description

Recidivism and Prison Gerrymandering:

A Primer for Delaware and the Rest of the United States

What is Recidivism?

“Recidivism refers to a person’s relapse into criminal behavior, often after the person receives sanctions or undergoes intervention for a previous crime.” - NIJ


Among criminal justice researchers and state agencies, recidivism is often defined in terms of rearrest, re-conviction, or return to prison among formerly incarcerated persons within three years of their release. Because the risk of recidivism varies from the time of a person’s release, at-risk individuals are grouped into cohorts by year of release, and the number of those recidivating in a cohort is divided by the total to calculate the recidivism rate for the cohort.


Various nuances need to be considered when calculating any one of these measures of recidivism, and states often make different decisions that complicate cross-state comparisons. In the state of Delaware, the Department of Corrections organizes all correctional facilities within the state into five tiers, ranging from Level V (24-Hour Incarceration), Level IV (Quasi Incarceration, often work-release), and various degrees of probationary supervision from Levels I to III.

Map of Delaware Correctional Facilities

Security Level of Prison in DE

Level 4
Level 5
A map of correctional facilities in Delaware, composed by scraping facility addresses from pages on the DE DOC website.

In its annual reports on recidivism, the Delaware Statistical Analysis Center (SAC) only considers releases and readmissions to Level V facilities when calculating the state’s three measures** of recidivism:

  1. Rearrest: “The first incident of arrest during the at-risk period for any qualifying offense,” where qualifying offenses were limited to “state felonies, serious misdemeanors, or most violations of probation or parole” and “rearrest” is counted at the reported date of the offense (i.e. not the arrest date itself).
  2. Recommitment: “the first secure custody readmission (excluding administrative recommitments,” due to custody from new arrests, sentencing for a new conviction, or violations of probation or other release conditions.
  3. Return to Prison: “a return to Level V status with a sentence length greater than 12 months during the at-risk period,” including multiple contiguous Level V sentences.
  4. Reconviction, another commonly used metric for measuring recidivism, is no longer published by the SAC following the passage of DE SB98. In prior publications, the SAC defined reconviction as follows:

  5. Reconviction: “the first conviction event on any qualifying offense during the at-risk period,” with offenses subject to the same qualifying conditions as rearrest events.

The SAC outlines further considerations for all metrics in the methodology sections of its reports.

Three-Year Recidivism Rates Among All Release Cohorts (DE)

Rearrest
Reconviction
Recommitment
Return to Prison
Percent of individuals who recidivated within three years of release from Level V institutions in Delaware from 2008 to 2019, grouped by year of release. 2017 was the last year the state reported reconviction as a measure of recidivism (see all the SAC’s PDF reports here).


Besides determining whether recidivism rates are comparable across states, accessing these rates is itself difficult because there is no unified or standardized process for publishing data on recidivism. In January 2023, the Virginia Department of Corrections identified 35 states (including VA itself) which reported statistically comparable three-year re-incarceration rates for cohorts released in 2018 (or 2017, if that data was the most recent available). Of these, Delaware reported the second-highest rate at 53.9% behind Alaska at 61.6%.


Three-year re-incarceration rates across 35 US states. See the VA DOC’s report for an explanation of which states were included and excluded from their analysis. Data from each state is the most recent available as of the report’s publishing in January 2023.

While state governments typically focus on recidivating behavior resulting in re-incarcerations, any level of recidivism carries significant consequences for those affected. If someone is arrested following their release, even if not convicted or imprisoned, time spent in detention and away from work could cost them their job. This in turn will increase their chances of returning to more serious and continued criminal behavior.

What is Prison Gerrymandering?

When drawing electoral maps, legislators engage in gerrymandering if, rather than setting boundaries along “natural” lines like roadways and neighborhood divisions, they split their constituencies geographically for partisan gain. Conventional gerrymandering tactics can be classified as “cracking” (i.e. separating voting blocs which would likely support the opposing party) or “packing” (i.e. condensing dispersed communities of supporters into a single district).

The US Census Bureau counts incarcerated people as residents in the towns where they are imprisoned, rather than at their permanent residence. Before mass incarceration, the country’s prison population was low enough to make this practice relatively inconsequential. Today’s prisons, however, form attractive and impactful targets for gerrymandering. By shifting electoral boundaries to include a prison, state legislatures and redistricting commissions can artificially increase the population of an electoral district and its number of state representatives, even if those incarcerated in the facility cannot vote (as is in the case in 48 states). While most state constitutions and statutes indicate that incarceration does not alter someone’s legal residence, using unadjusted Census data in redistricting enables state and local governments to circumvent these restrictions.

Regardless of whether any electoral boundaries are shifted, counting incarcerated persons away from their true places of residence wrongly inflates the population of districts home to large prisons at the expense of other districts in the state, particularly those where incarcerated persons should rightfully be counted. Since most incarcerated persons live in urban areas, the more rural locations of many prisons can mean prison gerrymandering will specifically disadvantage urban areas:

How are Prison Gerrymandering and Recidivism Related?

