Criminal justice reform — the sustained effort to reduce mass incarceration, address racial disparities, improve police-community relations, and redirect criminal justice resources toward evidence-based alternatives — is both the most politically consequential application of criminological knowledge and the domain where the gap between what research supports and what politics produces has been most persistently visible and most costly. The United States incarcerates approximately 2 million people — the largest prison population in the world and the highest incarceration rate among comparable democracies — at an annual cost exceeding $80 billion, with documented costs to families, communities, and public safety that dwarf the expenditure figures. Criminological research has established with substantial confidence that much of this incarceration produces marginal crime control benefit at enormous cost while imposing devastating collateral consequences on the communities most affected, making the evidence case for reform among the strongest in applied criminology.
Criminology and Public Policy engages criminal justice reform as both a policy topic — what specific reforms the evidence supports — and an institutional topic — why evidence-supported reforms are politically difficult to achieve and what strategies have proven effective for navigating that difficulty. Understanding why mass incarceration persists despite its documented inefficiency and injustice requires analysis of the political economy of criminal justice that evidence synthesis alone cannot provide. And understanding what reform has been achieved — through the evidence-based corrections movement, through justice reinvestment initiatives, through prosecutorial reform, through sentencing reform, through police reform following consent decrees — requires careful evaluation of what specific strategies have worked in specific institutional and political contexts.
Introduction
The evidence foundation for criminal justice reform rests on three decades of accumulated research whose core conclusions have become increasingly unambiguous. Mass incarceration does not produce proportional crime control benefits: research on the marginal incapacitation effect of imprisonment at current American incarceration levels consistently finds that each additional imprisoned individual produces smaller crime reduction benefits as incarceration grows, because the highest-rate offenders are incarcerated earliest while marginal additions increasingly involve lower-risk individuals whose incarceration produces minimal crime control return (Travis et al., 2014). The collateral consequences of mass incarceration — for the children of incarcerated parents, for the communities from which prisoners disproportionately come, and for the labor market prospects of formerly incarcerated individuals — impose social costs that substantially exceed the direct costs of incarceration.
The racial concentration of mass incarceration constitutes a social injustice that the evidence supports characterizing as structural rather than incidental: Black Americans are incarcerated at approximately five times the rate of white Americans, a disparity that has grown substantially since the mass incarceration era began, and the policies that produced mass incarceration were both predicted to have racially disproportionate effects and sustained despite documentation of those effects. The moral case for reform is complementary to the efficiency case: mass incarceration is not only an ineffective crime control strategy but an unjust one, and both arguments point toward the same policy direction.
The reform agenda that the evidence supports encompasses multiple complementary components: reducing incarceration by shortening sentences, expanding diversion, and limiting the use of incarceration for low-level and non-violent offenses; reforming policing to reduce racial disparities and improve community legitimacy; strengthening reentry support to reduce recidivism; reforming prosecutorial practices to reduce racial disparities in charging and plea bargaining; and investing in the upstream social conditions — education, housing, employment, mental health — that reduce crime more cost-effectively than downstream enforcement. Each component is supported by substantial evidence and has been advanced through specific reform strategies whose effectiveness the research literature has evaluated. This article examines each major reform domain in turn — analyzing what the evidence supports, what has been achieved, what barriers remain, and what the political and institutional conditions for continued progress look like.
The Political Economy of Reform
Why Mass Incarceration Persists
The persistence of mass incarceration despite its documented inefficiency and injustice is explicable through the political economy of criminal justice — the structure of political incentives, institutional interests, and public opinion dynamics that determine what criminal justice policies are politically viable regardless of what the evidence supports. Politicians who advocate harsher sentences signal toughness on crime to fearful electorates; those who advocate decarceration expose themselves to the political risk of being blamed for any subsequent crime by released individuals. This asymmetry — diffuse statistical credit, vivid specific blame — creates rational political incentives for punitive overcorrection that persist even when policymakers understand the evidence against incarceration’s marginal effectiveness.
