Policy Learning and Criminal Justice Reform in the U.S. States

Introduction

A national criminal justice reform movement is well underway in the United States (Schoenfeld 2011). Over the past fifteen years several notable bills have been signed into law at the national level. The Second Chance Act (2007) expanded prisoner reentry programs to assist returning inmates reintegrate into free society. The Fair Sentencing Act (2010) reduced sentences for crack cocaine. The law marked the first time Congress shortened federal penalties for drugs in more than four decades (Percival 2015). In 2017 the First Step Act shortened mandatory sentences for nonviolent federal drug crimes and expanded rehabilitation opportunities in federal prisons (Kelley and Rizer 2019).

These laws marked an important departure from law and order politics that characterized criminal justice policy making over the past generation (Beckett and Sasson 2000; Simon 2007). Even so, federal reforms have produced only modest reductions in the overall incarceration rate. ¹ The federal government plays a small role in the administration of criminal law. Nearly 90 percent of the 2.1 million prisoners in the U.S. are confined in state or local correctional institutions (U.S. Department of Justice 2020). Ultimately the extent to which the U.S. decarcerates will be decided by the policy decisions of state and local governments. ²

The criminal justice reform movement is the byproduct of a collective shift in attention in the criminal justice policy subsystem (see Jones and Baumgartner 2005). Scholarship on agenda setting and policy change shows political messages and problem indicators regularly flow into the political system. In most cases these indicators are ignored as policymakers working across decentralized institutions maintain the status quo (Baumgartner and Jones 1993). But periods of policy stability can be interrupted by a process of “alarmed discovery”—a moment when a problem long-ignored suddenly captures the attention of lawmakers. If attention shifts to new dimensions of an issue—that is, if enough policymakers come to understand a problem in a different or novel way—this can create opportunities for policy change (Jones and Baumgartner 2005).

Between 1972 and 2009 the state prison population increased 700 percent as tough on crime politics held a nearly universal grip (Ghandnoosh 2019). By the mid 2000s, with many states suffering from overcrowded prisons, rising prison costs, and deepening racial disparities, criminal justice reform emerged on the agenda as policymakers shifted their attention to the social and economic costs of mass incarceration (Percival 2015). Texas enacted “breakthrough” sentencing and probation reforms in 2007 (Fabelo 2010). Other conservative states including South Carolina (2010), North Carolina (2011), Georgia (2012), and Mississippi (2014) adopted reforms soon after. In the 2014–2015 legislative cycle, more than 200 bills were adopted in forty-six states (Silber, Subramanian, and Spotts 2016).

While the emergence of the issue represents a remarkable policy departure, scholars know relatively about the processes that affect state policymakers’ relative embrace of criminal justice reform. This article places emphasis on processes and mechanics of policy learning. What factors affect whether state policymakers, particularly conservative Republican lawmakers who have led the tough on crime movement, accept (or reject) less punitive policy alternatives? How does policy learning and Republican support for reform affect a state’s pathway to policy adoption?

Using a qualitative comparative case study of criminal justice reform in Texas and California between 2005 and 2011, this analysis reveals a policy learning process conditioned by each state’s political environment. Republicans in Texas legislated in an informational environment where conservative policy entrepreneurs had the willingness and wherewithal to frame reform in ways that matched Republican lawmakers’ core beliefs. Republican legislators enthusiastically embraced “smart on crime” reforms and joined with Democrats to build a bipartisan coalition for policy change.

In contrast, Republican lawmakers in California remained in steadfast opposition despite overwhelming pressure to address prison overcrowding and clear examples of successful, conservative-endorsed reforms in earlier adopting states. Republicans in the Golden State were legislating in a different information and political environment than their co-partisans in Texas. California had no conservative reform-focused entrepreneurs active in the debate. With few political incentives to support reform, Republicans in California demonstrated little interest in learning from the experiences of their co-partisans. As a result, California’s path to reform was slow, politically contentious, and chaotic. Overall, the analysis shows how policy learning shapes Republicans’ relative support for criminal justice reform.

