Community Safety Investments

Violence Prevention Initiatives Across New York State

by
Rebecca Balletto, Jeffrey A. Butts, Sheyla A. Delgado, Justice Evans, Gina Moreno, Michael Rempel, René Ropac, Elia Del Carmen Solano-Patricio, and Kathleen A. Tomberg

Research and Evaluation Center
&
Data Collaborative for Justice
February 2025

Preface and Executive Summary

Officials from New York State government asked researchers at John Jay College of Criminal Justice to assess the community safety benefits derived from more than ten years of investment in violence prevention initiatives managed at the county level. As of 2014, nearly two dozen counties in New York were operating at least one of the different violence prevention models funded by the State. The scale of the investments expanded significantly after 2021.

John Jay College research teams from the Research and Evaluation Center (JohnJayREC) and Data Collaborative for Justice (DCJ) examined changes in the incidence and rate of serious and violent offenses at the county level between 2010 and 2023 in an attempt to detect the effects of the State’s effort to build community safety at the local level.

Limitations

The report is not an evaluation. The analyses contained in this document are a quick scan of crime indicators and an assessment of possible associations between State funding and crime. The funding provided to the research team did not allow for a detailed evaluation. If, in the future, State officials seek an actual evaluation of these initiatives, a research team would need to include a number of important design elements. For example:

  • To establish credible claims of cause and effect, an evaluation would need some way of monitoring the efforts of each intervention beyond the simple tabulation of their budgets. The interventions described in this analysis pursue very different strategies, and evaluations of their work would need to measure their efforts in detail. This applies especially to the Crime Analysis Centers, which are mentioned in this report but not included in the data analytics. Effects of the CAC network are inherently less direct than the violence prevention initiatives, and researchers would need to collaborate with law enforcement to measure their operations with more detail than simple crime counts (e.g., investigations, arrests, clearances, convictions, etc.).

    •  

  • A fully designed evaluation of interventions to reduce violent crimes, especially firearm violence, needs to measure outcomes in very large areas and/or over extended periods of time to account for the relatively low incidence rate of reported violence. Especially in small communities, the number of violent crimes is thankfully low enough that statistical analyses are often unable to capture enough variation in key outcomes to calculate reliable outcome measures without aggregating data from several areas. This report includes other types of crime (i.e., serious property crime) as a way of compensating for low violent crime counts, and as a broader assessment of possible effects of the State’s prevention initiatives.

    •  

  • The varying sizes and characteristics of counties are particular challenges in New York. The State includes some of the most densely populated jurisdictions in the country as well as many small, rural communities. Using the common term “county” to describe all of them as a group does not eliminate these vast differences. Unless an evaluation was conducted over a very long period of time while monitoring and controlling inter-area variations, an analysis would have to create groups of similar areas with identical implementation measures and then track any changes in their crime indicators.

    •  

  • An evaluation of policy interventions designed to improve safety across varying geographic areas must include an intentional comparison strategy and measure relevant covariates. Outcomes in “treated” areas must be compared with those in “untreated” areas that are as similar as possible, controlling for demographics, socio-economic variables, recent history of crime indicators, and the presence and intensity of other related interventions. This analysis did not have the resources to collect primary data to build such a comparative approach. Rather, it simply divides the State by county and measures the extent of intervention funding on an annual basis.

    •  

  • Even if a research team had access to a wider array of data about all New York counties, an evaluation would still be hindered by the imperfect comparison areas represented by New York counties that did not receive funding for the initiatives examined here. A traditional method of countering this deficit is known as synthetic controls. Researchers create hypothetical matched comparison areas by compiling data from communities in other states that are similar politically, demographically, and economically to the intervention areas in the state being evaluated. The research team has experience with this method, but using it for this assessment was not possible given the available time and financial resources.

Key Findings

Despite various shortcomings, the research team found important indicators that suggest positive benefits of the State initiatives to prevent crime and violence. When researchers analyzed violent and property index crimes (i.e., aggravated assault, robbery, burglary, and larceny), the change in crime occurrences sometimes varied by the amount of funding received. Using 2010 as the base year and tracking crime rates through 2023, researchers found that total index crimes dropped 14 percent in counties receiving funding, but index crimes grew 13 percent in counties that received no funding for the three main initiatives. Researchers then constructed simple multivariate models to explore these statistical associations. The analyses did not include the type of covariates suggested above, and the results were not consistently positive, but they pointed to some possible benefits of the State’s initiatives — i.e., modest mitigation of firearm violence and serious assaults.