The network of institutional interests that has been labeled the prison industrial complex — corrections officer unions, private prison corporations, small towns economically dependent on prison employment, and prosecution and law enforcement agencies whose institutional strength is partly a function of criminal justice system scale — constitutes a political constituency for mass incarceration that actively resists reform and whose economic interests are directly threatened by decarceration (Gottschalk, 2014). Research documenting how these institutional interests have shaped criminal justice politics through legislative lobbying, campaign contributions, and organizational advocacy explains much of the disconnect between criminological evidence and criminal justice policy that would otherwise be difficult to account for.
Public opinion on crime and punishment is more nuanced than the “tough on crime” political dynamic suggests. Survey research consistently finds that when presented with detailed policy alternatives rather than abstract punitive/lenient frames, majorities support prevention, treatment, and rehabilitation alternatives over incarceration. Research by the Sentencing Project has documented that American public opinion has moved substantially toward reform over the past decade — particularly following the racial reckoning prompted by the killing of George Floyd — with majorities supporting reduced incarceration, police reform, and investment in social alternatives. This shift in public opinion has created political space for reforms that were not viable a decade earlier.
The Bipartisan Reform Coalition
The convergence of left and right on criminal justice reform — evident in the bipartisan coalition that produced the First Step Act of 2018 and in the growing number of states that have enacted sentencing reforms with bipartisan legislative majorities — represents a political development whose durability criminologists are still evaluating. The conservative case for sentencing reform, articulated through the Right on Crime initiative, emphasizes fiscal efficiency, federalism, and family values: incarceration is expensive and produces diminishing returns, states should experiment without federal mandates, and mass incarceration destroys families and communities. The progressive case emphasizes racial justice, human rights, and the investment in social alternatives that decarceration savings could fund.
Research evaluating the effects of state-level sentencing reforms — exploiting natural experiments created by variation in reform timing and scope — has generally found that sentencing reductions for low-level and non-violent offenses do not increase crime. A Brennan Center for Justice analysis examining 34 states that reduced incarceration between 2008 and 2016 found that these states also reduced crime at rates comparable to or greater than states that maintained or increased incarceration, strongly inconsistent with the proposition that reducing incarceration necessarily increases crime. The consistency of this finding across multiple analyses has provided the political cover that reform-skeptical legislators require before supporting decarceration.
Table 1. Criminal Justice Reform Strategies: Evidence and Assessment
| Reform Area | Specific Intervention | Crime Evidence | Equity Evidence | Key Challenge |
|---|---|---|---|---|
| Sentencing reform | Reducing mandatory minimums | No crime increase documented; cost reduction | Reduced racial disparities in some studies | Prosecutorial resistance; legislative politics |
| Prosecutorial diversion | Pre-arrest / pre-plea diversion to treatment | Comparable recidivism to prosecution; cost savings | Racial equity depends on eligibility criteria | Prosecutorial culture; treatment capacity |
| Bail reform | Eliminating cash bail; risk-based pretrial | No crime increase in strong evaluations | Reduces racial and economic disparity | Judicial resistance; risk assessment fairness |
| Police reform (consent decrees) | DOJ-negotiated comprehensive reform | Mixed; some find crime reductions | Reduced force; improved community relations | Departmental resistance; sustained monitoring |
| Justice reinvestment | Redirecting savings to prevention | Modest reductions in committed states | Varies by reinvestment priorities | Savings recaptured; political competition |
| Reentry investment | Comprehensive post-release support | Moderate recidivism reduction | Reduces inequality of collateral consequences | Interagency coordination; funding stability |
| Decriminalization | Reclassify low-level offenses | No crime increase documented | Substantial racial equity improvement | Political opposition; police culture |
| Record expungement | Automatic clearing after waiting period | Reduces recidivism by reducing employment barriers | Removes collateral consequence burden | Legislative politics; database systems |
The Racial Justice Dimension of Reform
Mass Incarceration as Racial Harm
The racial dimensions of mass incarceration are not incidental to the reform debate but central to it: the policies producing mass incarceration were racially targeted in their design and application, their racially disproportionate operation was documented in real time without producing reform, and the communities most devastated by those policies have had the least political power to change them. Michelle Alexander’s The New Jim Crow (2010) crystallized the racial justice argument for criminal justice reform for a broad public audience, arguing that mass incarceration had reconstituted racial caste through ostensibly race-neutral criminal law and enforcement mechanisms. The analytical debate about the adequacy of the caste analogy — whether mass incarceration is best understood as a mechanism of racial control or as a racially disparate response to genuine crime increases — has not diminished the reform mandate that the racial disparity evidence supports, regardless of the causal interpretation one adopts.