The Conditional Nature of Policy Learning

Policy learning is treated as a malleable concept in the state politics and public policy literatures. It is a process whereby lawmakers draw positive and negative lessons from a public policy (Karch 2007, 61). The direction of these lessons can change over time even for the same policy. Dobbins, Simmons, and Garrett (2007, 460) describe learning as a “process whereby policymakers change their beliefs about the effect of policies.” Sabatier and Jenkins-Smith (1988, 19) define learning as policymakers’ “relatively enduring alterations of thought or behavioral intentions that result from experience.”

Regardless of the exact conception, learning is purposeful. Lawmakers can incorporate new information to help update their opinions about policy. Major changes in lawmakers’ beliefs are relatively rare, however. This generates stability in most policy domains (Beland 2010). Yet there are instances when policies produce such adverse social consequences that they begin to weaken the policy’s bases of political support. In these cases, politics can “run against the grain of current policy: rolling back existing programs or otherwise effecting a larger shift in the distribution of policy benefits and burdens (Jacobs and Weaver 2014, 444).”

Policy learning and growing concern about the burdens and costs of the status quo can generate policy change. Learning appears to be an influential driver of reform at the national level. After years of pushing a tough on crime agenda many national-level Republicans in the mid-to-late 2000s began acknowledging the social costs of mass incarceration and searched for less punitive alternatives (Aviram 2015). Republicans’ newfound embrace of reform gave birth to new bipartisan coalitions in Congress. It allowed Democrats who began emulating conservatives’ punitive positions in the 1990s, to once again pledge support for rehabilitative and preventative solutions to crime by the 2010s (Percival 2015).

Yet important questions remain. To what extent do these interparty dynamics unfold at the state level and what are the conditions, content, and sources of information that persuade Republicans to change policy directions?

At the state level we might expect Republicans to learn from both internal and external information sources (Karch and Cravens 2014). Within a state, lawmakers are regularly exposed to policy entrepreneurs who draw attention to new problems, reframe the meaning of old problems, and propose new alternatives in the hope of achieving their preferred outcomes (Mintrom 1997).

The presence, strength, and political strategy of reform-minded policy entrepreneurs is likely to shape a state’s criminal justice informational environment and the degree to which lawmakers learn. But the challenge for policy entrepreneurs is that lawmakers with strong partisan and ideological commitments resist new information (Baumgartner and Jones 1993; Schattschneider 1960). Information that forces an admission of error, threatens partisans’ political identity, or that misaligns with their electoral interests will usually be rejected (Gilardi 2010). How policy entrepreneurs frame the costs of mass incarceration and the meaning of less punitive alternatives should affect how receptive Republicans are to new information and their willingness to learn. This is no easy task: incarceration and tough on crime policies became a major part of the Republican Party’s identity by the 1980s (Smith 2004).

Dagan and Teles (2015) argue that unlocking partisan cognitive barriers requires policy entrepreneurs who can communicate policy ideas that conform with lawmakers’ core political identity. Entrepreneurs proposing new ideas to an ideologically skeptical audience must be trusted. They must share certain experiences, ideological beliefs, and identities with the policymakers they are trying to persuade. As Dagan and Teles (2015, 128) note, these “identity vouchers” are “people of high esteem who can help address the contradictions that inevitably crop up in any movement of distinctive factions held up by an attachment to broad principles.” New policy ideas and policy frames can be persuasive if they conform with partisans’ foundational values and principles in overlooked or novel ways.

State lawmakers also learn from external sources, namely other states and jurisdictions who previously adopted an innovative policy (Karch 2007). One of the central insights of the policy diffusion literature is that lawmakers learn about the effectiveness of policies and programs from the experiences of earlier adopting states (Meseguer 2006; Mooney 2001). Officials draw positive or negative lessons from these policy experiences and emulate policies deemed effective and successful (Volden 2006).

