New York State Gun Violence Reduction Project

How to cite

LaForest-Tucker, Michael; Sharkey, Patrick. 2024. "New York State Gun Violence Reduction Project." AEA RCT Registry. AEARCTR-0012863. https://doi.org/10.1257/rct.12863-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Michael LaForest-Tucker* , Patrick Sharkey

* Primary investigator

Registry ID
AEARCTR-0012863
DOI
https://doi.org/10.1257/rct.12863-1.0
First registered
Last updated
Status
Completed
Trial dates
2012-09-01 – 2022-01-01
Countries
United States of America ()
Registry record
http://www.socialscienceregistry.org/trials/12863

Abstract

Can gun violence be deterred at low cost? We report the results of a randomized experiment of a messaging intervention which was designed to reduce gun violence among individuals under parole supervision with a prior violent felony conviction or firearm arrest. The intervention consisted of a group meeting in which high-risk paroled individuals were notified of the sanction they would face upon reoffending while being offered community resources to support re-integration into the community. The program did not lead to a reduction in gun violence or create community spillover effects but did reduce parole violations by 15%.

Experimental design

The New York State GVRP was implemented in five sites across the state of New York. Two sites were in Albany and Schenectady, mid-sized cities several hours north of New York City. The remaining three sites were in New York City, one each in the Bronx, Manhattan, and Brooklyn. Individuals living in these communities were eligible for the forums if they (1) were on parole, (2) had been convicted of a violent felony or had a prior firearm arrest, (3) did not have a mental health diagnosis or a prior sexual assault charge, (4) had lived within the targeted area for at least 30 consecutive days, and (5) had been released from prison within the prior two years. Forums took place between August 2012 and July 2014 in each of the five sites, with randomized assignment of attendees taking place between February 2013 and July 2014.

Treatment assignment involved two stages of randomization in each of the three NYC study sites. In a first stage of randomization, census tracts within each area were randomly assigned to a “neighborhood treatment group” and a “neighborhood control group.” Individuals who lived within the treated communities were eligible to be called in to the forums. Individuals in the control communities were not eligible to be called in. In the Albany and Schenectady sites, no neighborhood level randomization occurred due to the substantially smaller eligible populations in these cities. In these two sites, all eligible individuals were classified as belonging to the neighborhood treatment group.

In all five study sites, eligible individuals were then randomly assigned to an “individual treatment group” and an “individual control group.” Thus, in the NYC study sites, eligible individuals belonged to one of four study groups based on their combination of neighborhood treatment group and individual treatment group. Only individuals in both the neighborhood and individual treatment group (Group 1) were eligible to be called in to the forums. In the Albany and Schenectady sites, eligible individuals belonged to one of two study groups – an individual treatment group and an individual control group. In these sites, all individuals in the individual treatment group were eligible to be called in.

Design details

Randomization method
In office by a computer
Randomization unit
Treatment assignment involved two stages of randomization in each of the three NYC study sites. In a first stage of randomization, census tracts within each area were randomly assigned to a “neighborhood treatment group” and a “neighborhood control group.” Individuals who lived within the treated communities were eligible to be called in to the forums. Individuals in the control communities were not eligible to be called in. In the Albany and Schenectady sites, no neighborhood level randomization occurred due to the substantially smaller eligible populations in these cities. In these two sites, all eligible individuals were classified as belonging to the neighborhood treatment group.
Clusters
171
Observations
3557
Arms
Neighborhood treatment + individual treatment: 1101 Neighborhood treatment + individual control: 1054 Neighborhood control + individual treatment: 710 Neighborhood control + individual control: 692
Primary outcomes
Arrests and parole violations

Cited by

No citation data available for this registration yet.

Keywords