Research by Robert Sampson and colleagues using the Project on Human Development in Chicago Neighborhoods has documented that the communities experiencing the highest rates of incarceration are the same communities that already face the most concentrated disadvantage, weakest institutional infrastructure, and lowest collective efficacy — and that incarceration compounds these deficits rather than addressing them, by removing workers, parents, and community members from already fragile social structures. The community-level consequences of concentrated incarceration — reduced collective efficacy, weakened family structure, disrupted labor markets, and the criminogenic effects of returning mass incarceration graduates — may themselves elevate crime rates in affected communities, creating a vicious cycle in which incarceration produces the social conditions that produce more incarceration.
The reform implication is that reducing racial disparities in incarceration is not merely an equity objective but a crime control one: the social disorganization consequences of concentrated incarceration in the communities most affected may be undermining the crime control benefits that incarceration is designed to produce. Decarceration strategies that reduce the concentration of incarceration in the most affected communities — through sentencing reform, prosecutorial diversion, and supervision reform targeting the offenses most concentrated in those communities — simultaneously advance racial equity and address the social disorganization consequences that concentrated incarceration compounds.
Sentencing Reform and Decarceration
Evidence on Reducing Incarceration
The evidence case for reducing incarceration at the margins of the American prison population — by shortening sentences for low-level and non-violent offenders, expanding good-time credits, expanding eligibility for parole and early release, and reducing revocations to prison for technical supervision violations — is sufficiently well-established to support policy action with high confidence. Research on the crime consequences of the Fair Sentencing Act of 2010, which reduced the crack-powder cocaine sentencing disparity, and of the First Step Act’s retroactive application found no crime increases associated with the release of individuals whose sentences were reduced, consistent with the prediction that these individuals were not the high-risk, high-rate offenders whose incapacitation produces meaningful crime control benefits.
Research by Michael Mueller-Smith using administrative data from Houston found that for low-level offenses, prosecution and conviction substantially increased long-term criminal recidivism relative to diversion — a finding that reverses the naive expectation that criminal justice processing deters future offending and instead documents that processing is itself criminogenic for marginal offenders through the labeling, employment, and social disruption mechanisms that criminal justice contact produces (Mueller-Smith, 2015). This finding has particularly strong implications for the decarceration of low-level drug and property offenders, for whom prosecution and incarceration may be producing net crime increases rather than reductions.
Technical Violations and Supervision Reform
Technical violations of supervision conditions — returning to prison not for new crimes but for failing drug tests, missing appointments, or violating other conditions of probation or parole — account for approximately 25% of state prison admissions and represent one of the clearest targets for decarceration reform. Research on the deterrence value of incarceration for technical violations relative to graduated community-based sanctions has consistently found that swift, certain, but brief responses to violations produce better supervision compliance than the delayed, uncertain, but lengthy incarceration responses that conventional revocation practice relies on. The Hawaii’s Opportunity Probation with Enforcement (HOPE) program — which implements swift and certain but brief incarceration in response to supervision violations — demonstrated in a randomized trial that this approach reduced drug use and missed appointments among high-risk probationers more effectively than conventional supervision while substantially reducing total days incarcerated (Hawken & Kleiman, 2009).
Pretrial Justice and Bail Reform
The pretrial detention system — in which approximately 450,000 individuals sit in American jails on any given day, most of them awaiting trial rather than serving sentences, and many of them detained not because they pose genuine flight or safety risks but because they cannot afford to pay bail — has become a central target of criminal justice reform following research demonstrating its inequity and its crime consequences. Research by Will Dobbie, Jacob Goldin, and Crystal Yang using random assignment of bail judges as an instrument found that pretrial detention substantially increased the probability of conviction and of longer sentences among detained defendants, with effects operating through defendants’ reduced ability to participate in their own defense, their greater pressure to accept plea offers to secure release, and the direct testimony benefits that freedom provides compared to testifying from incarceration (Dobbie et al., 2018).