The information gleaned from earlier adopting states is often filtered by lawmakers’ partisan and ideological beliefs. Resource constrained policymakers use states’ ideological leanings as an informational shortcut (Grossback, Nicholson-Crotty, and Peterson 2004; Shipan and Volden 2008; Volden 2006). Drawing on experimental data, Butler and Pereira (2018) found officials expressed more interest in learning about a policy if it was endorsed by co-partisans in other jurisdictions. In a related study, Butler et al. (2017) found officials ideologically predisposed against a policy expressed less interest in learning from others. Importantly, however, these same officials were more likely to update their prior biases, and express interest in learning, if they were told the policy was previously adopted by co-partisans. Politicians are naturally reluctant to take a political risk in support of a program that doesn’t match their core political identity, but if they learn co-partisans in another jurisdiction already adopted it, they are more likely to want to learn more.

This body of research leads to several expectations about policy learning and criminal justice reform. First, Republicans’ acceptance of new policy ideas will be conditioned by a state’s political informational environment. A state with policy entrepreneurs with the capacity and wherewithal to frame criminal justice reform in a manner that conforms with conservatives’ core ideological and partisan beliefs should increase Republican support for reform. Republican learning and support for reform should facilitate bipartisan coalition building as Democratic lawmakers become less fearful about endorsing less punitive crime policies.

Consistent with the policy diffusion literature, Republicans in a later adopting state might be expected to give greater consideration to a policy reform endorsed by co-partisans in earlier adopting states. Yet learning might still be conditioned by electoral considerations (Gilardi 2010). A state’s internal political environment may create electoral incentives that decrease lawmakers’ willingness to learn from what is normally a trusted group of co-partisans in an earlier adopting state. Short term political calculations, in other words, may introduce cross pressures to the partisan learning process. In state legislatures party members make policy decisions for both sincere (ideological) and strategic reasons (Barrilleaux, Holbrook, and Langer 2002). They must consider how new information and new policy ideas fit in relation to the status quo.

Chappell and Keech (1986) show that parties can behave strategically to help attract voters who might normally vote for the opposite party. Suppose Republican lawmakers in a later adopting state, call it state “B,” hold little formal institutional power but through experience have learned that a law and order agenda is an influential piece of their reelection strategy. As a result they may be less willing to learn from criminal justice reforms adopted and endorsed earlier by Republicans governing in state “A.” For Republicans in state B, short term (internal) political and electoral considerations might be given greater weight than the ideological match to co-partisans in state A. The bottom line is that a state’s internal politics may diminish (or enhance) partisans’ willingness to learn.

Examining Criminal Justice Reform in California and Texas

To explore these ideas in greater depth I present a comparative case study of policy learning and reform in two states: California and Texas. Both states began a drive for criminal justice reform around 2005. Efforts in both states ended with the successful adoption of reforms that over time have contributed to a significant reduction in their respective prison populations. ³

Although each state achieved legislative success there are important differences between them. Texas adopted its reforms in 2007. The state might be thought of as an influential “early” adopter; indeed the “surprise” adoption of criminal justice reform in tough on crime Texas sent important signals about conservative leadership on the issue throughout the federalist system (T. Fabelo, pers. comm.). Republicans, the majority party in the Texas legislature, joined with Democrats to adopt a bipartisan package of justice reform bills. Legislation earmarked nearly $250 million for in-prison and post-release substance abuse treatment programs. Drug courts were expanded to assist “low-risk” offenders enter treatment programs. People sentenced to probation for low-level drug or property crimes saw their sentences cut by as much fifty percent (Fabelo 2010).

Reforms in California were adopted four years later in 2011. The state’s pathway to reform was far more challenging and chaotic. Republicans fought vehemently against policy change despite evidence of severe prison overcrowding, pressure from the federal courts to decarcerate, and clear examples of successful Republican-endorsed criminal justice reforms in Texas and other conservative states. Reforms were finally adopted when the Democratic-controlled legislature pushed through AB 109, a bill better known as “Prison Realignment,” on a partisan-line vote. Prison Realignment shifted responsibilities for managing “lower risk,” “non-violent” offenders to county governments. Virtually all people charged with drug and property crimes were no longer eligible to receive a state prison sentence (Petersilia and Snyder 2013). The law made it nearly impossible to return parolees or probationers to state prison on “technical violations.”