Prison Gerrymandering and recidivism are significantly interrelated as racial injustices within the United States. Because black and brown people are both likelier to live in urban areas and are arrested and sentenced to prison at disproportionate rates, prison gerrymandering dilutes the voting power of these communities. Local legislators rightfully perceive that if incarcerated individuals within their districts cannot vote, they face no incentive to represent their interests. And since criminal and racial justice issues are most salient among those communities which are directly impacted, prison gerrymandering dilutes the political will behind reform efforts (including reforms directed at ending prison gerrymandering itself). This latter point forms a vicious cycle with recidivism: if underlying causes remain unaddressed, recidivism rates and prison populations will remain high, amplifying the effects of prison gerrymandering and its role in preventing solutions from being implemented.


At their core, both prison gerrymandering and recidivism are issues of enfranchisement and integration into broader society. While formerly incarcerated individuals can vote in many jurisdictions, throughout their sentences they are instrumentalized to serve interests other than their own. Upon release, barriers to their full reintegration into society persist, perpetuating a cycle of disenfranchisement and marginalization. These barriers can include limited access to employment opportunities, housing discrimination and instability, and the stigma associated with a criminal record. Without adequate support and resources, individuals leaving prison face significant challenges in rebuilding their lives and participating fully in civic life.

What can policymakers do to address recidivism?

With so many barriers inhibiting formerly incarcerated persons from re-integrating into society, policymakers must take a multifaceted approach to lowering recidivism rates, removing legal impediment and investing in comprehensive rehabilitation programs.

Clearing Criminal Records

A criminal record presents a significant obstacle to securing employment and housing, in addition to carrying a general stigma throughout an individual’s life. The Clean State Initiative outlines a model for state governments to automatically expunge criminal records for individuals who refrain from criminal behavior for a given time. Twelve US States have passed laws that clear CSI’s five baseline standards for Clean Slate legislation:

Map of Clean Slate Laws

Clean Slate Law Status

No Clean Slate Law
Clean Slate Law
Map of US states which passed legislation meeting clean-state standards to automatically expunge criminal records (CSI)

With SB111 and SB112, in 2021 Delaware became the fifth state to pass legislation meeting Clean Slate criteria. Set to be implemented on August 1st 2024, the Clean Slate Act will automatically expunge arrest and misdemeanor records and “makes additional felony convictions eligible for mandatory expungement 10 years after conviction” for individuals without prior or subsequent convictions.

Incorporating Mental Health Throughout Re-entry Programs

Social programs intended to assist formerly incarcerated persons in re-entering society should incorporate mental health treatment throughout the process, as recidivism expertDr. Debra Mason has argued throughout her career. “If you pay someone’s rent for 3 months and get them a job, but they can’t handle the stress at work, they will quit and then become homeless again,” Dr. Mason says. Even if someone released from prison does not return to criminal behavior, if their unaddressed mental health or substance abuse disorders prevent them from obtaining housing and employment, whatever re-entry programs were in place failed to meet their needs. The struggles with mental health that currently and formerly incarcerated individuals face often stem from traumatic experiences at young ages related to environmental and social factors associated with living in communities affected by generational poverty and systemic racism. “Re-entry should begin all the way back in someone’s school years,” says Dr. Mason. “Something must have gone seriously wrong in a person’s mental development for them to be able to shoot a childhood friend.”


Some of the scientific literature has failed to find a clear link between mental health diagnoses and recidivism. In one example published last year, a retrospective review of nearly 10,000 incarcerated persons released from the New Jersey Department of Corrections found no correlation between placement on the prison system’s Mental Health Special Needs Roster and three-year recidivism rates following release when controlling for substance abuse disorders. But social stigma poses a significant barrier to seeking mental health care and to reporting one’s symptoms with full transparency; in Dr. Mason’s experience, this is especially a problem in the Black community. “It’s much easier to say you ‘have an aunt who’s an alcoholic,’ versus saying your aunt has anxiety and depression that she deals with via alcoholism.” Furthermore, “mental health” in this context is often construed too broadly when the focus should instead be placed on the undiagnosed trauma endemic to communities impacted by mass incarceration. While conditions like schizophrenia or bipolar disorder are certainly prevalent among incarcerated populations, they are not the most widespread consequences of systemic racism on mental health and may not be the appropriate conditions to study when considering the connection between mental health and recidivism. Both of these considerations suggest that under-diagnosis of the relevant mental health conditions might be obscuring what connections are in fact present. In a similar vein, Dr. Mason points out that trauma has long been known to increase the risk of substance abuse, and that both substance abuse and dismissive attitudes towards it are common in the black community.