While this analysis was not a rigorous evaluation designed to establish precise estimates of cause and effect, the assessment did reveal encouraging indicators to suggest that the State’s investments may have lowered the occurrence of violence and increased the safety and security of neighborhoods across New York. Given the scale and scope of previous investments, policymakers would be well-advised to investigate further with an intentional and more robust approach to evaluation.

Introduction

The John Jay College of Criminal Justice Research and Evaluation Center (JohnJayREC) partnered with John Jay’s Data Collaborative for Justice (DCJ) to help New York State government officials estimate the benefits of more than ten years of investment in violence prevention initiatives across the state.

Estimating the public safety benefits of enhanced crime-prevention investments is complicated due to many pre-existing programs and policies across all New York communities. There is no such thing as an “untreated” comparison area to inform the analysis. Thus, it was not possible to construct a rigorous, quasi-experimental evaluation design. Instead, researchers tracked the incidence of crime and violence across the state. The study examines the extent to which county-level crime measures might be correlated with State government spending on new initiatives at the county level.

The project focused on four major initiatives:

      1. SNUG (guns spelled backward), a community-level intervention program inspired by Cure Violence,
      2. GIVE (Gun Involved Violence Elimination), inspired by law enforcement and place-based models,
      3. Project RISE (Respond, Invest, Sustain and Empower), and
      4. An expansion of the State’s Crime Analysis Centers (CAC).

Some jurisdictions in the state received funding for the GIVE initiative in all eleven years between 2014 and 2024, while others received various combinations of GIVE, SNUG, and Project RISE for varying periods (Table 1).

SNUG uses a public health approach to reduce gun violence. First implemented in 2009 but then reduced temporarily due to state budget shortfalls, SNUG deploys outreach workers, social workers, case managers, and hospital responders to engage with individuals in high-risk communities, mediate the underlying causes of violence, and partner with community-based nonprofit groups to enhance safety. SNUG focuses on interrupting cycles of violence by promoting conflict resolution and offering support services that foster safer neighborhoods and help to reduce gun-related incidents. By 2024, after the New York State budget had recovered from previous shortages, 14 New York counties were once again building public safety efforts with SNUG funding.

New York’s GIVE initiative, launched in 2014, uses multiple strategies and frameworks inspired by law enforcement and place-based models to reduce firearm-related crimes. The program tries to build trust between law enforcement and communities, including enhanced crime analysis approaches, problem-oriented policing tactics, focused deterrence, hot spots policing, street outreach, and crime prevention through environmental design (or CPTED). As of 2024, police departments in more than two dozen counties had received GIVE funding.

Project RISE, launched in 2022, was designed to build partnerships between established, well-resourced lead organizations and smaller grassroots groups to address issues related to gun violence and public safety. It provides funding and support for gun violence prevention efforts to implement and expand programs serving youth and families, focusing resources on at-risk individuals, and fostering community-driven solutions. The model was intended to empower communities and grassroots organizations that could create sustainable solutions by building local capacity to address the root causes of violence. Ten cities across New York State had received Project RISE funding by 2024.

In addition to these three initiatives focused on violence, especially gun violence, officials expanded funding for New York’s network of Crime Analysis Centers (CACs). These multi-agency units use data analysis to identify crime trends and patterns more comprehensively. Each CAC across the state operates as an information-sharing hub, providing tactical support, analytical resources, and assistance with investigations upon request. The components of the CACs include dedicated analysts, data systems to monitor trends and crime patterns, and partnerships with federal entities (e.g., ATF, FBI), the National Guard, and regional law enforcement agencies. Their multi-agency efforts allow CACs to gather and analyze broad ranges of data to support effective crime-solving. The 11 existing CACs currently assist over 350 local, state, and federal law enforcement agencies across numerous counties in New York state.

Descriptive Analysis

The network of CACs helps state officials track patterns of crime while the three prevention initiatives (GIVE, SNUG, and Project RISE) focus on changing those patterns. Researchers were asked to analyze whether those three efforts and the significant expansions of funding in 2014 and 2021 could be associated with changes in crime (Figure 1).

The project focused on two key questions:

      • Was the timing of investment associated with changes in crime and violence at the county level?
      • Was the financial scale of investment associated with changes in crime and violence at the county level?

