Donate
Independent, objective, nonpartisan research

Realignment and Recidivism in California

Database

This is the content currently stored in the post and postmeta tables.

View live version

object(Timber\Post)#3742 (44) { ["ImageClass"]=> string(12) "Timber\Image" ["PostClass"]=> string(11) "Timber\Post" ["TermClass"]=> string(11) "Timber\Term" ["object_type"]=> string(4) "post" ["custom"]=> array(5) { ["_wp_attached_file"]=> string(13) "r_1217mbr.pdf" ["wpmf_size"]=> string(6) "465432" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(61723) "DECEMBER 2017 Mia Bird, Ryken Grattet, and Viet Nguyen with research support from Justin Goss Supported with funding from the National Institute of Justice Realignment and Recidivism in California © 2017 Public Policy Institute of California PPIC is a public charity. It does not take or support positions on any ballot measures or on any local, state, or federal legislation, nor does it endorse, support, or oppose any political parties or candidates for public office. Short sections of text, not to exceed three paragraphs, may be quoted without written permission provided that full attribution is given to the source. Research publications reflect the views of the authors and do not necessarily reflect the views of our funders or of the staff, officers, advisory councils, or board of directors of the Public Policy Institute of California. SUMMARY CONTENTS Introduction The BSCC–PPIC Multi-County Study Realigned Offenders Released from Prison (PRCS) Realigned Offenders Sentenced under 1170(h) Conclusions and Policy Implications References About the Authors Acknowledgments 5 7 9 14 19 21 22 22 A technical appendix to this paper is available on the PPIC website. PPIC.ORG California has experienced significant changes in its criminal justice landscape since the 2011 implementation of public safety realignment—which shifted the management of lower-level offenders from the state prison and parole system to county jail and probation systems. The prison population has dropped dramatically, and though jail populations rose, overall incarceration levels have declined. One goal of realignment was to reduce California’s persistently high recidivism rates. Using data from 12 counties representative of the state, this report examines rearrest and reconviction rates after release from custody for two groups of offenders affected by realignment: those on post-release community supervision (PRCS) and those sentenced under section 1170(h) of the California Penal Code. Overall, we find realignment had modest effects on recidivism, with considerable variation across offender groups and counties. Specifically:  Individuals on PRCS have somewhat higher recidivism than similar individuals released before realignment. PRCS offenders are released from state prison after serving time for certain lower-level felonies and receive supervision by county probation agencies. In the two years following realignment, we find that 71.9 percent of these individuals were rearrested (2.6 percentage points higher than before realignment), and 56.4 percent were reconvicted (2.4 points higher).  Realignment did not have a consistent effect on recidivism for individuals sentenced under 1170(h). These offenders are sentenced for a specific set of lower-level felonies and, under realignment, serve time in county jail rather than state prison. In the two years following realignment, we find that 74.5 percent of these individuals were rearrested (2.3 percentage points higher than their pre-realignment counterparts) and 54.9 percent were reconvicted (2.0 points lower).  Offenders who received straight sentences have the same or lower rates of recidivism. Realignment created two types of 1170(h) offenders: those who receive both jail time and probation supervision (known as a “split” sentence) and those who receive jail time with no supervision (known as a “straight” sentence). The group serving “straight” sentences had the best outcomes: the same two-year rearrest rates and lower two-year reconviction rates (by 3.0 percentage points). Those who received “split” sentences had higher rates of rearrest (by 7.8 points) but lower rates of reconviction (by 3.4 points) compared with similar individuals before realignment.  The effects of realignment on recidivism vary substantially across counties. For example, overall we find reconviction rates were higher for those on PRCS after realignment, but in fact nine counties saw lower Realignment and Recidivism in California 3 rates of reconviction—indicating that the overall finding is driven by a small number of counties. County variation in recidivism outcomes is likely linked to demographic, economic, and geographic differences, as well as the range of county capacity and experiences providing evidence-based interventions before realignment. However, some of this variation may be due to different intervention strategies, creating the potential for counties to learn from each other over time. Notably, offenders who received a jail term and no supervision stand out as having better outcomes on all measures of recidivism, when compared with similar individuals released before realignment. This finding suggests we need to carefully consider the complex relationship between supervision and recidivism. While it could simply be easier to detect reoffending when an individual is under supervision, the requirements of supervision could also create more opportunities for non-criminal violations. With a longer follow-up window and more recent data, the relationship between supervision and recidivism, as well as the overall effects of realignment, may change as counties build capacity and experience with evidence-based practices. Policymakers, practitioners, and researchers must continue to work together to develop the data and expertise necessary to understand the impacts of California’s corrections reforms and to identify effective strategies to reduce recidivism. PPIC.ORG Realignment and Recidivism in California 4 Introduction In 2011, California passed public safety realignment, one of the most far-reaching criminal justice reforms in recent US history.1 Realignment has been referred to as “revolutionary and sudden” (Weisberg 2011), “the most significant correctional reform in decades” (Misczynski 2012), and “the biggest penal experiment in modern history” (Santos 2013). This law was motivated by a US Supreme Court decision, which found that California’s prison system could not adequately meet inmates’ health care needs due to extreme overcrowding. The court mandated that the state either dramatically increase prison capacity or reduce the inmate population by tens of thousands. At the time, California faced a severe budget crisis, with limited ability to build new prisons or contract out beds and services to other entities. Under these constraints, the state elected to shift (or “realign”) the management of lower-level felony offenders from the state prison and parole system to county jail and probation systems. Realignment created two new populations of offenders:  Individuals on post-release community supervision (PRCS). The PRCS population refers to individuals released from prison after serving terms for non-serious and non-violent felony offenses. Under realignment, individuals on PRCS are supervised by county probation departments rather than the state parole system. Prior to realignment, people released from prison after serving time for these offenses would have been revoked to prison custody for supervision violations—which contributed to prison overcrowding. Now, the PRCS population serves any revocations in county jails.  Individuals sentenced under 1170(h). The 1170(h) population refers to non-serious, non-violent, nonsexual felony offenders who serve their sentences in county jail. Prior to realignment, individuals convicted for specific offenses defined in §1170(h) of the California Penal Code would have been sent to state prison if their custody sentences exceeded one year. Like the PRCS population, 1170(h) offenders are revoked to local jail custody, not prison custody, for supervision violations. Realignment further created two types of 1170(h) offenders: those who receive a “straight” jail sentence for the full term and no probation following release, and those who have their sentence “split” between a jail term and probation. Realignment effectively reduced the state’s prison population by more than 27,000 in the first year of implementation (Lofstrom and Martin 2015). As expected, jail populations increased and, in many counties, jails reached or exceeded capacity (Martin and Lofstrom 2014). However, the total statewide jail population increased by fewer than 9,000 inmates, or about one-third of the decline in the prison population. As a result, realignment led to a reduction in both the prison population and overall incarceration levels. The reform prompted considerable debate over the relationship between incarceration and public safety. Proponents of realignment anticipated the reduction in incarceration and argued that pre-realignment incarceration levels had been too high. Some supporters claimed that increased corrections spending on incarceration was not a cost-effective way to improve public safety, and others maintained that the pre-realignment system created perverse incentives for counties to shift the burden of preserving public safety to the state, resulting in overly punitive sanctions against lower-level offenders. However, opponents of realignment had serious concerns that an increase in “street time” for former offenders would lead to rising crime rates. Critics were also concerned that the perceived reduction in penalties would lead to increased reoffending. To date, research examining the effects of realignment on crime has found no increase in violent crime under realignment and small increases in property 1 