The analysis that follows examines how learning, as indicated by Republicans’ relative embrace of reform, affected each state’s pathway to reform and the relative ease by which legislation was adopted. The cases were selected following a loosely based application of the most similar systems design (Anckar 2008). ⁴ Using this approach two cases (e.g. states) are chosen that share similarities across background conditions that might be relevant to the outcome of interest (Seawright and Gerring 2008, 304). The goal is to keep “constant” as many extraneous variables as possible.

In terms of “background” conditions, Texas and California shared many commonalities as the drive for reform began in the mid 2000s. California (liberal) and Texas (conservative) have very different ideological orientations. However, Jonathan Simon (2007) persuasively argues that California possesses an orientation around crime that shares many similarities to the punitive culture in southern states. In addition to similar “crime cultures,” each state had a large prison population. In 2006 Texas’s prison population reached 152,000, second only to California’s 173,000 (Pew Charitable Trusts 2007). Law and order politics had flourished in each state for decades. Both states, for example, made heavy use of death sentences. Since the reinstatement of the death penalty in 1976, Texas has executed more people than any other state. California, meanwhile, has had the largest death row population in the country. Both states’ prison systems placed about equal budgetary pressure on the government. In 2007, corrections comprised 8.7 percent of Texas’s general fund budget, placing it sixth highest among states in terms of state general fund tax dollars spent on corrections. California was ranked fifth with 9.1 percent (Pew Charitable Trusts 2007). In both states tough on crime policies were pushed by both Republicans and Democrats. Texas went on an incarceration binge under Democratic governor Anne Richards. In California, Democratic governor Gray Davis took hardline stances on the death penalty, criminal sentencing, and parole. Democratic lawmakers in the California Legislature regularly joined with Republicans to pass punitive sentencing laws (Page 2011).

During the reform process, however, the two states differed on one key dimension: Republican lawmakers’ relative support and enthusiasm for reform. This variation, which I attribute to differences in policy learning among Republicans, is shown to affect the ease of policy adoption as assessed by two indicators: the presence or absence of bipartisan votes on reform legislation and the timing of reform.

One of the benefits of a small-N methodological design is that it allows the researcher to incorporate richer, more contextualized detail into a study (Nicholson-Crotty and Meier 2002). I collected information from in-person interviews, newspapers, policy reports, and legislative archives to present a contextualized analysis of the processes that contributed to Republicans’ support or opposition to reform and how learning, as measured by a qualitative assessment of Republican lawmakers’ discourse and position taking on reform, affected each state’s pathway to policy adoption. In Texas I document the actions and strategies of policy entrepreneurs who framed criminal justice reform around dimensions that appealed to conservative Republicans’ core identity. Some of these same entrepreneurs also supplied technical policy analysis which helped lawmakers design more effective programs. In California I provide detail of a very different political environment absent conservative reform-focused policy entrepreneurs. I explore how this caused Republicans to reject internal calls for reform but also policy ideas endorsed by co-partisans in early adopters like Texas.

A drawback to small-N studies is they often lack external validity (Nicholson-Crotty and Meier 2002). These studies risk overselling the generalizability of results. Texas and California have particularly punitive penal cultures with unusually large prison systems (Perkinson 2010; Simon 2007). And as we’ll see in the case of California, an unusual intervention by the U.S. Supreme Court placed pressure on Democrats to adopt reforms on a party-line vote in the eleventh hour. Despite these drawbacks and limitations these two states provide a useful environment to offer an initial test of an underlying theory of policy learning.

Criminal Justice Reform in Texas

In 2005, Texas was forecasting a 17,000 prison bed shortage (Fabelo 2010). State lawmakers could have done the familiar: build more prisons. Prison overcrowding in the 1970s had led to strict oversight of the state’s prison system and hard population caps. When prisons filled up Texas quickly built new ones. It was thus a surprise when Republican House Speaker Tom Craddick, a hard-nosed fiscal conservative, gave a directive to the new chair of the House Corrections Committee, Republican Jerry Madden: “Don’t build more prisons—they cost too much (J. Madden, pers. comm.).”