The bail reform movement has pursued multiple strategies: legislative elimination of cash bail for low and medium-risk defendants (enacted in New Jersey and other states); use of validated risk assessment instruments to guide release decisions; and advocacy for alternatives to detention including supervision, reminder calls, and court date assistance that research has found reduce failures to appear without the pretrial detention that bail is ostensibly designed to prevent. Research on New Jersey’s 2017 bail reform — which eliminated cash bail for most defendants and replaced it with risk-based pretrial assessment — found reductions in pretrial detention rates and in racial and economic disparities in pretrial outcomes, without documented increases in crime or failure to appear that reform opponents predicted.
Police Reform
What the Evidence Supports
The Department of Justice’s Civil Rights Division has conducted pattern-and-practice investigations of police departments finding systemic constitutional violations and has negotiated consent decrees requiring comprehensive police reform. Research evaluating consent decree implementation has found reductions in use-of-force incidents and civilian complaints, with studies by Stephen Rushin and colleagues finding no evidence of crime increases in consent decree cities and some evidence of crime reductions, directly contradicting the “Ferguson effect” narrative that police reform undermines crime control (Rushin & Edwards, 2017). The consent decree evidence provides the most systematic available evaluation of comprehensive police reform and supports its expansion as a tool for addressing systemic racial bias and excessive force.
The 21st Century Policing Task Force recommendations — developed under the Obama administration following the 2014 Ferguson protests — provide a comprehensive framework for police reform organized around six pillars: building trust and legitimacy, policy and oversight, technology and social media, community policing and crime reduction, training and education, and officer wellness and safety. Research evaluating jurisdictions that have implemented task force recommendations has found improvements in multiple dimensions of police-community relations and, in some cases, reductions in crime rates, suggesting that legitimacy-oriented policing is not merely normatively superior but operationally effective — producing better crime control outcomes through improved community cooperation alongside the equity improvements that legitimate policing provides directly.
Prosecutorial Reform
The Data-Driven Accountability Movement
The election of reform-oriented prosecutors in major American cities — including Philadelphia, Los Angeles, San Francisco, and Chicago — has created natural experiments in prosecutorial reform whose crime and racial equity consequences criminologists are actively evaluating. Research on Philadelphia District Attorney Larry Krasner’s sentencing reform policies found no significant increases in crime alongside substantial reductions in incarceration rates and in racial disparities in prosecution outcomes, providing early evidence that prosecutorial reform can achieve racial equity goals without the public safety costs that opponents predicted (Doherty, 2019).
The Institute for Innovation in Prosecution and the Vera Institute‘s work on prosecutorial accountability have developed data infrastructure — case management systems that enable prosecutors to analyze their own charging and disposition data for racial and economic disparities — that creates transparency accountability mechanisms the discretionary nature of prosecution has historically lacked. Making prosecutorial data publicly available and requiring regular disparity analysis are structural reforms that the evidence suggests would reduce racial disparities by making them visible and holding prosecutors accountable for their reduction.
Collateral Consequences Reform
The approximately 40,000 federal and state laws that impose legal disabilities on individuals with criminal records — restricting employment, housing, voting, education, and public benefits — represent a secondary punishment system that operates independently of the formal sentence and that extends criminal justice consequences indefinitely after sentence completion. The American Bar Association’s National Inventory of the Collateral Consequences of Conviction has catalogued these restrictions by state and offense category, providing the empirical foundation for advocacy targeting the most harmful and least justified restrictions. The reform of collateral consequences — through automatic expungement, removal of lifetime bans on public housing and student loans for non-violent offenders, individualized licensing assessment, and restoration of voting rights — reduces the collateral consequence burden that compounds disadvantage while the research suggests would reduce recidivism through the employment, civic reintegration, and stability pathways that collateral consequence reform addresses.