Public safety realignment was initiated through the passage of Assembly Bill 109 in 2011. Subsequent legislation has further defined the details of realignment implementation. PPIC.ORG Realignment and Recidivism in California 5 crime, driven specifically by an increase in motor vehicle theft (Lofstrom and Raphael 2013; Lofstrom and Raphael 2015). For proponents of realignment, another goal was to reduce the state’s persistently high rates of recidivism, which are among the highest in the nation. Supporters put forth the notion that “locals can do it better,” referring to a belief that local justice systems are better positioned than the state prison and parole system to provide rehabilitative interventions shown by research to be effective in reducing recidivism. While local justice systems supported this idea in principle, they also voiced concerns about their abilities to ramp up capacity in time and with the level of funding that was provided—particularly with respect to the use of evidence-based interventions. Findings on the effects of realignment on recidivism are still emerging from the state and research community— in part because evaluating the impact of a policy change on recidivism rates requires a substantial follow-up period after individuals are released from custody. In October 2017, the California Department of Corrections and Rehabilitation (CDCR) published reports on the recidivism outcomes of individuals released from prison, including the PRCS population, after realignment (CDCR 2017a, 2017b). As intended, returns to prison custody have declined dramatically. CDCR describes an increase in reconviction for offenders released from prison during the first fiscal year after realignment and a decrease for offenders released in the second year, relative to the reconviction rates of individuals released in the two years prior to realignment.2 These are descriptive findings, and as CDCR notes, the population of individuals released from prison is changing in composition over time. As realignment rolls out, this group is increasingly composed of serious and violent offenders released following sentences imposed for new convictions (rather than for revocations).3 Previous research generally finds that recidivism rates have not yet improved for individuals released from prison under realignment (Lofstrom, Raphael, and Grattet 2014; Lofstrom, Bird, and Martin 2016). Those released on PRCS have been found to have higher rates of recidivism relative to their pre-realignment counterparts (Bird and Grattet 2015; Lofstrom, Bird, and Martin 2016). However, recidivism rates vary in important ways across counties, and some evidence suggests this variation may be linked to county capacity and the propensity to use rehabilitative programs and services with the PRCS population (Bird and Grattet 2015; Bird and Grattet 2016). Thus far, research on the effects of realignment on recidivism has been limited in two important ways. First, prior studies have not been able to examine recidivism among the 1170(h) population.4 In contrast to the PRCS population, which spiked in the first year of realignment and leveled out, over time, individuals sentenced under 1170(h) will make up a larger portion of the realigned population. Second, existing research has not captured returns to jail custody—a key measure of recidivism—for those on parole, those on PRCS, or those sentenced under 1170(h). This report addresses those gaps. We examine recidivism outcomes for the PRCS and 1170(h) populations for a representative group of 12 California counties. Our analysis focuses on the impact of realignment on one- and two-year recidivism rates for individuals released from prison or jail between October 2011 and September 2013—relative to those released from prison custody before realignment, from October 2009 to September 2011. We also present differences in one-year recidivism rates before and after realignment for each of the 12 counties. 2 For consistency with the state definition of recidivism (“conviction of a new felony or misdemeanor committed within three years of release from custody”), CDCR uses reconviction as the primary measure of recidivism. Section 3027 of the California Penal Code required the Board of State and Community Corrections to develop a statewide definition of recidivism. 3 CDCR has shared data with researchers to allow for recidivism analyses that attempt to adjust for the changing composition of the population released from prison, as well as to focus on outcomes for the PRCS population. 4 There is no statewide data source capturing the 1170(h) population and their recidivism outcomes. Nor is there a statewide data source capturing returns to jail custody for the parole, PRCS, and 1170(h) populations. PPIC.ORG Realignment and Recidivism in California 6 The BSCC–PPIC Multi-County Study Our analysis relies primarily on data from the BSCC–PPIC Multi-County Study (MCS), a collaborative effort between the California Board of State and Community Corrections (BSCC) and PPIC.5 The MCS was established in the wake of public safety realignment with the goal of bringing together the data needed to rigorously evaluate the statewide effects of this policy reform and to identify the most effective recidivism-reduction interventions at the local level. To achieve these goals, we identified a group of counties representative of the state as a whole and partnered with these counties to examine how individuals move through local jail and probation systems after realignment. Figure 1 shows the 12 counties participating in the study: Alameda, Contra Costa, Fresno, Humboldt, Kern, Los Angeles, Orange, Sacramento, San Bernardino, San Francisco, Shasta, and Stanislaus. FIGURE 1 The 12 counties participating in the MCS represent California’s geographic diversity SOURCE: BSCC–PPIC Multi-County Study. NOTE: These were the only 12 counties approached by the MCS team. No counties were approached and declined to participate. Taken together, these counties comprise 60 percent of California’s population and represent the state’s geographic diversity, as well as its overall demographic and economic characteristics. Table 1 summarizes the characteristics of the MCS counties relative to the statewide population. While quite similar, the MCS counties tend to be more 5 PPIC research publications reflect the views of the authors and do not necessarily reflect the views of the BSCC, CDCR, the Department of Justice, or the participating counties. PPIC.ORG Realignment and Recidivism in California 7 urban (as measured by population density) and have higher shares of African Americans, Asian Americans, and Latinos. In addition, the poverty and unemployment rates are slightly higher among the MCS counties. TABLE 1 MCS counties are similar to the state overall in demographic and economic characteristics California MCS counties Demographic characteristics Male 49.7% 49.4% African American 6.5% 8.0% Asian American 14.8% 16.2% Latino 38.6% 40.6% Native American 1.7% 1.5% White 73.3% 70.8% Two or more 3.7% 3.6% Under 20 26.4% 26.2% Age 20–39 28.9% 29.5% Age 40–59 26.5% 26.8% Age 60+ 18.2% 17.5% Population density (population per square mile) 244.6 454.7 Economic characteristics Unemployment rate 9.1% 9.2% Poverty rate 16.5% 17.6% Total population 38,335,203 22,847,093 SOURCES: Demographic and population density characteristics are from the US Census. Poverty rates are from the Small Area Income and Poverty Estimates (SAIPE) program within the Census Bureau. Unemployment rates are from the Bureau of Labor Statistics. NOTES: Characteristics for the MCS county group are population-weighted for the year 2013. In addition to the data provided by the counties, the California Department of Justice (DOJ) and the CDCR provided essential data to fill out the state-local picture. Altogether, the newly available data used in this analysis includes demographic characteristics, offender background and criminal history, and recidivism outcomes. The participation of the MCS counties allows us to expand on previous research by assessing outcomes for individuals sentenced to serve time in county correctional agencies, who, because they pass through local systems, are not tracked at the state level. We are also able to examine returns to jail custody within the MCS counties. Finally, sentencing information from these counties allows us to explore the relationship between split sentencing—a new tool under realignment—and recidivism for the first time. PPIC.ORG Realignment and Recidivism in California 8 Measuring Recidivism There are many ways of measuring recidivism.6 The state has adopted reconviction—often considered a more accurate measure of reoffending behavior, relative to rearrest—as its primary recidivism measure. Reconvictions also represent a substantial resource burden and thus may be of particular interest to correctional systems. However, when comparing the reconviction rates of individuals released before and after realignment, it is important to consider how realignment changed the likelihood that criminal justice systems would pursue formal convictions. Before realignment, individuals who violated the conditions of their parole or who were suspected of a new offense could be revoked to prison custody. Under realignment, most offenders can only be sent to prison following conviction on a new, prison-eligible offense—meaning that correctional systems may be shifting away from revocations and toward more formal rearrests and reconvictions. Therefore, if we were to see increases in reconviction rates after realignment, we should not necessarily assume these differences were driven by an increase in reoffending behavior. However, realignment may have also created new incentives for individual offenders. Research on recidivism generally aims to identify how changes in policy or practice affect underlying criminal behavior, but it is often