Auden Johnson/Shutterstock

Why Recidivism Isn’t a Strong Metric for Determining a Housing Program’s Success

Reduced recidivism rates are commonly used to determine whether a policy change or program intervention for justice-involved people has been successful. In turn, they drive decisions around funding allocations and program and research design (PDF).

However, recidivism is inconsistently measured (PDF) across studies, is limited by fragmented data sources (PDF), and fails to capture the nuances of successful reentry and service provision, such as behavioral health support and housing services. Because of this, it cannot accurately capture the impact of changes to individual criminal activity, but reflects “system decision-making” (PDF). And given the systemic racism embedded into the past and present criminal legal system, using recidivism to measure the success of reentry is itself a form of racial bias that does not improve public safety.

Researchers should consider using alternative measures of program success to ensure that program evaluations do not perpetuate systemic racism within the criminal legal system and do not widen disparities in access to critical reentry resources.

How does the field define recidivism?

The US Department of Justice’s National Institute of Justice defines recidivism as “criminal acts that resulted in rearrest, reconviction or return to prison with or without a new sentence during a three-year period following the person’s release.”

This binary definition doesn’t account for important factors such as differences in risk level (PDF), length of criminal record, or initial crime and sentence (PDF), all of which can affect a person’s likelihood to commit crimes. This can lead researchers to make comparisons across dissimilar populations, reducing the accuracy of recidivism as a measure of program or policy success (PDF).

Further, a “yes” or “no” measure of failure does not account for variation in subsequent crimes committed, such as desistance from five arrests in one year to one arrest in the following year. A reduction in the severity or amount of crime committed over time is an important metric to understanding how a policy or program affects criminal behavior, and ultimately, whether it improves public safety.

Using this definition of recidivism to measure success limits what is considered “successful” in program outcomes. It doesn’t account for the wide variety of program goals and designs that address the root causes of recidivism, such as housing instability. This may lead to inaccurate comparisons across programs that provide different services to clients.

How disparate law enforcement influences recidivism

Racial and economic disparities in recidivism reflect the intersecting legacies of disinvestment from communities of color and the racist origins of policing. These policies have led to higher concentrations of poverty and a greater risk for being overpoliced in communities of color than in white communities (PDF). This is especially true in Black communities. There are historic and ongoing racial disparities in many aspects of the criminal legal system, ranging from increased interactions with the police to higher rates of pretrial detention and incarceration rates.

A common measure of recidivism is rearrest. But because law enforcement agencies have historically—and still to this day—perpetuated racist policies and practices, utilizing rearrest rates can result in higher measured recidivism rates for communities of color.

Traffic stops are among the most common forms of police interactions (PDF) that result in an arrest—a frequently used indicator of recidivism. In jurisdictions across the country, Black drivers are more likely to be pulled over by police than white drivers, and both Black and Latinx drivers are more likely to be searched than white drivers. The same study found that Black drivers were less likely to be stopped by police after sundown, when officers are less able to discern a person’s race, indicating a potential bias in police decisionmaking. Similarly, though marijuana use is nearly equal across racial groups, Black people are nearly four times more likely to be arrested for marijuana than white people.

Ordinances that criminalize poverty by outlawing everyday activities of people experiencing homelessness—such as bans on camping and panhandling—contribute to the homelessness-jail cycle. These repeated interactions with the police increase opportunities for arrest, and therefore recidivism. Housing instability thus may influence a person’s likelihood of recidivating. Housing instability and housing insecurity are both more likely to be experienced by Black, Latinx, and Indigenous Americans compared with white Americans, further increasing the likelihood that there will be racial inequality in rates of rearrest.

Research shows that calls for service for low-level crimes increase in gentrifying neighborhoods where community members with low incomes, particularly those areas with high concentrations of Black and Latinx populations, are overpoliced for nuisance crimes such as graffiti or loud noises. This may lead to a disproportionate level of arrests for people in low-income communities.

How disparate community supervision influences recidivism

Access to stable and affordable housing is a key component of successful reentry and desistance from crime (PDF), yet formerly incarcerated people face significant barriers in securing and maintaining housing upon release.

For example, people on community supervision face strict conditions, such as restrictions on living with a person with a criminal history, which can result in parole violations if a person’s living options are limited. Though these violations do not constitute additional criminal activity, they can result in reincarceration via a revocation, which is often reflected as a recidivism event in the data.

Further, there are measured racial disparities in who receives revocations, with a 2014 Urban Institute (PDF) study finding that Black people on probation received revocations at higher rates than both white and Latinx people. The disproportionate impact of supervision on Black communities has important implications for equity, given the harm of incarceration on mental health outcomes and housing stability.


To effectively measure program success, researchers can consider the following:

  • Adopt inclusive community engagement strategies to ensure definitions and measurements of program success accurately reflect the priorities of people receiving services and community members. By grounding measures in the experiences of people both receiving and providing services, researchers can move away from measures, like recidivism, that do not meaningfully measure success for people in programs.
  • Focus research on measures of positive behavioral change and markers of desistance (PDF), such as decreases in the frequency and severity of criminal activity. Measures that examine reductions in risk factors for recidivism, such as whether a program successfully improves access to behavioral health services, employment, or gainful employment, as well as measures of retention, can help researchers understand program success.
  • Emphasize tracking housing placement for formerly incarcerated people in programs and more nuanced measures, such as the number of days a person has been successfully housed. Housing is a foundational key to successful reentry (PDF), and ensuring it is understood through measures that capture stability and accessibility of housing for justice-involved people is essential to understanding a program’s success.

Measuring success this way may require linking across administrative datasets or integrating data more intentionally at the outset of a program. However, once the infrastructure has been built, it may be easier for other programs to replicate research designs and could create more opportunities for collaboration across agencies interested in improving reentry for community members.

By reframing measurement from recidivism—a measure of failure—to measures that examine positive behavioral change, researchers can begin to more equitably measure and research what services contribute to a successful reentry program, as defined by people receiving and providing services. In future posts, I’ll explore alternative measures of success and innovative approaches service providers are taking to capture program achievements.