Texas’s capacity to design innovative and effective prison alternatives required access to good data. Texas had long relied on its own Criminal Justice Policy Council for policy analysis, but Governor Rick Perry cut the agency’s budget in 2003, thus stripping the state of badly needed expertise (Bernstein 2005). State Democratic Senator John Whitmire, the chair of Senate Criminal Justice Committee and close partner of Madden in the drive for reform, noted the problem at the time. “We’ve been handicapped for good information. We wouldn’t be in this crisis about capacity if we had had good information. We’ve been hurt tremendously (Burka 2007).”

Madden began looking to alternative sources he could trust. But to understand policy learning and Republicans’ eventual support for reform it’s important to take a step back and ask why Tom Craddick set Madden forth on a new mission in the first place. How did the debate around prison shift to thinking about costs and savings rather than simply locking people up no matter what the cost? The answer is that Texas lawmakers were hearing new messages and policy ideas from trusted conservative interest groups.

One of the most influential entrepreneurs in this area was Prison Fellowship Ministries (PFM), a non-profit evangelical organization whose work in Texas focused on “biblical justice” and rehabilitating prisoners through the teachings of Jesus Christ (Percival 2015, 99). The PFM was founded by Chuck Colson in 1976 following his release from federal prison for Watergate-related crimes. Pat Nolan headed the PFM’s political lobbying unit called the Justice Fellowship. A former leader in the California Assembly in the 1980s, Nolan led dozens of tough on crime bills through the California legislature (Percival 2015). His elected career came to an abrupt stop after he was sentenced to two years in prison on federal political corruption charges (Mai-Duc 2019). Nolan’s time in prison laid bare the dehumanizing experience of incarceration.

The Justice Fellowship worked to reframe debate along a different moral dimension. For many evangelical conservatives, the moral, spiritual, and expressive features of public policy affirm core cultural values of justice. Policies that express morally right values become ends in themselves. Punitive criminal sanctions were understood as expressions of moral outrage against criminals and drug users (Tonry 2011). Those who broke the social contact needed to be punished. Now the Prison Fellowship Ministries was framing prison reform around a new morality that resonated with many conservatives. They argued for rehabilitation-focused policies based on a belief that prisoners deserved the right to human dignity—a right shared among all people. They argued that people who completed their sentence and served their debt to society deserved a “second chance.” They decried the moral failings of mass incarceration.

When Texas Republicans learned to think about criminal justice through a lens of prisoner rehabilitation and “second chances” it offered a conceptual bridge to understanding prison reform through a second foundational conservative belief: government efficiency and shrinking the size of government. Republican Governor Rick Perry linked the two dimensions in a State of State Address in 2007, “I believe we can take an approach to crime that is both tough and smart…there are thousands of nonviolent offenders in the system whose future we cannot ignore. Let’s focus more resources on rehabilitating these offenders so we can ultimately spend less money locking them up again (Zeidman 2012).”

The work of reframing criminal justice reform around themes of government efficiency and cost savings was led by the Texas Public Policy Foundation (TPPF), a conservative-leaning Think Tank that was a member of the libertarian State Policy Network (Dagan and Teles 2015). One of the TPPF’s major donors was Tim Dunn, an evangelical Christian with strong libertarian leanings. Dunn had grown dismayed over seeing Texas’s prison population continue to rise: “I had come to see our justice as imperial, as intent on maintaining the authority of the king. It was no longer communal or restorative (King 2013).”

During the summer of 2006, Madden, Whitmire, and other legislative leaders worked closely with the TPPF. Around the same time the legislature received technical assistance from the non-partisan Council of State Governments (CSG) Justice Center. The CSG’s technical analysis identified three drivers of prison growth in the state: Too few prisoners were paroled; probationers and parolees were being readmitted to prison on technical violations at high rates; the state lacked sufficient residential and community drug treatment capacity (T. Fabelo, pers. comm.).