The Clean Slate Initiative‘s legislative progress — with automatic expungement enacted in a growing number of states — represents the most consequential collateral consequences reform movement of the current period. The Sentencing Project‘s documentation that approximately 5.2 million Americans are disenfranchised by felony conviction laws — unable to vote in elections that determine the criminal justice policies most affecting them — provides both the equity case and the democratic legitimacy case for restoring voting rights at release from incarceration, a reform already adopted by Maine, Vermont, and the District of Columbia and supported by growing majorities in state-level polling.
Justice Reinvestment
The Evidence from State Experiments
The Justice Reinvestment Initiative (JRI), administered by the Council of State Governments Justice Center, has worked with over 35 states to analyze corrections data, identify evidence-based alternatives to incarceration, and develop legislation implementing reforms projected to reduce both incarceration costs and recidivism. Research evaluating JRI states has found that states that implemented both sentencing reforms and reinvestment in evidence-based programs achieved modest but consistent recidivism reductions alongside substantial cost savings, while states that enacted sentencing reforms without reinvestment achieved cost savings with less clear recidivism benefits.
The most important lesson from the JRI experience is that the reinvestment component is as essential as the decarceration component. The crime control case for reform rests not only on ending the marginal inefficiency of mass incarceration but on redirecting those resources toward prevention and reentry investments that produce larger crime reductions per dollar than the marginal incarceration they replace. States that failed to protect reinvestment from legislative competition — with corrections savings captured by general revenue reductions rather than reinvested in prevention — failed to achieve the crime reduction improvements that the full JRI framework predicts and that more committed implementers have achieved.
Record Expungement and the Second Chance Infrastructure
The Growing Expungement Movement
Record expungement — the legal process by which criminal convictions are sealed or cleared from public records, removing or reducing their consequences for employment, housing, and civic participation — has emerged as one of the fastest-growing criminal justice reform areas, with 28 states enacting new or expanded expungement laws since 2018. The evidence on expungement’s crime prevention benefits supports expansion: research by J.J. Prescott and Sonja Starr using Michigan administrative data found that individuals whose records were expunged showed substantially reduced recidivism relative to comparable individuals whose records remained, with the effects operating through improved employment outcomes that the removal of the criminal record barrier enabled (Prescott & Starr, 2020).
The Clean Slate Initiative — a national advocacy campaign supporting automatic record clearing after conviction-free waiting periods — has successfully passed automatic expungement legislation in Pennsylvania, Michigan, Utah, Connecticut, and other states, with additional states considering similar legislation. Automatic expungement addresses the principal limitation of petition-based expungement — that only individuals who navigate a complex legal process benefit, systematically excluding those with least access to legal assistance — by clearing records automatically without requiring any individual action. Research on the uptake rates for petition-based expungement consistently finds that a small fraction of eligible individuals successfully petition for expungement, suggesting that most of the crime prevention benefits of expungement access currently go unrealized because of procedural barriers that automation would remove.
The intersection of expungement with employer background check practices illustrates the systemic challenges of criminal justice reform: even when criminal records are legally expunged, commercial background check companies may retain and report expunged records that have not been removed from their databases, undermining expungement’s employment benefits through data systems that the expungement order does not reach. Addressing this gap — through data furnisher requirements, audit mechanisms, and private right of action for improper expunged record reporting — is a technical but consequential component of expungement reform that the Clean Slate movement is beginning to address.
Community Reinvestment and the Social Determinants of Crime
The most comprehensive criminal justice reform frameworks argue that reducing mass incarceration without simultaneously investing in the communities most affected by both crime and criminal justice enforcement will not produce the public safety improvements that justify the political investment reform requires. Research on the social determinants of crime — reviewed in detail in the articles on poverty, housing, education, employment, and community-based prevention — consistently documents that upstream social investment produces larger crime reductions per dollar than equivalent downstream enforcement investment at the margins of the American criminal justice system. Justice reinvestment frameworks that explicitly connect decarceration savings to upstream social investment — in schools, housing, jobs, mental health, and community organizations in the communities most affected — provide the policy architecture for a comprehensive reform agenda whose crime prevention rationale is as strong as its equity rationale.