difficult to determine whether changes in recidivism measures reflect changes in criminal behavior or, alternatively, changes in criminal justice processes. In this report, we account for potential changes in criminal justice processes by constructing a measure of reconviction that includes both formal reconvictions and revocations. We use a similar measure for rearrests that includes both formal rearrests and any revocations not preceded by a formal arrest. Using these two measures, we estimate the effects of realignment on recidivism for individuals on PRCS and individuals sentenced under 1170(h).7 Realigned Offenders Released from Prison (PRCS) As defined in realignment law, post-release community supervision refers to a group of offenders who are released from state prison after having been committed for a non-serious, non-violent felony offense.8 The law disqualifies individuals from PRCS who are determined through a CDCR assessment process to be “high-risk sex offenders” or “mentally disordered offenders” (Couzens and Bigelow 2016). Importantly, PRCS does not affect the length of time a person serves in custody, but it does change where one is supervised upon release. People on PRCS are supervised in the community by county probation departments. Prior to realignment, they were supervised by the state parole system.9 In the case of a supervision violation, individuals on PRCS are revoked to jail instead of prison custody. In addition to the change in which agency supervises them, realignment altered other conditions of their supervision as well, including their eligibility for early discharge and new limits on revocations.10 6 Indeed, it is important to note that the best measures of recidivism depend on the purpose of the analysis. For example, if the primary goal is to understand the effect of a policy change on the cost of operating correctional institutions, then the most appropriate metric to examine may be returns to custody. If the goal is to understand how a policy change has affected the demands placed on police departments, then the best recidivism measure may be rearrest rates. No single measure of recidivism fits all purposes. 7 In the Technical Appendix, we also present traditional rearrest and reconviction measures without adjustments for revocations. Note that the 1170(h) population, given their local sentence length, is limited to those likely to have gone to prison prior to realignment, as explained in more detail in the Technical Appendix. 8 PRCS offenders may have prior serious or violent offenses in their criminal history. 9 State parole and county probation have many similarities in terms of the types of supervision practices, legal authority, and surveillance tools available. However, there are some important differences as well. A large portion of parole officers spent at least some portion of their careers working in prison as correctional guards. Probation officers are less likely to have such experiences. Parole officers carry firearms, while probation officers typically do not. Probation is subject to variation by county, given that it is a county government agency, dependent on resources provided by the County Boards of Supervisors. Parole is more centralized and uniformly administered across the state. 10 If individuals on PRCS have no violations that result in a custodial sanction, they are eligible for early discharge from supervision after six months. Prior to realignment, eligibility for discharge from parole commenced at 12 months. In addition, realignment imposed limits on the amount of time an individual under PPIC.ORG Realignment and Recidivism in California 9 Characteristics During the first two years under realignment, more than 35,000 individuals were released on PRCS to MCS counties.11 To estimate the effects of realignment on recidivism, we compare the outcomes of those on PRCS after realignment to those of similar individuals who were released from prison in the two years before realignment (between October 2009 and September 2011). A key concern when comparing outcomes of those on PRCS to those of the pre-realignment population is that there are considerable differences between these two groups. Demographically, the pre-realignment population—which includes all individuals released from state prison in the two years preceding realignment—is younger, less likely to be white, more likely to be Latino, and less likely to face mental health challenges than those on PRCS (Table 2). Overall, more than one-quarter of the pre-realignment population is incarcerated for a crime against a person, compared with less than one-sixth of the PRCS population. The pre-realignment population is also more likely to be incarcerated on a conviction for a new offense (rather than a revocation). Those released from prison before realignment also have lower numbers of past arrests and convictions when compared with individuals on PRCS. These differences in characteristics suggest the pre-realignment population includes more serious offenders, but these offenders may have a lower likelihood of recidivism than the PRCS population. We use a matching method to refine the pre-realignment population to a control group that is more similar to the PRCS population. This approach allows us to better isolate the effects of realignment when examining differences in recidivism.12 The demographic characteristics and criminal histories of the pre-realignment control group and the PRCS population are quite similar. However, those on PRCS are still less likely to have been incarcerated for a new conviction and, instead, are more likely to be released after serving a revocation. Our analysis below on recidivism outcomes uses a regression model to adjust for these remaining differences between the prerealignment control group and those on PRCS. TABLE 2 The PRCS population differs considerably from the pre-realignment population of those released from prison Demographic characteristics Age Male African American Asian American Latino Native American White Other race Pre-realignment population 36.2 88.9% 28.4% 1.0% 43.5% 0.6% 23.3% 3.3% Pre-realignment control group PRCS population 37.1 89.3% 28.8% 0.9% 40.0% 0.7% 26.5% 3.2% 37.5 89.9% 29.0% 0.9% 39.8% 0.7% 26.5% 3.1% supervision could be revoked to custody for violations, to a maximum of six months. It also changed the location where those revocations are to be served, from state prison to county jail. 11 Here we refer to individuals. The total number of PRCS releases would be higher than the total number of individuals released over the period. This report focuses on recidivism outcomes for this subset of the statewide PRCS population because the MCS dataset provides information on revocations to jail custody, which is not available for the full, statewide PRCS population. 12 We use a propensity score matching method to select individuals from the larger pre-realignment population that are similar to the PRCS population. This method is described in more detail in the Technical Appendix. PPIC.ORG Realignment and Recidivism in California 10 Pre-realignment population Pre-realignment control group PRCS population Offender background and criminal history Age at first conviction 23.3 22.6 22.4 History of mental illness 15.9% 18.5% 19.0% Length of stay in prison (in days) 636.9 488.4 504.4 New conviction 67.8% 59.6% 56.2% Parole violator with new conviction 19.1% 30.3% 34.5% Parole violator 11.9% 9.3% 9.3% Number of past arrests 15.1 18.1 18.8 Number of past felony arrests 9.4 11.2 11.6 Number of past convictions 5.9 7.1 7.4 Number of past felony convictions 3.2 3.9 4.2 Number of prior serious convictions 0.2 0.3 0.4 Number of prior violent convictions 0.1 0.1 0.1 High-risk for a person offense 11.1% 10.7% 11.5% High-risk for a property offense 9.2% 11.9% 11.6% High-risk for a drug offense 5.0% 7.4% 7.7% Moderate risk for reoffending 21.2% 23.1% 23.0% Low risk for reoffending 16.5% 15.8% 15.1% Second-striker 13.4% 17.1% 18.6% Offense type Murder 0.9% 0.0% 0.0% Manslaughter 0.9% 0.0% 0.1% Robbery 7.1% 0.0% 0.0% Other person offense 14.1% 14.1% 13.7% Rape 0.4% 0.0% 0.0% Molestation 2.3% 0.2% 0.2% Other sex offense 2.2% 2.0% 1.9% Burglary 11.3% 10.7% 10.2% Grand theft 3.6% 4.3% 4.0% Petty theft 4.2% 5.0% 5.1% Vehicle theft 4.4% 6.0% 6.2% Other property offense 7.4% 9.0% 8.9% Controlled substance sales, possession, manufacturing 26.4% 32.2% 32.7% Marijuana substance sales, possession, manufacturing 2.2% 2.1% 2.0% DUI 3.1% 2.8% 2.7% Weapon possession 5.9% 9.2% 10.3% Other offense 3.6% 2.4% 2.3% Number of observations 84,577 26,261 35,218 SOURCES: BSCC–PPIC Multi-County Study data. NOTES: We limit the analysis to first releases in the pre- and post-realignment periods. Therefore, these counts will not correspond perfectly with the total number of prison releases, as some individuals are incarcerated for multiple prison spells in a given time period. PPIC.ORG Realignment and Recidivism in California 11 Recidivism Outcomes Our findings suggest that realignment increased recidivism by a small amount for people on PRCS. This increase could be driven by a rise in offending behavior related to a perceived reduction in sanctions under realignment. It is also possible that the PRCS population was monitored more closely while under probation supervision than their pre-realignment counterparts under parole supervision. If this were the case, the higher rates of recidivism may be due to a greater likelihood of detecting misconduct under realignment or because the conditions of supervision created more opportunities for violations that led to revocations. Conditions of supervision might include drug testing, frequent reporting, required programs, or required meetings with probation officers, among others. We focus on two key measures of recidivism: rearrest and reconviction outcomes after release. As described earlier, we use a measure of rearrest that