Protection Details
Police staffing shortages, public awareness, and policy preferences
Hunter Boehme et al.
Journal of Criminal Justice, March–April 2026
Abstract:
Since the summer of 2020, police agencies across the United States have continued to experience police staffing shortages. The consequences of understaffed police agencies are well studied, but little is known about whether the public is aware of the staffing problem and its consequences for public safety and police behavior. This study presents a pre-registered factorial survey experiment to 3105 US adults, varying the information respondents see on the consequences of staffing shortages (e.g., public safety, police behavior) and the messenger of this information (e.g., police chief, mayor). Respondents are then asked about their levels of concern about police staffing, the downstream consequences of staffing shortages on crime and public safety, and their willingness to pay more taxes to resolve the issue. We find a strong effect of the informational treatments on the public's concern about understaffed police departments and their willingness to pay more taxes, with surprisingly little effect heterogeneity by respondent race and partisanship. Despite well-documented polarized views toward contemporary policing among Americans, we find strong support across partisan and racial dimensions for resolving police staffing shortages. Our findings suggest that voters understand the link between short-staffed police departments and the consequences for public safety, and are even willing to pay more taxes in response. As public opinion is the primary heuristic elected policymakers rely on for forming policy agendas, these results indicate a real opportunity for policy action and proposals. Relevant stakeholders are likely to benefit from being clear about the consequences of understaffing on public safety.
The “Rainey Street Ripper”: Social Media Apophenia and Linkage Illusions
Kim Rossmo, Edward Anderson & Zena Rossouw
Homicide Studies, forthcoming
Abstract:
The theory that a serial murderer is drowning men from Austin’s Rainey Street District has been advanced by social and mainstream media for 3 years now, repeatedly fueled by speculation set off by the recovery of bodies from nearby Lady Bird Lake. We conducted an evidence-based study to detect any existence of a serial murderer. A search of police data produced 189 drowning incidents. From these, a target sample of 58 cases matching the general modus operandi and victimology of the alleged killer was identified for analysis. Our study found neither direct evidence nor indirect warning signs of a serial murderer. The frequency of drowning incidents is consistent with historical patterns, average drowning risk, and population growth. Several prior years experienced more drownings than 2023, when allegations of a serial killer began to escalate. We found a number of similar claims of mysterious drowning serial killers in other jurisdictions, suggesting an issue with social media apophenia and crime linkage illusions. Online forums exaggerate numbers of deaths by invoking longer time periods, larger geographic areas, and vaguer definitions of “suspicious.” Such allegations have costs in the real world, wasting valuable resources that could be used to help solve real crimes.
As Seen on TV: Local Television News and the Shaping of Rural Political Preferences
Lukas Alexander
American Politics Research, forthcoming
Abstract:
How do information environments shape political preferences among rural Americans? This study examines how variation in local television news exposure shapes rural Americans’ political preferences about government spending on law enforcement. These governmental spending preferences are often shaped by skepticism grounded in rural place-based identity. Drawing on agenda-setting theory, I argue that exposure to different media markets -- particularly those from more populous urban areas where coverage of crime and law enforcement is more prevalent -- creates distinct informational contexts that influence political preferences. Using survey data from rural respondents in the 2016, 2020, and 2022 Cooperative Election Studies and a matching design, the analysis finds that rural Americans exposed to local television news from urban media markets express greater support for increasing law enforcement spending, even at the expense of other public services. These findings challenge the idea that rural Americans hold rigid, identity-based political preferences rooted in place. Instead, this study demonstrates that rural political preferences are malleable and shaped by information environments.
An economic evaluation of a police–mental health co-response program: Data from a pragmatic randomized controlled trial
Meret Hofer et al.
Journal of Experimental Criminology, March 2026, Pages 31-45
Methods: Eligible 911 calls-for-service were randomized to receive a police-as-usual or a co-response. Next, we record-linked randomized events to emergency medical services, jail, outpatient services, and emergency department data to assess outcomes. We calculated per-person costs of service utilization following the randomized event from a public-sector perspective.
Results: Our analysis revealed no cost-savings from the co-response. Persons who received a co-response team response had greater 12-month post-randomized incident costs associated with outpatient behavioral health encounters and emergency department visits.
Identifying the general equilibrium effects of narcotics enforcement
Zachary Porreca
Journal of Urban Economics, March 2026
Abstract:
I analyze the demand side impacts of a supply-side intervention into the market for illegal drugs in what has been described as America’s largest open-air drug market. Beginning in 2018, the Pennsylvania Attorney General’s office and the Philadelphia Police Department engaged in an ambitious effort to shut down the drug market in Philadelphia’s Kensington neighborhood. The intervention involved increased police presence in the targeted area alongside a series of targeted “kingpin” sweeps which were intended to remove the most pervasive operators from the market. I employ highly granular SafeGraph cell phone location data to track changes in traffic flows between census block groups, observing that the Initiative led to sizeable and persistent reductions in traffic flows to the target area. Additionally, in contrast to substitution effects observed in other work, I observe that the Initiative led to reductions in traffic flows to other regional drug markets and large declines in overdose mortality in the Philadelphia metropolitan area as a whole, suggesting a genuine reduction in the quantity of illegal narcotics demanded. With a combination of theory and empirics, I argue that this reduction in the quantity demanded regionally is able to be achieved due to the Initiative disrupting a supply-chain that data indicates flows from the target area outwards to smaller satellite markets. Together this all suggests that, despite the inelastic demand for narcotics, regionally linked markets can be impacted broadly by location specific interventions.
