order. By April 2020, strict lockdown orders

had been implemented in numerous countries around the world, whereas a few coun-

tries, such as Sweden, became known worldwide for adopting “lighter” stay-at-home

recommendations. Although recent studies of North American and European cities

have shown a substantial impact on the volume and distribution of crime, in particular

violence (e.g., Gerell et al., 2020; Mohler et al., 2020; Rosenfeld & Lopez, 2020), the

international literature that compares different cities in different countries in terms of

crime levels and geography is still lacking. Findings of comparative studies of this

type can be particularly relevant for criminological research to illustrate the impact of

changes on local criminogenic conditions imposed by changes in routine activity. As

stated by Stickle and Felson (2020), these circumstances provide researchers with

naturally occurring, quasi-randomized control experiment conditions that allow more

advanced testing of criminological theories.

To contribute to this knowledge base, we report here variations between crime lev-

els and crime geography in 2019 and compare those variations with crime reported

during the 2020 stay-at-home recommendations in three major cities—New York, São

Paulo, and Stockholm—in different country contexts: the United States, Brazil, and

Sweden. Using multiple analytical strategies, we compare crime levels and geography

before and after the stay-at-home recommendations in these three cities. Temporal

quantitative methods, spatial statistics techniques, and Geographical Information

Systems (GIS) underlie the exploratory methodology used in this study. Note that the

research presented in this article is exploratory in nature because the analysis illus-

trates how differences in crime levels and geography are affected overall by the lock-

down and how key concepts from environmental criminology theories can help

interpret these changes. The study does not specifically test measures (for instance, of

guardianship, ambient population, number of people close to their residential nodes)

that would have to be tested to determine the applicability of the theories in interpret-

ing the impact of the lockdown restrictions.

The cities constitute interesting case studies for several reasons. First, they repre-

sent different stay-at-home restriction models imposed by local governments during

the pandemic. On one hand, New York City serves as an example of “strict lockdown”

measures put into operation from Week 12, with basic services and schools closed

until mid-June 2020. On the other hand, Stockholm represents a case of “mild lock-

down” during the same time period, with businesses and primary schools open (but

most secondary schools and universities closed), whereas São Paulo is the in-between