These findings influenced the legislation that Madden developed with the help of the TPPF, the CSG, and a growing coalition of partners on the political left. Reform bills invested in drug treatment, expanded prison diversion programs, and reformed the state’s probation systems.

Texas Republicans’ embrace of criminal justice can be directly tied to an informational and political environment conducive to learning. Trusted conservative policy entrepreneurs effectively framed the issue in ways that resonated with Republicans’ moral beliefs, and their fidelity to fiscal conservatism and smaller government. Criminal justice reform came to personify what it meant to be a conservative in Texas.

Learning appears to be facilitated by more than ideology, however. That Republican lawmakers spent resources designing effective reforms suggests they saw political benefits in adopting a “smart on crime” agenda. Republican support helped produce super majority, bipartisan votes in the legislature. A parole reform bill, SB 166, passed 31-0 in the Senate and 143-1 in the House (T. Fabelo, pers. comm.). On a drug treatment bill, HB 1, 64 percent of majority vote across the two chambers were cast by Republicans (Chettiar and Gupta 2011). The importance of Republican leadership on reform was not lost on Democrat John Whitmire. “For a Democrat you need political cover. I had to be tough and smart.…we needed to do much better on nonviolent offenders. And to do this, to get smart, I also needed Republican support (J. Whitmire, pers. comm.).” Republicans enjoyed the privileges and power of the majority party in the Texas Legislature. Their support of reform offered a chance to reorient their party image around innovative “smart on crime” policy solutions and the promise of bringing tax dollar savings to their conservative base (Madden 2010).

Criminal Justice Reform in California

In the mid aughts California was facing a similar yet more acute correctional crisis after years of failed reform attempts. The prison population climbed past 163,000 in 2006, double the system’s designed capacity. Crime policy in the Golden State been ruled for years by the California Correctional Peace Officers Association (CCPOA), the powerful prison guards union. The union spent lavishly on politics. More than 35 percent of its membership revenue was spent on lobbying (Page 2011). They contributed money to both Republican and Democratic candidates. Their political work undermined reform proposals and kept attention focused squarely on dimensions of the crime policy debate that produced more prisoners. This meant elevating the public’s fear about crime and raising public confidence in the utility of prisons. Over time the Republican Party’s identity on crime mirrored that of the CCPOA’s; Republicans raised fear and supported policies like “three strikes” and other sentencing enhancements.

By 2009, and with numerous reform efforts already thwarted, Republican Governor Arnold Schwarzenegger and the Democratic-controlled legislature faced pressure from a federal three-judge panel to reduce the population of the state’s unconstitutionally crowded prisons by 40,000 inmates. If the state wasn’t successful it would face a court ordered (and likely less structured and vetted) prisoner release.

A Schwarzenegger-backed bill in the legislature, SB 18, called for many of the same reforms adopted in Texas. It included enhanced good time credits for inmates who participated in rehabilitation programming, parole reform, and sentence reductions for low-level drug and property crimes. When the bill was debated in the state Senate Republican lawmakers immediately attacked it using fear-mongering rhetoric straight out of the tough-on-crime playbook. They called the bill a “get out of jail free card” and warned against “early release.” Republicans in the state Assembly warned of “blood in streets” (Yi 2009). Without Republican support, support among Democrats cratered. Assembly Speaker Karen Bass dropped several ambitious reforms just to secure a simple majority. The final bill was projected to reduce the prison population by only 10,000–a number far short of the court mandate (Percival 2015).

The Prison Realignment bill described earlier was introduced in the spring of 2011. By that time, the long running prisoner rights’ case (now named Brown vs. Plata) had reached the Supreme Court. The added pressure of an imminent Supreme Court decision and the very real possibility the Court would mandate the release of thousands of prisoners worked to the advantage of newly elected Democratic governor Jerry Brown, who cobbled together enough Democratic votes to pass the bill. Republicans once again stood in staunch opposition.

