Predicting Crime Through Incarceration: The Impact of Rates of Prison Cycling on Rates of Crime in Communities

• NCJ Number: 247318
• Author(s): Todd R. Clear; Natasha A. Frost; Michael Carr; Geert Dhondt; Anthony Braga; Garrett A.R. Warfield
• Date Published: May 2014
• Length: 141 pages
• Grant Number(s): 2009-IJ-CX-4037
• Agencies: NIJ-Sponsored
• Publication Type: Report (Study/Research); Report (Grant Sponsored)
• Award: Predicting Crime Through Incarceration: Assessing The Impact Of Rates Of Prison Cycling On Rates Of Crime In Communities

Annotation

This study estimated the impact of "prison cycling" (the flow into and out of prison) on crime rates in communities, with attention to neighborhoods that have high rates of prison cycling.

Abstract

The study found strong support for the impact of prison cycling on neighborhood crime rates, i.e., when resident removal rates due to incarceration were high, crime rates decreased; and when reentry rates from prison were high in a neighborhood, the crime rate increased. This involved controlling for neighborhood characteristics. Prison cycling had different effects in different neighborhoods, consistent with the concept of a "tipping point" (the level of removal and reentry in prison cycling); however, neighborhood differences are more clearly expressed as an interaction between crime control policy in a neighborhood and type of neighborhood. Further research will examine whether this neighborhood interaction holds in other sites. It will also assist in determining how neighborhood change over time affects the prison cycling-crime relationship. The issues to be answered in future research are whether neighborhoods that improve their conditions begin to benefit from incarceration policy or whether current incarceration policy is a factor that inhibits neighborhood improvement. The analyses focused on the impact of prison removal and returns on crime rates in communities (defined as census tracts) in the cities of Boston, MA and Newark and Trenton, NJ, as well as across rural municipalities in New Jersey. The data collection developed a uniquely comprehensive crime and incarceration dataset over time. The dataset enabled the modeling of the relationship between crime and incarceration by using a range of techniques (fixed effect panel models, Arrellano-Bond estimations, and vector auto-regression). 36 tables, 9 figures, and 133 references