The World Health Organization's declaration of COVID-19 as a global health emergency in March 2020 prompted widespread implementation of lockdown measures, including stay-at-home orders. These restrictions, while crucial for public health, had extensive societal, economic, and human rights implications. Data indicate significant reductions in daily movement during peak lockdown periods. This study investigates the association between stay-at-home restrictions and changes in six crime categories (assault, burglary, robbery, theft, vehicle theft, and homicide) across 27 diverse cities. The cities selected represent a wide spectrum of restriction stringency, ranging from voluntary recommendations to strictly enforced orders. This broad geographic and policy scope allows for a more generalized evaluation of the effects of stay-at-home orders on crime compared to previous single-city studies. Existing criminological theories offer contrasting predictions: strain theories suggest increased stress and negative emotions might lead to higher crime rates, while opportunity and routine activity theories predict crime decreases due to reduced opportunities and increased guardianship. Early studies on the COVID-19 pandemic's impact on crime revealed mixed results, suggesting variations across countries and crime types. This study aims to address these inconsistencies by examining a larger, more diverse dataset.

Several theories attempt to explain the relationship between population movement restrictions and crime rates. Strain theories posit that restrictions increase stress and negative emotions (anxiety, frustration, anger), potentially increasing criminal motivation. This could manifest in increased domestic violence, child maltreatment, and substance abuse due to social isolation and reduced freedom. Conversely, opportunity and routine activity theories suggest that stay-at-home policies disrupt the convergence of motivated offenders, suitable targets, and the absence of capable guardianship, leading to lower crime, particularly in public spaces. Prior research on the impact of COVID-19 on crime showed inconsistent results, with some studies reporting crime decreases and others noting variations across countries and crime categories. These variations highlight the need for a more comprehensive analysis considering diverse contexts and policy stringency.

This study employed interrupted time series (ITS) analysis, a quasi-experimental design suitable for evaluating the impact of interventions on time series data. The researchers collected daily police-recorded crime data for six major crime categories from 27 cities across the Americas, Europe, the Middle East, and Asia. The data spanned from January 2018 or 2019 to the most recent available date. The treatment variable was a dummy variable indicating the implementation of stay-at-home restrictions in each city. Poisson generalized linear models with a logit-link function were used to analyze the count data, accounting for seasonality, autocorrelation, potential outliers, and average daily temperature as a covariate. Meta-analytic techniques were used to summarize effect sizes across cities and crime types. To explore the heterogeneity in effect sizes, meta-regression analyses examined the relationship between the stringency of stay-at-home restrictions (measured on a 0-3 scale) and the size of the crime reduction. Additional analyses considered other COVID-19-related policy responses (school closures, workplace closures, etc.) and Google's COVID-19 Community Mobility Reports to assess their impact on crime trends. The Oxford COVID-19 Government Response Tracker data was used to measure policy stringency.

The analysis revealed a substantial overall decrease in urban crime (-37%) following the implementation of stay-at-home restrictions. However, the magnitude of the effect varied significantly across crime types and cities. The largest decreases were observed for robbery and theft (-46% and -47%, respectively), followed by vehicle theft (-39%), assault (-35%), and burglary (-28%). Homicide showed a smaller decrease (-14%). Meta-regression analysis indicated a significant positive association between the stringency of stay-at-home restrictions and the magnitude of crime reduction for burglary, robbery, theft, and vehicle theft. More stringent restrictions were associated with larger crime declines. For assault, while the association was not statistically significant at p<0.05, a visual inspection of scatterplots hinted at the potential influence of outliers. While other COVID-19-related policies were investigated, their independent effects on crime reduction were less consistent than those of stay-at-home orders. The inclusion of an overall stringency index did not substantially improve model fit, suggesting that stay-at-home restrictions were the primary driver of crime reduction. Analyses using Google's mobility data largely supported the findings, demonstrating that cities with larger declines in public space usage experienced greater crime decreases (except for homicide).

The findings strongly support the predictions of opportunity and routine activity theories. The study demonstrates that crime levels are highly responsive to changes in opportunity structures and constraints, with reduced mobility being a primary factor in crime reduction. The short-lived nature of the crime declines, with a return to pre-restriction levels within weeks, suggests that the initial effect was primarily driven by changes in routine activities rather than sustained changes in offender motivation. The varying effects across crime categories likely reflect differences in opportunity structures. Crimes requiring the convergence of offenders and victims in public spaces (e.g., theft, robbery) experienced the largest declines. Homicide showed a smaller reduction potentially due to its frequent occurrence in domestic contexts or its association with organized crime, which is less susceptible to changes in daily routines. The study's results highlight the importance of considering the interplay between opportunity, guardianship, and offender motivation in understanding the dynamics of crime. The study also emphasizes the importance of stay-at-home restrictions as the primary driver of crime reduction, although more research is needed to dissect the contributions of other concurrent policies.

This global analysis provides robust evidence that stay-at-home restrictions during the COVID-19 pandemic were associated with substantial reductions in urban crime. The findings underscore the critical role of opportunity structures in shaping crime levels, highlighting the immediate impact of mobility restrictions on crime rates. Future research should investigate the longer-term effects of these policies, explore potential crime displacement to online or domestic settings, and conduct more granular analyses within cities to better understand contextual factors influencing crime trends. Analyzing the relative impacts of different policy components and their enforcement is also crucial.

The study's limitations include a non-random sample of cities, primarily from Europe and the Americas, and reliance on police-recorded crime data, which is subject to underreporting and variations in definitions and recording practices across jurisdictions. Data limitations prevented the investigation of crime displacement to online platforms or other crime types. The study's reliance on macro-level analysis prevents a detailed examination of neighborhood-level heterogeneity in crime reduction. These limitations warrant caution in generalizing the findings to all urban contexts.