Throughout California’s drive for reform Republicans demonstrated no signs of policy learning. Their commitment to law and order remained firmly entrenched. Republican opposition to the Schwarzenegger-backed reform bill in 2009 slowed the process down and weakened and narrowed the scope of reforms. Republicans showed no appetite for studying what ailed the system or how best to fix it. This left Democrats scrambling in the eleventh hour to pass the Realignment bill. This too came with costs as county officials’ concerns about financing gaps and a variety of administrative challenges went unresolved.

Importantly, Republicans in California were not receiving the same messages that their co-partisans in Texas received. There was no equivalent to the Prison Fellowship Ministries or the Texas Public Policy Foundation. Republicans were operating in a different informational environment entirely. Interestingly the CCPOA publicly endorsed the Prison Realignment bill but the union’s endorsement was instrumentally motivated. By 2011 prison conditions were so dire that guards faced growing threats of injury or death on the job (Abramsky 2008). The union saw reform as a way to improve workplace safety. It made no effort to frame reform around core conservative principles that might have appealed to Republicans.

Lawmakers can also learn from policies in earlier adopting states, but this did not happen in California. Four years after the adoption of reforms in Texas, Republicans in California now had several successful conservative-endorsed reforms from which to learn. In addition, the TPPF’s advocacy work mutated into a “Right on Crime” campaign that sold the Texas “miracle” to conservative lawmakers and policy networks across the country. In 2011, the editorial board of the Los Angeles Times cited the Texas example and reminded state lawmakers, “Many of the progressive innovations in criminal justice are coming not from supposedly liberal states and officials, but from conservatives who are determined to focus on the cost and outcomes while keeping justice in the forefront” (Los Angeles Times 2011).

California Republicans simply refused to change their position, possibly due to electoral calculations shaped by their longtime status as the minority party. Prisons were one of the few public goods Republicans used to leverage the little power they held in legislature. For years, Republicans used tough on crime politics to attract Independent and some leaning Democratic voters fearful of crime (Simon 2007). Prisons remained popular with voters in many rural, conservative, and economically marginalized legislative districts in the Central Valley and high desert regions of the state (Thorpe 2015). California Republicans’ resistance to learning from external examples of policy success suggests new avenues for research. How and in what ways do a state’s internal political dynamics decrease the likelihood that lawmakers emulate policy ideas, even when they are endorsed by co-partisans in other states?

Discussion and Areas for Future Research

This article applied a policy learning framework to the criminal justice reform movement in the U.S. states. Policy learning helps shape Republicans’ support for reform, and it might influence the movement in ways beyond those explored in this article. Seeds (2017) argues policymakers have applied a “bifurcated” logic to reform. Policies have disproportionately benefited “low risk” or “non-violent” offenders but punitive approaches are still applied to “violent” or “repeat” offenders. Research should investigate how policy learning at the elite and mass level shapes these choices. This work might include building measures of the depth or shallowness of state reforms and their relationship to decarceration.

State governments have experienced varying degrees of success in reducing their prison population over the past decade. Thirty-two states saw their prisoner population decline. Another eighteen states experienced prisoner population increases. What drives these differences? The depth of conservative support and the strength of law enforcement organizations may shape the range of reforms considered acceptable and how they are implemented. Republicans have shifted their views on crime, but not to a degree where now anything is possible. Most states, including Texas, have resisted policies that would soften penalties for violent offenses.

Scholars working within a policy feedback framework (Pierson 1993) should investigate how the success or failure of “early” reforms shapes future reforms. Black Lives Matter, Color Of Change, and other civil rights groups have emerged because previous reforms inadequately addressed the problem of systemic racism in the criminal justice system. Progressive activists are increasingly organizing at the local level with the hope of electing more progressive prosecutors, judges, and sheriffs. How has the emergence of these new progressive groups with a renewed focus on local politics helped to modify, expand, or deepen reforms over time (see Karch and Cravens 2014)? Do these efforts risk activating a counter-reform movement intent on rolling back reformers’ hard-won gains? Investigating how these forces unfold in and across different state contexts will help clarify the promise and limitations of criminal justice reform in the U.S. states.