In her own dissertation research, Dr. Mason presented individuals with questions from the ACE assessment for Adverse Childhood Experiences, i.e. potentially traumatic events in childhood with known links to negative long-term outcomes on health, education, and employment. Dr. Mason favors the ACE assessment versus common alternative questionnaires for its focus on trauma, its shorter length, and simple questions, all factors which encourage accurate self-reporting.. While limited to a small convenience sample, her dissertation study found that individuals with an ACE score of five or more are more likely to offend after completing a re-entry program.

Probation Reform

As in the case of recidivism, it can be difficult to access the data necessary to fully evaluate the efficacy of a state’s probation and re-entry systems. One clear connection between recidivism and probation is its duration: in many states, probation and parole can last from 5 to 10 years. Over such a long period, “it doesn’t matter who you are – if you’re under supervision, they’re bound to find something,” Dr. Mason says, speaking to her own experiences as a probation officer in Delaware. The state has made massive strides by reducing the maximum duration of probation to 1.5 years, barring any violations. Beyond that direct policy change, it’s less clear what interventions may be effective. The wide range of discretion afforded to probation officers can be abused but also be beneficial, as probation officers are in a better position than any other state official to exercise judgment in individual cases. Dr. Mason also suggests that probation officers might not be getting the tools and support they need; for instance, official probation risk assessment forms might be biased against women, who are more likely to seek care, but are then penalized for it. Much of her work on the Corrections and Law Enforcement Committee, falling under the governor’s Behavioral Health Consortium, is dedicated to exploring these issues.

What can policymakers do to address Prison Gerrymandering?

Until the US Census Bureau starts counting incarcerated people as residents of their home addresses, state and local governments must pass legislation regulating the use of Census data for redistricting.

The Prison Gerrymandering Project at the Prison Policy Initiative lists solutions for state, county, and local governments to address prison gerrymandering at their respective level of jurisdiction.

  • State governments should implement a state-level census to collect the home addresses of incarcerated persons and use its findings to adjust the official US Census data. To move toward this specialized census, a state’s Department of Corrections can collect home addresses as part of a standardized intake process.
  • Local governments should omit prison population counts during redistricting; state governments can expedite this process with state-level legislation enforcing the practice.

Numbers of Prison Gerrymandering Bills Passed

Number of Bills per State

No bills passed
1 bill passed
2 bills passed
3 bills passed
4 bills passed
A map of states which have passed legislation ending prison gerrymandering as of August 2023, color-coded by the number of bills passed. See this quick reference chart from the PPI for more details.

The Delaware state legislature has passed bills in both the house (HB 384 , 2010) and senate (SB 171, 2021) to properly count incarcerated persons at their last known residences. With the passage of SB171, the Wilmington city council voted likewise to end prison gerrymandering within the city’s municipal districts. Since the 2010 redistricting cycle, residents of the Howard R. Young Correctional Institution were counted as residents of the city’s third council district. Drawn in 2021, elections will be held in the districts for the first time in 2024.

loading...
A map of Wilmington’s city council districts depicting how the 2020 Census misallocated the incarcerated population of Howard R. Young versus the allocation determined by SB171. See the city council website for the reference map and data tables.

Further Reading

The Prison Policy Initiative includes a helpful overview of its resources devoted to the issue of prison gerrymandering.

  1. Legislative History: An in-depth page recording state and local legislation.
  2. National Briefing Packet: Includes selected fact sheets, articles, news clippings, and resolutions.
  3. Model Legislation to end prison gerrymandering at a state level

The Redistricting Data Hub provides extensive datasets and articles on the issue of gerrymandering the United States.

  1. Estimating the Effect of Prison Gerrymandering in 11 states without legislative safeguards.
  2. Redistricting Basics: A short FAQ on redistricting and prison gerrymandering.
  3. Racially Polarized Voting Analysis: An introduction to statistical methods used to validate redistricting maps for compliance with the Voting Rights Act of 1965

Mass Incarceration Draws Its Own Maps and Creates a Country in Its Image (The New Republic)

Criminal Recidivism in Inmates with Mental Illness and Substance Use Disorders (March 1st, 2024)

  1. “A retrospective review of medical and legal records of all 9,669 inmates released from the [New Jersey Department of Corrections] in 2013.”
  2. The study explored the correlation between placement on the prison system’s Mental Health Special Needs Roster and diagnosis substance abuse disorders while in prison and recidivism following release.
  3. No correlation was found between mental illness and three-year recidivism rates when controlling for co-morbid substance-abuse disorders.
    • Three-year recidivism rates of offenders with a mental illness, but not a substance use disorder, were statistically similar to rates of offenders with neither problem (characterized as consistent with scientific literature by the study authors).
    • “Our third hypothesis, i.e., that recidivism would be highest among mentally ill inmates with co-occurring substance use disorder, was not supported by our data” (characterized as surprising given the scientific literature).

The Impact of Criminal Records on Employment, Earnings, and Tax Filing (September 29, 2023)