includes any revocations that are not preceded by a formal arrest, and a measure of reconviction that also includes revocations.13 After adjusting for differences between these groups, we estimate that those on PRCS have a one-year rearrest rate of 61.7 percent, or a notable 5.2 percentage points higher than their pre-realignment counterparts (Figure 2). When we examine two-year rearrest rates, however, the difference shrinks considerably. The two-year rearrest rate for individuals on PRCS is 71.9 percent, only 2.6 percentage points higher than that of similar individuals prior to realignment. Those on PRCS also have slightly higher reconviction rates when compared with similar individuals released before realignment: 40.1 percent (1.8 points higher than before realignment) of individuals on PRCS are reconvicted within one year of release and 56.4 percent (2.4 points higher) are reconvicted within two years. Although we have used regression models to adjust for differences in the characteristics available in the data, it is possible that people on PRCS differ from the pre-realignment control group in ways that we cannot account for and that would lead to higher rates of recidivism. It is also possible that the outcomes of realigned offenders may improve over time as counties become more experienced in managing the PRCS population and better able to provide evidence-based, recidivism-reduction interventions. Indeed, the findings presented may reflect some improvement over time, as the difference in the two-year rearrest rate is much lower than the difference in the one-year rearrest rate. Share of population (%) FIGURE 2 Individuals on PRCS have slightly higher recidivism rates than similar individuals before realignment 80 61.7 60 56.5 69.3 71.9 54.0 56.4 Pre-realignment PRCS 38.3 40.1 40 20 0 One-year rearrest Two-year rearrest One-year reconviction Two-year reconviction SOURCE: BSCC–PPIC Multi-County Study data. NOTE: Chart shows regression-adjusted rates, which should be interpreted as causal findings. All differences in recidivism rates between the pre-realignment control group and the PRCS population are statistically significant. Please see the Technical Appendix for more details. 13 The first- and second-year release cohorts are included in the one-year rearrest rate, but only the first-year release cohort is included in the two-year rearrest rate due to limitations in the date range over which we have data. PPIC.ORG Realignment and Recidivism in California 12 County Variation In California, there is considerable variation in recidivism outcomes across counties both before and after realignment. Across the state, counties have vastly different demographic compositions, economic resources, and geographic environments. For these reasons, counties should not be directly compared to each other or judged to be performing better or worse under realignment on the basis of this analysis. However, it will be important for counties to assess their own outcomes over time as they make efforts to improve. It is important to note that, while the recidivism rates presented below account for changes in the composition of county populations over time, they do not adjust for differences in population or other characteristics across counties. Therefore, these estimates should be compared within a given county over time, but not across counties. As shown in Figure 3, in some counties, such as Contra Costa, Humboldt, and Kern, those on PRCS have oneyear rearrest rates that are lower than the rates of their pre-realignment counterparts.14 In addition, in San Francisco County, we find similar rearrest rates for the PRCS population and the pre-realignment population. Overall, however, those on PRCS in most of the MCS counties have higher rearrest rates compared with similar individuals released to parole before realignment. FIGURE 3 In three out of twelve counties, those on PRCS have lower rearrest rates than similar individuals before realignment Alameda Contra Costa Fresno Humboldt Kern Los Angeles Orange Sacramento San Bernardino San Francisco Shasta Stanislaus Pre-realignment PRCS 0 20 40 60 80 One-year rearrest rates (%) SOURCE: BSCC–PPIC Multi-County Study data. NOTE: These are regression-adjusted recidivism rates. This analysis draws on small sample sizes in some counties and, therefore, we caution against causal interpretations. In addition, regression adjustments are made for changes in the composition of county populations over time, within county. Adjustments are not made for differences in population composition across counties and, therefore, these estimates should be compared within county over time, but not across counties. Figure 4 presents one-year reconviction rates for the 12 MCS counties. In nine of the twelve counties, individuals on PRCS have lower reconviction rates than similar individuals before realignment. This demonstrates that the overall finding of higher reconviction rates for the PRCS population (see Figure 2 above) is driven by higher rates in a few counties. It is important to note that about one-third of individuals on PRCS statewide and nearly one- 14 Our county-level analysis uses one-year rearrest and reconviction rates because two-year rates are based only on the first-year release cohort and the size of the 1170(h) populations (i.e., the treatment groups) are quite small for some counties. Though the PRCS populations (i.e., the treatment groups) are larger, we use the same approach for both groups of offenders for consistency. PPIC.ORG Realignment and Recidivism in California 13 half of individuals on PRCS in the MCS counties were released to Los Angeles County. Because Los Angeles is such a large county, it is a driving factor in the overall recidivism outcomes of the MCS group.15 FIGURE 4 In nine out of twelve counties, those on PRCS have lower reconviction rates than similar individuals before realignment Alameda Contra Costa Fresno Humboldt Kern Los Angeles Orange Sacramento San Bernardino San Francisco Shasta Stanislaus Pre-realignment PRCS 0 10 20 30 40 50 60 One-year reconviction rates (%) SOURCE: BSCC–PPIC Multi-County Study data. NOTE: These are regression-adjusted recidivism rates. This analysis draws on small sample sizes in some counties and, therefore, we caution against causal interpretations. In addition, regression adjustments are made for observable changes in the composition of county populations over time, within county. Adjustments are not made for differences in population composition across counties and, therefore, these estimates should be compared within county over time, but not across counties. Realigned Offenders Sentenced under 1170(h) Realignment law also defined a group of felony offenders under California Penal Code §1170(h) who now are ineligible for state prison and must serve their sentences in county correctional systems. In general, the 1170(h) population is made up of non-serious, non-violent, and non-sexual offenders who are sentenced to county jail under realignment.16 Those offenders who have a serious or violent offense in their past conviction history in California or another state are ineligible to be sentenced under 1170(h). Registered sex offenders are also excluded.17 Judges may sentence qualifying individuals to a straight sentence or a split sentence.18 Those who receive a straight sentence serve the entirety of their term in jail and, upon release, have no further supervision associated with the sentence. In contrast, those who receive a split sentence serve a portion of their term in jail and a portion under supervision by county probation. During this period of supervision, these individuals may be revoked to jail 15 Although overall reconviction rates for the PRCS population among MCS counties are higher than for similar individuals before realignment, when Los Angeles County is removed from the analysis we find the one-year reconviction rate for the PRCS population is 5.7 percentage points lower and the two-year reconviction rate is 2.2 percentage points lower than for those released before realignment. Los Angeles County is not alone among the MCS counties in having higher reconviction rates for those on PRCS, but it does have a disproportionate effect on the overall results because of its size. Please see the Technical Appendix for more details. 16 This includes over 1,200 separate sections of the California Penal Code. 17 In addition, offenses included in over 70 separate sections of the California Penal Code were deemed ineligible by the legislature, even though these offenses are not serious, violent, or sexual in nature. 