Researchers investigated the questions with statistical analyses including regression models to test whether funding and implementation of the crime prevention programs appeared to be associated with changes in crime trajectories measured with various reported crime incidents between 2010 and 2023. (Some analyses explored the use of 2024 crime counts estimated with data from the first six months of that year.)

Data Sources

The research team considered analyses of violent crimes specifically involving the use of firearms across the state, but data for measuring gun crimes was unavailable at the local level before 2015. Analyzing trends after 2015 was less than ideal since so much of the funding was received by counties before 2015. Furthermore, some gun-specific crime measures in New York are organized at the jurisdiction level (i.e., city and township). Thus, jurisdiction population counts would be required to create county crime rates, and these may vary slightly from the county-level crime data used in the primary analyses conducted for this study.

Researchers focused on the timing and amount of funding received by New York counties for each of the three gun violence prevention initiatives. The study used a form of dose analysis to examine whether the scale (dollars per capita) and breadth (number of initiatives) of investment were associated with changes in crime across counties. Regression analyses tested relationships among these variables and several outcome variables built from the crime incident data provided by New York State between 2010 and 2023.

Researchers reviewed available datasets that could be used to assess changes in crime and violence and detect their association with the three gun violence prevention initiatives. Counties served as the primary geographic units of analysis because funding amounts were reported at the county level.

The primary data source was New York State’s compilation of crime incident reports organized with offense categories used by the Federal Bureau of Investigation (FBI) to standardize reporting across states. The FBI’s “index” offenses comprise two categories – violent crime and property crime. The former includes murder, rape, robbery, and aggravated assault, while the latter consists of burglary, larceny, and motor vehicle theft.

The analyses focus primarily on the violent and property crime indices as well as the individual offenses most likely to occur in sufficient numbers at the county level for statistical purposes and most related to the goals of the three funding initiatives: 1) total index crimes, 2) total violent index crimes, 3) total property index crimes, 4) robbery, 5) aggravated assault, 6) burglary, and 7) larceny.

The results suggest the prevention initiatives may have had their intended effect (Figure 2). Between 2010 and 2023, total index offenses declined 14 percent in counties receiving State funding while counties receiving no funding for any of the three initiatives experienced an increase of 13 percent in total index crimes between 2010 and 2023.

Researchers examined total violent index offenses in more detail and found that funded counties experienced a very small decrease in violent index crimes (–1%) while unfunded counties saw violent index crimes increase (+8%). The two largest categories of violent index crimes (robbery and aggravated assault) displayed different patterns, but both were consistent with a possible effect of state funding. All counties saw large declines in robberies, but counties receiving funding for the three initiatives experienced a decline of 31 percent in robberies while unfunded counties saw robberies fall 22 percent. Aggravated assaults, on the other hand, increased in all counties, but funded counties experienced less of an increase compared with unfunded counties (up 12% versus 23%).

Trends in the two largest categories of property index crimes were also starkly different, but the changes were again consistent with the goals of state funding. Burglaries declined considerably in funded counties (–59%) and less so in unfunded counties (–41%). Larceny offenses dropped 15 percent in funded counties but increased 23 percent in counties that did not receive funding for any of the three violence prevention initiatives.

Crime changes across all New York counties after 2020 reflect the social and economic interruptions experienced worldwide during the COVID-19 pandemic. From 2010 to 2020, total index crimes in New York declined 30 percent in counties receiving funding for the three prevention initiatives, while counties receiving no funding declined as well, but only by 16 percent (Table 2). Between 2020 and 2023, all counties experienced increased index crimes, but the increase in funded counties (+23%) was smaller than the increases experienced by unfunded counties (+35%). Similarly, total property index crimes grew 26 percent in funded counties, but the increase was larger in areas that did not receive funding (+41%).

Comparative changes varied between the larger offenses included in the two crime indices. Between 2020 and 2023, robberies grew 19 percent in funded jurisdictions and 21 percent in unfunded areas. In counties that received funding, the number of aggravated assaults grew 9 percent between 2020 and 2023, and the increase was slightly larger in counties not receiving funding (11%).

Larceny offenses increased in both funded and unfunded areas from 2020 to 2023, but the increase was larger in counties not receiving funding for violence prevention initiatives (46% versus 21%). Burglaries, on the other hand, dropped more in unfunded counties than in counties receiving funding (–8% versus –4%). One other property index category showed a very different pattern. Motor vehicle thefts increased sharply between 2020 and 2023, and the increase was much larger in counties receiving funding (116% versus 66%).