Does Uber affect crime? Evidence from U․S․
Jian Zhang et al.
Journal of Economic Behavior & Organization, April 2026
Abstract:
This paper investigates the relationship between the ride-hailing platform Uber and local crime. Using a difference-in-differences approach with a sample of 28,258 city-year observations across the U.S. from 2009 to 2019, we find that Uber’s entry significantly reduces crime rates. Exploring the underlying mechanism, we show that Uber’s crime-reducing effect is significantly stronger in areas facing greater liquidity constraints -- specifically, regions with less bank credit supply, fewer local job opportunities, higher personal bankruptcy risk, and greater household financial stress. Consistent with this mechanism, further analysis reveals that this effect is particularly pronounced for crimes with explicit financial motives, such as robbery and larceny. Overall, our findings demonstrate that Uber may enhance urban safety by providing an economic safety net that mitigates the motivation for crime.
No man’s hand: Artificial intelligence does not improve police report writing speed
Ian Adams et al.
Journal of Experimental Criminology, March 2026, Pages 137-154
Methods: In a pre-registered randomized controlled trial, we test this claim within the patrol division of a medium-sized police department (n = 85) at the individual report level (n = 755). Analyses utilize mixed-effects regression accounting for the nested structure of report-writing.
Results: AI assistance did not significantly affect the duration of writing police reports. Alternative specifications beyond those specified in the pre-registration, including a difference-in-differences approach observing report duration over a full year (n = 6084), confirm the null findings are robust.
Factions and Brokers in the Russian Mafia: Investigating the Structure of the Thieves-in-Law
Niles Breuer, Federico Varese & Elena Racheva British
British Journal of Criminology, March 2026, Pages 306-330
Abstract:
This paper investigates the internal structure of the fraternity of the vory-v-zakone -- the high-ranking members of the Russian Mafia -- using a co-signing network based on 50 signed edicts (progony) produced by the fraternity. We find that the vory are a multi-ethnic, polycentric association grouped into factions based on vertical patronage relationships; these factions are interconnected by key brokers that bridge between groupings. While this structure allows for effective collective action, it also points to possible fractures that can be exploited by authorities to weaken the fraternity. This research is the first quantitative network analysis of documents produced by a higher-level coordination body within a mafia and contributes to the literature on the emergence of coalitions and factions within organizations.
Illuminating Crime Prevention: Evidence from Washington, DC's Street Lighting Upgrade
David Mitre-Becerril & Dylan Spencer
University of Connecticut Working Paper, January 2026
Abstract:
We examine how large-scale investments in urban infrastructure influence crime by evaluating Washington, DC’s 2023 citywide conversion of street lighting to LED technology. Using incident-level crime data from 2014 to 2025 and leveraging the staggered rollout of the upgrade, we estimate its effects using a staggered difference-in-differences design. Our primary analysis uses street segment–month data to capture immediate, micro-level impacts at the locations where lighting was upgraded, complemented by broader community-level effects. The results show significant reductions in nighttime offenses, driven largely by financially motivated crimes and some evidence of decreases in gun-related offenses. The effects persist for at least one year after the intervention. Crime did not reallocate to other times of day, and there is no evidence of displacement into adjacent streets.
How Do Inmates "Do Time"? Recidivism Impacts of Within-Prison Programs
Matias Axelrod
Michigan State University Working Paper, January 2026
Abstract:
I evaluate the impacts of several programs offered to inmates held at various prison facilities in the Arizona Department of Corrections, Rehabilitation, and Reentry (ADCRR) through a novel application of a quasi-experimental shift-share IV (SSIV) research design, which leverages plausibly exogenous variation across facilities over time in how exposed inmates were to three different programs. I find that a cognitive-behavioral therapy program, Changing Offender Behavior (COB), reduced one-year reincarceration rates by 3.4 percentage points (14%) and that these impacts did not dissipate by the third year. I also find suggestive evidence that this program reduced new violent felonies within two years. Substance-use disorder treatment reduced reincarceration rates by 3.7 percentage points (15%) within the first year but dissipated thereafter. Heterogeneity analyses reveal that most of the cognitive-behavioral therapy programs recidivism effects were due to fewer new felonies, while the substance-use disorder treatment's shorter-term effects were entirely due to fewer technical violations.