18 For individuals sentenced after January 1, 2015, split sentencing became the presumptive sentence under California Penal Code §1170(h). PPIC.ORG Realignment and Recidivism in California 14 to serve out the remainder of their sentence. Realignment law did not change the length of sentences, so a split sentence is the same amount of time as a straight sentence for the same conviction offense.19 Characteristics The 1170(h) population includes nearly 13,500 individuals sentenced under 1170(h) and released from MCS county jails during the first two years of realignment.20 In contrast to the number of individuals on PRCS, which reached a peak in the first year of realignment and then leveled out, the number of people sentenced under 1170(h) more than doubled in the second year of realignment. This is to be expected, as 1170(h) convictions began only after the implementation of realignment, and those sentenced under 1170(h) must first complete their jail terms before they enter a release cohort. In addition, as realignment implementation rolls out over time, fewer individuals sent to prison will be eligible for PRCS, as non-serious, non-violent, non-sexual offenders are increasingly serving sentences in county correctional systems. Table 3 summarizes the characteristics of those released after serving 1170(h) sentences. As was the case with the PRCS population, the full pre-realignment population is demographically different from the 1170(h) population. In this case, the pre-realignment population is older, more likely to be male, and more likely to be African American than those sentenced under 1170(h). In terms of criminal history, the pre-realignment population has a higher number of prior serious convictions and a greater share are incarcerated for a crime against a person. But those sentenced under 1170(h) have higher rates of past arrests and convictions. These differences suggest that while offenders in the pre-realignment population may have more serious criminal histories, they may also be less frequent offenders than the 1170(h) population. As a result, we might expect the pre-realignment population to have a lower overall likelihood of recidivism. We again create a matched pre-realignment control group of similar individuals to serve as a better comparison for those sentenced under 1170(h). The characteristics of this comparison group are much more similar to those of the 1170(h) population. However, the pre-realignment control group is still slightly more likely to be African American and has slightly lower rates of past arrest and conviction. Therefore, below we use regression analysis, which adjusts for these remaining differences, to estimate the effects of realignment on recidivism for the 1170(h) population. TABLE 3 Individuals sentenced under 1170(h) differ in notable ways from those who were released from prison before realignment Demographics Age Pre-realignment population 36.2 Pre-realignment control group 34.8 1170(h) population 34.6 Male 88.9% 84.2% 84.1% African American Asian American Native American Latino White 28.4% 1.0% 0.6% 43.5% 23.3% 23.5% 1.4% 0.3% 43.6% 29.0% 22.4% 1.8% 0.2% 43.8% 29.7% 19 Moreover, in most cases a judge may elect to sentence an individual convicted for an 1170(h) offense to felony probation in lieu of a straight or split sentence. This option existed prior to realignment and remains an option today for both 1170(h)-eligible and prison-eligible offenses. 20 The Technical Appendix provides details on the construction of the MCS 1170(h) population. As explained therein, this population will not match total counts of 1170(h) convictions. PPIC.ORG Realignment and Recidivism in California 15 Pre-realignment population Pre-realignment control group 1170(h) population Other race 3.3% 2.3% 2.1% Offender background and history Age at first conviction 23.3 23.2 22.9 Length of stay in jail or prison (in days) 636.9 243.2 233.2 Number of past arrests 15.1 18.3 19.1 Number of past felony arrests 9.4 11.0 11.3 Number of past convictions 5.9 7.0 7.3 Number of past felony convictions 3.2 3.8 4.0 Number of prior serious convictions 0.2 0.1 0.1 Number of prior violent convictions 0.1 0.1 0.1 Offense type Crime against persons 27.8% 2.8% 2.6% Burglary 11.3% 16.5% 16.0% Grand theft 3.6% 5.0% 4.7% Petty theft 4.2% 4.6% 4.3% Vehicle theft 4.4% 7.9% 8.7% Other property offense 7.4% 15.3% 16.3% Controlled substance sales, possession, manufacturing 26.4% 35.4% 36.2% Marijuana substance sales, possession, manufacturing 2.2% 2.8% 2.7% DUI 3.1% 3.6% 3.3% Weapon possession 5.9% 3.6% 3.0% Other offense 3.6% 2.6% 2.3% Number of observations 84,577 10,869 13,435 SOURCES: BSCC–PPIC Multi-County Study data. NOTES: We limit the analysis to first releases in the pre- and post-realignment periods. Therefore, these counts will not correspond perfectly with the number of prison releases in any period or the counts of 1170(h) convictions, as some individuals are incarcerated for multiple prison or jail spells. In addition, for the analysis of the effects of realignment on recidivism, we further limit the 1170(h) population to those likely to have been sent to prison if they had been convicted prior to realignment. This limitation is described in more detail in the Technical Appendix. Recidivism Outcomes Overall, we find that realignment did not have a consistent effect on recidivism among the 1170(h) population. After adjusting for differences in the characteristics of the pre-realignment control group and the 1170(h) population, we find no differences in one-year rearrest rates (Figure 5). Those sentenced under 1170(h) have a two-year rearrest rate of 74.5 percent, slightly higher (by 2.3 percentage points) than that of similar individuals before realignment.21 However, when we examine reconviction rates, the 1170(h) population has a one-year reconviction rate of 39.1 percent, slightly lower (by 3.8 points) than that of the pre-realignment control group. We also estimate that those sentenced under 1170(h) have a two-year reconviction rate of 54.9 percent, 2.0 percentage points lower than that of their pre-realignment counterparts. 21 One-year recidivism rates are based on the first- and second-year release cohorts under realignment. Two-year rates are based only on the first-year release cohort due to limitations in the follow-up time period for which we have data. PPIC.ORG Realignment and Recidivism in California 16 These findings suggest the 1170(h) population had better outcomes under realignment—at least as measured by reconviction rates—than the PRCS population. However, the 1170(h) population includes both those with straight jail sentences and split sentences (a jail term followed by probation). When we conduct a separate analysis for these two groups, we find that those who received split sentences have higher one-year rearrest rates (by 7.2 percentage points) but lower one-year reconviction rates (by 5.4 percentage points) when compared to similar individuals released before realignment. The two-year rearrest rates for those who received split sentences are higher than those of the pre-realignment group (by 7.8 percentage points) and the two-year reconviction rates are lower than those of the pre-realignment control group (by 3.4 percentage points). In contrast, we find that those who received straight sentences have lower one-year rearrest rates (by 1.4 percentage points) and lower oneyear reconviction rates (by 4.1 percentage points) when compared to their pre-realignment counterparts. The twoyear rearrest rates for those given straight sentences are about the same as for their pre-realignment counterparts. However, the two-year reconviction rates are 3.0 percentage points lower than those of the pre-realignment comparison group. Therefore, lower rearrest rates among the full 1170(h) population are driven by lower rates for those receiving straight sentences. (Additional details are in Table A4 in the Technical Appendix.) FIGURE 5 The effects of realignment on recidivism vary among those sentenced under 1170(h) 80 72.2 74.5 61.6 61.7 60 56.9 54.9 42.9 39.1 40 Pre-realignment 1170(h) Share of population (%) 20 0 One-year rearrest Two-year rearrest One-year reconviction Two-year reconviction SOURCE: BSCC–PPIC Multi-County Study data. NOTE: Chart shows regression-adjusted rates, which should be interpreted as causal findings. The small difference in the one-year rearrest rate is not statistically significant. All other differences in recidivism rates between the pre-realignment control group and the PRCS population are statistically significant. Please see the Technical Appendix for more details. County Variation As is the case for individuals on PRCS, recidivism rates vary considerably across counties for those released after serving 1170(h) sentences. Here we see that more than half of the MCS counties have lower one-year rearrest rates among those released for 1170(h) sentences, when compared to a similar group of offenders released prior to realignment (Figure 6). However, as shown above (see Figure 5), we find no evidence of an overall difference in PPIC.ORG Realignment and Recidivism in California 17 one-year rearrest rates for the 1170(h) population after realignment—suggesting higher rearrest rates in a minority of counties outweigh lower rearrest rates in the majority of counties.22 FIGURE 6 Over half of counties have lower rearrest rates among 1170(h) offenders than similar individuals before realignment Alameda Contra Costa Fresno Humboldt Kern Los Angeles Orange Sacramento San Bernardino San Francisco Shasta Stanislaus 0 Pre-realignment 1170(h) 10 20 30 40 50 60 70 80 90 One-year rearrest rates (%) SOURCE: BSCC–PPIC Multi-County Study data. NOTE: These are regression-adjusted recidivism rates. This analysis draws on very small sample sizes in some counties and, therefore, we caution against causal interpretations. In addition, regression adjustments are made for changes in the composition of county populations over time, within county. Adjustments are not made for differences in population composition across counties and, therefore, these estimates should be compared within county over time, but not across counties. In the case of reconviction, more than half of the MCS counties have lower one-year reconviction rates, consistent with the overall findings (Figure 7). The variation across counties in the recidivism rates before and after realignment suggest there is a great deal to learn from the various county experiences under realignment. At the same time, the variation across counties in the pre-realignment period suggests that counties had different starting points and have considerably different populations to manage. 