Trend Analyses

The possible effects of the State’s investments were apparent when researchers examined the number of crimes and percentage change in crimes between 2010 and 2023 (Figure 3). Counties generally reported falling numbers of index crimes from 2010 to 2019 before the sharp increases associated with the onset of the COVID-19 pandemic.

When viewed in terms of percentage change, it appears that counties that received funding for violence prevention initiatives sustained the declining number of index crimes slightly longer compared with unfunded counties. By 2023, the number of crimes in funded counties was still lower than the level of 2010, while index crimes in unfunded counties remained more than 10 percent above the level of 2010.

All jurisdictions experienced crime increases during the years of the COVID-19 pandemic, but perhaps the extra State funding helped some areas recover more effectively than others.

When index crimes are divided into violent offenses and property offenses, the patterns are similar while overall numbers vary (Figure 4). Violent index crimes generally declined in all counties between 2010 and 2019 before the onset of the COVID-19 pandemic. Between 2020 and 2023, violent index crimes in counties receiving funding grew 11 percent while violent index crimes in counties not receiving funding grew 12 percent.

The number of violent index crimes in funded counties in 2023 was just below the level of 2010 (down 1%). The number of crimes in counties not receiving funding was up eight percent in 2023 compared with 2010.

A stronger effect is apparent in the case of property index offenses. Both groups of counties experienced falling numbers of property crimes through 2019, but funded counties continued to decline through 2021 while unfunded counties saw property crimes grow in 2020 and 2021. All areas surged in 2022 and 2023, but the number of property index offenses in funded counties was lower in 2023 than in 2010 (–16%) while unfunded counties saw 15 percent more property index crimes in 2023.

To reduce the effect of varying population sizes, researchers calculated per capita crime rates, or the total number of crimes reported in each county group for every 100,000 residents (Figure 5).

The analysis revealed the similarity of index crime trends across all areas of the state. All counties experienced generally falling crime after 2012 before crime began to increase at some point between 2018 and 2020. Violent crime rates in funded counties were consistently lower than in non-funded counties throughout the period of the study: 2010 to 2023.

In contrast, the rate of property index crimes was greater in counties that began receiving funding after 2010, but the rate fell 35 percent between 2010 and 2020, from 2,105 to 1,363 crimes per 100,000 residents. Non-funded counties also saw property index crimes fall, but only by 23 percent. The different extent of change in per capita crime rates is clear when researchers plot the percentage change in index crime rates relative to the year 2010 (Figure 6).

Violent index crimes per capita were 10 to 15 percent lower by 2018 or 2019 than in 2010. Both county groups saw declines. Between 2020 and 2022, all counties combined experienced sharp increases in the violent crime rate. In funded counties, however, the surge returned violent crimes per capita to their 2010 level. Unfunded counties, on the other hand, saw violent crime rates increase to a level nearly 15 percent higher than in 2010. Violent crime rates then fell in both groups between 2022 and 2023.

Changes in property crimes per capita were even more dramatic and pointed to a potential effect of State funding. Index property crimes fell more in funded than unfunded counties, dropping to nearly 40 percent below the 2010 level by 2021. Even with sharp increases after 2021, the property index rate in 2023 remained almost 20 percent below the 2010 level, while the 2023 rate in unfunded counties was nearly 10 percent greater than in 2010.

Multivariate Analysis

The research team next considered the combined effects of expanded investment on reported crimes across New York. Researchers generated statistical models to test whether funding levels for the different initiatives were associated with various crime indicators.

The analysis addressed the study’s two key questions in more detail:

      • Overall association: Was the timing of new investment associated with reductions in crime rates?
      • Larger vs. smaller investments: Was the scale of new investment associated with reductions in crime rates? Specifically, was there an effect of multiple investments across all the initiatives?

Researchers examined the timing and scale of investments and their association with various crime categories in 22 counties. As shown in Table 1, 21 counties received funding for GIVE, and 13 of those (in addition to the Bronx) received funding for SNUG at some point. Of 13 counties funded to operate SNUG, six also received funding for Project RISE.

When the effect of State funding was examined across individual counties, the relationship between program support and reported index crimes was less apparent (Figure 7). Counties with larger populations were more likely to experience rising crime between 2010 and 2023, and except for the Bronx, most of the funding for violence prevention focused on midsized New York counties. Most of those jurisdictions did see index crimes fall between 2010 and 2023. Outside of New York City (i.e., Bronx, New York, Kings, Queens, and Richmond), only Oneida County reported more index crimes in 2023 relative to 2010.