The Violence Interrupters model and the federal community violence intervention funding in the American Rescue Plan represent investments in the community-level crime prevention that criminal justice reform’s decarceration component makes space for by redirecting resources from incarceration to prevention. Research on community violence intervention programs’ effectiveness, reviewed in the article on community-based crime prevention policy, provides the evidence base for these investments. Building the political coalitions that can sustain this multi-component reform agenda — combining decarceration, police reform, prosecutorial reform, and social investment in a coherent framework — is the defining political challenge for criminal justice reform in the current period.
Research on the consequences of the 1994 Crime Bill — the Violent Crime Control and Law Enforcement Act signed by President Clinton — and its role in driving mass incarceration has generated a productive criminological debate about how to evaluate complex legislative interventions with multiple simultaneous components. The bill included truth-in-sentencing incentives, the assault weapons ban, Violence Against Women Act provisions, and community policing investment alongside the mandatory sentencing escalations most associated with mass incarceration’s growth. Research attempting to isolate each component’s crime and incarceration consequences has found that the truth-in-sentencing provisions contributed to incarceration growth without proportional crime benefits, while the COPS program’s community policing investment was associated with crime reductions that the enforcement components did not achieve independently. This decomposition — finding within a single landmark legislative package that the enforcement components were counterproductive and the community investment components effective — illustrates the kind of rigorous policy evaluation that criminal justice reform most needs and that the political environment least reliably produces.
Conclusion
Criminal justice reform is the domain where criminological evidence most directly confronts political resistance — where the gap between what research supports and what political institutions produce is most visible and most consequential for the millions of people whose lives are shaped by criminal justice policy. The progress of the past decade — the First Step Act, growing state-level justice reinvestment, prosecutorial diversion expansion, police reform through consent decrees, the movement toward bail reform — represents meaningful advance driven significantly by criminological research. The First Step Act alone has produced the release of thousands of individuals whose sentences exceeded any crime control justification, with documented non-increases in recidivism that have expanded the political viability of further reform.
The measurement of reform progress itself requires more systematic attention than the criminal justice reform movement has consistently provided. Tracking incarceration rates, racial disparities, recidivism rates, and crime rates at the jurisdiction level — with the methodological rigor that distinguishes genuine reform assessment from advocacy metrics — enables the cumulative knowledge building that evidence-based reform requires. The Council of State Governments Justice Center‘s data dashboards, the Sentencing Project‘s state-level tracking, and the Bureau of Justice Statistics‘s national data infrastructure all contribute to the measurement ecosystem that accountability requires. Building this infrastructure more comprehensively — and ensuring that measurements are conducted with sufficient independence and methodological rigor to support causal conclusions rather than merely documenting trends — is among the most important investments the criminal justice reform movement can make in the evidence base for continued reform.
The communities most affected by mass incarceration have yet to experience the systemic change in their relationship to criminal justice that the research evidence supports and the reform movement aspires to achieve. Building the political conditions for that change requires the sustained communication of criminological evidence to policy audiences, the development of reform-oriented criminal justice leadership from within the institutions that must change, and the cross-coalition partnerships — between conservatives concerned about fiscal efficiency and progressives concerned about racial justice, between practitioners willing to challenge their own institutional cultures and advocates pressing for accountability — that make reform politically viable in the adversarial environment that surrounds criminal justice in American politics.
References
- Doherty, F. (2019). Prosecutorial discretion and mass incarceration. In Oxford Research Encyclopedia of Criminology. https://doi.org/10.1093/acrefore/9780190264079.013.471
- Gottschalk, M. (2014). Caught: The prison state and the lockdown of American politics. Princeton University Press.
- Hawken, A., & Kleiman, M. (2009). Managing drug involved probationers with swift and certain sanctions: Evaluating Hawaii’s HOPE. National Institute of Justice. https://www.ojp.gov/pdffiles1/nij/grants/229023.pdf
- Mueller-Smith, M. (2015). The criminal and labor market impacts of incarceration. University of Michigan Working Paper.
- Rushin, S., & Edwards, G. (2017). An empirical assessment of consent decrees. Stanford Law Review, 69, 1555–1618.
- Travis, J., Western, B., & Redburn, F. S. (Eds.). (2014). The growth of incarceration in the United States. National Academies Press. https://doi.org/10.17226/18613