22 For example, when we remove Los Angeles from the analysis, we find that the one-year rearrest rate among 1170(h) offenders is 1.4 percentage points lower and the two-year rearrest rate is 0.7 percentage points lower, compared with those released before realignment. Please see the Technical Appendix for more details. PPIC.ORG Realignment and Recidivism in California 18 FIGURE 7 Over half of counties have lower reconviction rates among 1170(h) offenders than similar individuals before realignment Alameda Contra Costa Fresno Humboldt Kern Los Angeles Orange Pre-realignment 1170(h) Sacramento San Bernardino San Francisco Shasta Stanislaus 0 10 20 30 40 50 60 70 One-year reconviction rates (%) SOURCE: BSCC–PPIC Multi-County Study data. NOTE: These are regression-adjusted recidivism rates. This analysis draws on very small sample sizes in some counties and, therefore, we caution against causal interpretations. In addition, regression adjustments are made for changes in the composition of county populations over time, within county. Adjustments are not made for differences in population composition across counties and, therefore, these estimates should be compared within county over time, but not across counties. Conclusions and Policy Implications Public safety realignment marked a new era for corrections and rehabilitation—and raised questions about the reform’s effects on crime and recidivism. Our analysis examines recidivism among the two realigned groups of offenders using data from the BSCC–PPIC Multi-County Study, an integrated state-local collaboration with the Board of State and Community Corrections and 12 counties representative of the state. We find that individuals released under post-release community supervision during the first two years of realignment have slightly higher rates of rearrest and reconviction than their pre-realignment counterparts. Recidivism among individuals sentenced under 1170(h) is more mixed. The 1170(h) population has a higher two-year rearrest rate than their pre-realignment counterparts, but lower one- and two-year reconviction rates. Notably, among 1170(h) offenders, we find higher rearrest rates for those who received probation supervision following their jail term (known as a split sentence), but lower reconviction rates. Those who served a jail term with no probation after release (known as a straight sentence) have the same or lower recidivism rates when compared with their pre-realignment counterparts—the only group for which there is evidence of consistently better outcomes after realignment. Realignment has created a unique opportunity for counties to learn from each other and for the state to leverage the diversity of practices to identify the most effective recidivism-reduction strategies. We find substantial variation in both overall recidivism rates across counties and changes in those rates within counties over the PPIC.ORG Realignment and Recidivism in California 19 realignment period. While some of this variation is likely linked to differences in the demographic, economic, and geographic characteristics of these counties, it may also be explained by differences in county capacity and experience with evidence-based interventions before realignment. However, if some of this variation is due to different intervention strategies, counties could potentially improve their outcomes by learning from each other over time. Taken together, these findings suggest three key takeaways. First, the effects of realignment appear to vary across offender groups and counties. Second, as more data becomes available we may observe the effects of realignment changing over time. The majority of the realigned population will shift from PRCS offenders to 1170(h) offenders, and in the coming years, counties will continue to build capacity and experience with evidence-based and datadriven practices. And third, we need to carefully consider the relationship between supervision and higher recidivism rates. Our findings indicate higher rates of recidivism for the groups supervised following release: those on PRCS and those who received split sentences under 1170(h). While it may simply be easier to detect reoffending when an individual is under supervision, it is also possible that supervision conditions create more opportunities for non-criminal violations. Additional research is needed to further investigate the likely complex relationship between supervision and recidivism. The next steps in the Multi-County Study will be to gather and analyze the data necessary to help counties identify the program, service, and sanction interventions that offer promise in reducing recidivism. A greater understanding of these effects will be essential to help policymakers and practitioners identify strategies that will reduce recidivism and maintain public safety. PPIC.ORG Realignment and Recidivism in California 20 REFERENCES Bird, Mia, and Ryken Grattet. 2015. “Policy Change and Recidivism: The Effects of California’s Realignment and Local Implementation Strategies on Rearrest and Reconviction.” Criminal Justice Policy Review 28 (6): 601–623. Bird, Mia, and Ryken Grattet. 2016. “Realignment and Recidivism.” The Annals of the American Academy of Political and Social Science 664 (1): 176–195. California Department of Corrections and Rehabilitation. 2017a. 2016 Outcome Evaluation Report: An Examination of Offenders Released in Fiscal Year 2011–12. California Department of Corrections and Rehabilitation, Office of Research, October. California Department of Corrections and Rehabilitation. 2017b. 2017 Outcome Evaluation Report: An Examination of Offenders Released in Fiscal Year 2012–13. California Department of Corrections and Rehabilitation, Office of Research, October. Couzens, J. Richard, and Tricia A. Bigelow. 2016. Felony Sentencing after Realignment. California Courts, May. Lofstrom, Magnus, Mia Bird, and Brandon Martin. 2016. California’s Historic Corrections Reforms. Public Policy Institute of California. Lofstrom, Magnus, and Steven Raphael. 2015. Realignment, Incarceration, and Crime Trends in California. Public Policy Institute of California. Lofstrom, Magnus, and Steven Raphael. 2013. Public Safety Realignment and Crimes Rates in California. Public Policy Institute of California. Lofstrom, Magnus, Steven Raphael, and Ryken Grattet. 2014. Is Public Safety Realignment Reducing Recidivism in California? Public Policy Institute of California. Lofstrom, Magnus, and Brandon Martin. 2015. Public Safety Realignment: Impacts So Far. Public Policy Institute of California. Martin, Brandon, and Magnus Lofstrom. 2014. Key Factors in California’s Jail Construction Needs. Public Policy Institute of California. Misczynski, Dean. 2012. Corrections Realignment: One Year Later. Public Policy Institute of California. Santos, Michael. 2013. “California's Realignment: Real Prison Reform or Shell Game?” The Blog, HuffPost, March 11. Weisberg, Robert. 2011. “California’s De Facto Sentencing Commissions.” Stanford Law Review 64 (1): 1–7. PPIC.ORG Realignment and Recidivism in California 21 ABOUT THE AUTHORS Mia Bird is a research fellow in the areas of corrections and health and human services at the Public Policy Institute of California. She also serves on the faculty of the Goldman School of Public Policy at the University of California, Berkeley. At PPIC, she cofounded and directs a collaborative project between the California Board of State and Community Corrections (BSCC), PPIC, and 12 California counties, known as the BSCC–PPIC Multi-County Study. This data collection and evaluation effort is designed to estimate the effects of realignment on recidivism outcomes and identify best practices for recidivism reduction at the local level. She also leads a project focused on the impact of the Affordable Care Act on enrollment and recidivism outcomes for the criminal justice population. Her past work has covered topics such as the allocation of realignment funding, healthcare for the correctional population, and the use of data to improve policymaking. She holds a PhD in public policy, an MA in demography, and an MPP from the University of California, Berkeley. Ryken Grattet is an adjunct fellow at the Public Policy Institute of California and professor and chair of sociology at the University of California, Davis. Previously, he served as assistant secretary of research in the California Department of Corrections and Rehabilitation. He is a cofounder of a collaborative project between the California Board of State and Community Corrections (BSCC), PPIC, and 12 partner counties, known as the BSCC–PPIC Multi-County Study. This data collection and evaluation effort is designed to estimate the effects of realignment on recidivism outcomes and identify best practices for recidivism reduction at the local level. He is the co-author of Making Hate a Crime: From Social Movement to Law Enforcement, Parole Violations and Revocations in California, and numerous articles in professional and policy publications. He holds a PhD in sociology from the University of California, Santa Barbara. Viet Nguyen is a research associate at the Public Policy Institute of California. He conducts research on corrections policy, including the effects of realignment. He is also the data manager for a collaborative project between the California Board of State and Community Corrections (BSCC), PPIC, and 12 partner counties, known as the BSCC–PPIC Multi-County Study. Before joining PPIC, he was a survey specialist and operations analyst at NORC at the University of Chicago. He holds a BA in political science, with a minor in public policy, from the University of California, Los Angeles. ACKNOWLEDGMENTS This report benefited from valuable feedback from our external reviewers, Alison Anderson and Scott MacDonald. We also appreciate the helpful feedback of our internal reviewers, Magnus Lofstrom and Shannon McConville; insights from our director of government affairs, Deborah Gonzalez; and editorial support from Vicki Hsieh and Lynette Ubois. This work would not be possible without the support of the Board of State and Community Corrections and data provided by the California Department of Corrections and Rehabilitation and the California Department of Justice. This report also draws on data provided by the 12 counties participating in the BSCC–PPIC Multi-County Study. We are grateful to these partner counties for their willingness