In the remaining midsized counties (i.e., between the 30th and 57th largest jurisdictions), the relationship between State funding and changes in crime was less than clear. Even when the top six funded counties are considered separately, it is not possible to identify an association between funding and crime changes.

The research team conducted a regression analysis in an attempt to identify the possible effects of funding on crime incidents reported by counties. Researchers first conducted interrupted time series (ITS) analyses for each county and each crime category. The ITS results would indicate whether counties experienced changes in crime trends in the post-funding period relative to the pre-funding period (with positive coefficients indicating an increase in crime trends and negative coefficients indicating a decrease in crime trends). Next, the effect estimates were used as outcome variables in linear regression models to examine whether funding increases were associated with changes in crime trends.

Researchers generated bivariate regression models including all 22 funded counties and 11 of the most comparable unfunded counties to examine the association between the funding increases and changes in crime rates. Comparable counties included those with resident populations most similar to funded counties (i.e., Cattaraugus, Sullivan, Clinton, Wayne, Steuben, Putnam, St. Lawrence, Ontario, Oswego, Saratoga, and Richmond).

The analysis examined the association between funding of the three violence prevention initiatives with reductions in crime. In most crime categories, the association was not statistically significant, but the analysis showed possible beneficial effects with two offense types: violent crime involving firearms and aggravated assault (Table 3).

For example, the statistically significant coefficient for violent crimes with guns shows that, on average, funded counties saw a decrease in crime trends that was 0.052 crimes per 100,000 people per month larger (or their increase was less severe) compared to unfunded counties.

A second set of regressions examined whether the scale and breadth of investments were associated with changes in crime trends among the 22 funded counties. The independent variables in the regression models included: (1) if a county received funding for SNUG; (2) if a county received funding for Project RISE; (3) the year-over-year (YOY) funding increase for GIVE; (4) the YOY funding increase for SNUG; (5) the YOY funding increase for Project RISE; and (6) the YOY increase for all three initiatives combined.

The working hypothesis was that greater investments would be associated with crime reductions — or, at minimum, less severe increases in crime trends in funded counties versus unfunded counties. The regression results indicated no overall association between the scale and breadth of new investments and changes in crime trends at the local level, but the analysis may underscore the fact that other factors play a substantial role in shaping the general crime trajectories of communities.

Conclusion

Researchers at John Jay College of Criminal Justice analyzed public safety benefits from more than ten years of investment in local violence prevention initiatives in New York. The study was a retrospective statistical examination of associations between crime trends and the amount and timing of investments. It was not a rigorous evaluation to establish cause and effect and rule out extraneous factors, but when researchers analyzed crime incidences involving serious offenses against persons and property, changes in crime sometimes varied by levels of State funding. Researchers then employed multivariate regression models to identify statistical relationships between crime and funding more accurately. The analyses were not consistently positive, but two offense categories appeared to be related to the funding initiatives (violence committed with firearms and aggravated assault).

The study revealed encouraging indicators to suggest that investments in violence prevention may help the efforts of local jurisdictions to guard the safety of communities. With additional research involving specifically designed evaluation methods, policymakers could learn even more about how public financing of crime prevention efforts may support communities and produce beneficial, actionable results for New York residents.

Funding
The New York Governor’s Office provided funding support for this analysis through a contract between the State’s Division of Criminal Justice Services and the Research and Evaluation Center at the John Jay College of Criminal Justice, but all conclusions are those of the authors. Funders and partners of the Research and Evaluation Center and the Data Collaborative for Justice are not responsible for any findings presented in our publications.

Acknowledgments
The authors appreciate the guidance received from staff in the Office of the New York Deputy Secretary for Public Safety and the Division of Criminal Justice Services who helped to conceptualize the analysis presented in this report and who reviewed drafts of project deliverables as they became available. Their contributions helped to shape the final product.

Recommended Citation
Balletto, Rebecca, Jeffrey A. Butts, Sheyla A. Delgado, Justice Evans, Gina Moreno, Michael Rempel, René Ropac, Elia Del Carmen Solano-Patricio, and Kathleen A. Tomberg (2025). Community Safety Investments — Funding Violence Prevention Initiatives Across New York State. New York, NY: John Jay College of Criminal Justice, City University of New York.

John Jay College of Criminal Justice,
City University of New York
524 West 59th Street
New York, NY 10019