to participate in this effort and their interest in data-driven practice. This kind of sustained support for research among California’s justice agencies enables efforts to investigate the effects of major criminal justice reforms and to identify promising practices for the future. Finally, we acknowledge our colleague, Sonya Tafoya, who is an original cofounder of the MCS and greatly contributed to the project while she was a research associate at PPIC. This project was supported by Award No. 2013-R2-CX-0010, awarded by the National Institute of Justice, Office of Justice Programs, US Department of Justice. The opinions, findings, and conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect those of the Department of Justice. PPIC.ORG Realignment and Recidivism in California 22 PUBLIC POLICY INSTITUTE OF CALIFORNIA Board of Directors Mas Masumoto, Chair Author and Farmer Mark Baldassare President and CEO Public Policy Institute of California Ruben Barrales President and CEO, GROW Elect María Blanco Executive Director University of California Immigrant Legal Services Center Louise Henry Bryson Chair Emerita, Board of Trustees J. Paul Getty Trust A. Marisa Chun Partner, McDermott Will & Emery LLP Chet Hewitt President and CEO Sierra Health Foundation Phil Isenberg Former Chair Delta Stewardship Council Donna Lucas Chief Executive Officer Lucas Public Affairs Steven A. Merksamer Senior Partner Nielsen, Merksamer, Parrinello, Gross & Leoni, LLP Leon E. Panetta Chairman The Panetta Institute for Public Policy Gerald L. Parsky Chairman, Aurora Capital Group Kim Polese Chairman, ClearStreet, Inc. Gaddi H. Vasquez Senior Vice President, Government Affairs Edison International Southern California Edison The Public Policy Institute of California is dedicated to informing and improving public policy in California through independent, objective, nonpartisan research. Public Policy Institute of California 500 Washington Street, Suite 600 San Francisco, CA 94111 T: 415.291.4400 F: 415.291.4401 PPIC.ORG PPIC Sacramento Center Senator Office Building 1121 L Street, Suite 801 Sacramento, CA 95814 T: 916.440.1120 F: 916.440.1121" } ["___content":protected]=> string(135) "

Realignment and Recidivism in California

" ["_permalink":protected]=> string(84) "https://www.ppic.org/publication/realignment-and-recidivism-in-california/r_1217mbr/" ["_next":protected]=> array(0) { } ["_prev":protected]=> array(0) { } ["_css_class":protected]=> NULL ["id"]=> int(13053) ["ID"]=> int(13053) ["post_author"]=> string(1) "4" ["post_content"]=> string(0) "" ["post_date"]=> string(19) "2017-12-11 19:32:41" ["post_excerpt"]=> string(0) "" ["post_parent"]=> int(12749) ["post_status"]=> string(7) "inherit" ["post_title"]=> string(40) "Realignment and Recidivism in California" ["post_type"]=> string(10) "attachment" ["slug"]=> string(9) "r_1217mbr" ["__type":protected]=> NULL ["_wp_attached_file"]=> string(13) "r_1217mbr.pdf" ["wpmf_size"]=> string(6) "465432" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(61723) "DECEMBER 2017 Mia Bird, Ryken Grattet, and Viet Nguyen with research support from Justin Goss Supported with funding from the National Institute of Justice Realignment and Recidivism in California © 2017 Public Policy Institute of California PPIC is a public charity. It does not take or support positions on any ballot measures or on any local, state, or federal legislation, nor does it endorse, support, or oppose any political parties or candidates for public office. Short sections of text, not to exceed three paragraphs, may be quoted without written permission provided that full attribution is given to the source. Research publications reflect the views of the authors and do not necessarily reflect the views of our funders or of the staff, officers, advisory councils, or board of directors of the Public Policy Institute of California. SUMMARY CONTENTS Introduction The BSCC–PPIC Multi-County Study Realigned Offenders Released from Prison (PRCS) Realigned Offenders Sentenced under 1170(h) Conclusions and Policy Implications References About the Authors Acknowledgments 5 7 9 14 19 21 22 22 A technical appendix to this paper is available on the PPIC website. PPIC.ORG California has experienced significant changes in its criminal justice landscape since the 2011 implementation of public safety realignment—which shifted the management of lower-level offenders from the state prison and parole system to county jail and probation systems. The prison population has dropped dramatically, and though jail populations rose, overall incarceration levels have declined. One goal of realignment was to reduce California’s persistently high recidivism rates. Using data from 12 counties representative of the state, this report examines rearrest and reconviction rates after release from custody for two groups of offenders affected by realignment: those on post-release community supervision (PRCS) and those sentenced under section 1170(h) of the California Penal Code. Overall, we find realignment had modest effects on recidivism, with considerable variation across offender groups and counties. Specifically:  Individuals on PRCS have somewhat higher recidivism than similar individuals released before realignment. PRCS offenders are released from state prison after serving time for certain lower-level felonies and receive supervision by county probation agencies. In the two years following realignment, we find that 71.9 percent of these individuals were rearrested (2.6 percentage points higher than before realignment), and 56.4 percent were reconvicted (2.4 points higher).  Realignment did not have a consistent effect on recidivism for individuals sentenced under 1170(h). These offenders are sentenced for a specific set of lower-level felonies and, under realignment, serve time in county jail rather than state prison. In the two years following realignment, we find that 74.5 percent of these individuals were rearrested (2.3 percentage points higher than their pre-realignment counterparts) and 54.9 percent were reconvicted (2.0 points lower).  Offenders who received straight sentences have the same or lower rates of recidivism. Realignment created two types of 1170(h) offenders: those who receive both jail time and probation supervision (known as a “split” sentence) and those who receive jail time with no supervision (known as a “straight” sentence). The group serving “straight” sentences had the best outcomes: the same two-year rearrest rates and lower two-year reconviction rates (by 3.0 percentage points). Those who received “split” sentences had higher rates of rearrest (by 7.8 points) but lower rates of reconviction (by 3.4 points) compared with similar individuals before realignment.  The effects of realignment on recidivism vary substantially across counties. For example, overall we find reconviction rates were higher for those on PRCS after realignment, but in fact nine counties saw lower Realignment and Recidivism in California 3 rates of reconviction—indicating that the overall finding is driven by a small number of counties. County variation in recidivism outcomes is likely linked to demographic, economic, and geographic differences, as well as the range of county capacity and experiences providing evidence-based interventions before realignment. However, some of this variation may be due to different intervention strategies, creating the potential for counties to learn from each other over time. Notably, offenders who received a jail term and no supervision stand out as having better outcomes on all measures of recidivism, when compared with similar individuals released before realignment. This finding suggests we need to carefully consider the complex relationship between supervision and recidivism. While it could simply be easier to detect reoffending when an individual is under supervision, the requirements of supervision could also create more opportunities for non-criminal violations. With a longer follow-up window and more recent data, the relationship between supervision and recidivism, as well as the overall effects of realignment, may change as counties build capacity and experience with evidence-based practices. Policymakers, practitioners, and researchers must continue to work together to develop the data and expertise necessary to understand the impacts of California’s corrections reforms and to identify effective strategies to reduce recidivism. PPIC.ORG Realignment and Recidivism in California 4 Introduction In 2011, California passed public safety realignment, one of the most far-reaching criminal justice reforms in recent US history.1 Realignment has been referred to as “revolutionary and sudden” (Weisberg 2011), “the most significant correctional reform in decades” (Misczynski 2012), and “the biggest penal experiment in modern history” (Santos 2013). This law was motivated by a US Supreme Court decision, which found that California’s prison system could not adequately meet inmates’ health care needs due to extreme overcrowding. The court mandated that the state either dramatically increase prison capacity or reduce the inmate population by tens of thousands. At the time, California faced a severe budget crisis, with limited ability to build new prisons or contract out beds and services to other entities. Under these constraints, the state elected to shift (or “realign”) the management of lower-level felony offenders from the state prison and parole system to county jail and probation systems. Realignment created two new populations of offenders:  Individuals on post-release community supervision (PRCS). The PRCS population refers to individuals released from prison after serving terms for non-serious and non-violent felony offenses. Under realignment, individuals on PRCS are supervised by county probation departments rather than the state parole system. Prior to realignment, people released from prison after serving time for these offenses would have been revoked to prison custody for supervision violations—which contributed to prison overcrowding. Now, the PRCS population serves any revocations in county jails.  Individuals sentenced under 1170(h). The 1170(h) population refers to non-serious, non-violent, nonsexual felony offenders who serve their sentences in county jail. Prior to realignment, individuals convicted for specific offenses defined in §1170(h) of the California Penal Code would have been sent to state prison if their custody sentences exceeded one year. Like the PRCS population, 1170(h) offenders are revoked to local jail custody, not prison custody, for supervision violations. Realignment further created two types of 1170(h) offenders: those who receive a “straight” jail sentence for the full term and no probation following release, and those who have their sentence “split” between a jail term and probation. Realignment effectively reduced the state’s prison population by more than 27,000 in the first year of implementation (Lofstrom and Martin 2015). As expected, jail populations increased and, in many counties, jails reached or exceeded capacity (Martin and Lofstrom 2014). However, the total statewide jail population increased by fewer than 9,000 inmates, or about one-third of the decline in the prison population. As a result, realignment led to a reduction in both the prison population and overall incarceration levels. The reform prompted considerable debate over the relationship between incarceration and public safety. Proponents of realignment anticipated the reduction in incarceration and argued that pre-realignment incarceration levels had been too high. Some supporters claimed that increased corrections spending on incarceration was not a cost-effective way to improve public safety, and others maintained that the pre-realignment system created perverse incentives for counties to shift the burden of preserving public safety to the state, resulting in overly punitive sanctions against lower-level offenders. However, opponents of realignment had serious concerns that an increase in “street time” for former offenders would lead to rising crime rates. Critics were also concerned that the perceived reduction in penalties would lead to increased reoffending. To date, research examining the effects of realignment on crime has found no increase in violent crime under realignment and small increases in property 1 Public safety realignment was initiated through the passage of Assembly Bill 109 in 2011. Subsequent legislation has further defined the details of realignment implementation. PPIC.ORG Realignment and Recidivism in California 5 crime, driven specifically by an increase in motor vehicle theft (Lofstrom and Raphael 2013; Lofstrom and Raphael 2015). For proponents of realignment, another goal was to reduce the state’s persistently high rates of recidivism, which are among the highest in the nation. Supporters put forth the notion that “locals can do it better,” referring to a belief that local justice systems are better positioned than the state prison and parole system to provide rehabilitative interventions shown by research to be effective in reducing recidivism. While local justice systems supported this idea in principle, they also voiced concerns about their abilities to ramp up capacity in time and with the level of funding that was provided—particularly with respect to the use of evidence-based interventions. Findings on the effects of realignment on recidivism are still emerging from the state and research community— in part because evaluating the impact of a policy change on recidivism rates requires a substantial follow-up period after individuals are released from custody. In October 2017, the California Department of Corrections and Rehabilitation (CDCR) published reports on the recidivism outcomes of individuals released from prison, including the PRCS population, after realignment (CDCR 2017a, 2017b). As intended, returns to prison custody have declined dramatically. CDCR describes an increase in reconviction for offenders released from prison during the first fiscal year after realignment and a decrease for offenders released in the second year, relative to the reconviction rates of individuals released in the two years prior to realignment.2 These are descriptive findings, and as CDCR notes, the population of individuals released from prison is changing in composition over time. As realignment rolls out, this group is increasingly composed of serious and violent offenders released following sentences imposed for new convictions (rather than for revocations).3 Previous research generally finds that recidivism rates have not yet improved for individuals released from prison under realignment (Lofstrom, Raphael, and Grattet 2014; Lofstrom, Bird, and Martin 2016). Those released on PRCS have been found to have higher rates of recidivism relative to their pre-realignment counterparts (Bird and Grattet 2015; Lofstrom, Bird, and Martin 2016). However, recidivism rates vary in important ways across counties, and some evidence suggests this variation may be linked to county capacity and the propensity to use rehabilitative programs and services with the PRCS population (Bird and Grattet 2015; Bird and Grattet 2016). Thus far, research on the effects of realignment on recidivism has been limited in two important ways. First, prior studies have not been able to examine recidivism among the 1170(h) population.4 In contrast to the PRCS population, which spiked in the first year of realignment and leveled out, over time, individuals sentenced under 1170(h) will make up a larger portion of the realigned population. Second, existing research has not captured returns to jail custody—a key measure of recidivism—for those on parole, those on PRCS, or those sentenced under 1170(h). This report addresses those gaps. We examine recidivism outcomes for the PRCS and 1170(h) populations for a representative group of 12 California counties. Our analysis focuses on the impact of realignment on one- and two-year recidivism rates for individuals released from prison or jail between October 2011 and September 2013—relative to those released from prison custody before realignment, from October 2009 to September 2011. We also present differences in one-year recidivism rates before and after realignment for each of the 12 counties. 2 For consistency with the state definition of recidivism (“conviction of a new felony or misdemeanor committed within three years of release from custody”), CDCR uses reconviction as the primary measure of recidivism. Section 3027 of the California Penal Code required the Board of State and Community Corrections to develop a statewide definition of recidivism. 3 CDCR has shared data with researchers to allow for recidivism analyses that attempt to adjust for the changing composition of the population released from prison, as well as to focus on outcomes for the PRCS population. 4 There is no statewide data source capturing the 1170(h) population and their recidivism outcomes. Nor is there a statewide data source capturing returns to jail custody for the parole, PRCS, and 1170(h) populations. PPIC.ORG Realignment and Recidivism in California 6 The BSCC–PPIC Multi-County Study Our analysis relies primarily on data from the BSCC–PPIC Multi-County Study (MCS), a collaborative effort between the California Board of State and Community Corrections (BSCC) and PPIC.5 The MCS was established in the wake of public safety realignment with the goal of bringing together the data needed to rigorously evaluate the statewide effects of this policy reform and to identify the most effective recidivism-reduction interventions at the local level. To achieve these goals, we identified a group of counties representative of the state as a whole and partnered with these counties to examine how individuals move through local jail and probation systems after realignment. Figure 1 shows the 12 counties participating in the study: Alameda, Contra Costa, Fresno, Humboldt, Kern, Los Angeles, Orange, Sacramento, San Bernardino, San Francisco, Shasta, and Stanislaus. FIGURE 1 The 12 counties participating in the MCS represent California’s geographic diversity SOURCE: BSCC–PPIC Multi-County Study. NOTE: These were the only 12 counties approached by the MCS team. No counties were approached and declined to participate. Taken together, these counties comprise 60 percent of California’s population and represent the state’s geographic diversity, as well as its overall demographic and economic characteristics. Table 1 summarizes the characteristics of the MCS counties relative to the statewide population. While quite similar, the MCS counties tend to be more 5 PPIC research publications reflect the views of the authors and do not necessarily reflect the views of the BSCC, CDCR, the Department of Justice, or the participating counties. PPIC.ORG Realignment and Recidivism in California 7 urban (as measured by population density) and have higher shares of African Americans, Asian Americans, and Latinos. In addition, the poverty and unemployment rates are slightly higher among the MCS counties. TABLE 1 MCS counties are similar to the state overall in demographic and economic characteristics California MCS counties Demographic characteristics Male 49.7% 49.4% African American 6.5% 8.0% Asian American 14.8% 16.2% Latino 38.6% 40.6% Native American 1.7% 1.5% White 73.3% 70.8% Two or more 3.7% 3.6% Under 20 26.4% 26.2% Age 20–39 28.9% 29.5% Age 40–59 26.5% 26.8% Age 60+ 18.2% 17.5% Population density (population per square mile) 244.6 454.7 Economic characteristics Unemployment rate 9.1% 9.2% Poverty rate 16.5% 17.6% Total population 38,335,203 22,847,093 SOURCES: Demographic and population density characteristics are from the US Census. Poverty rates are from the Small Area Income and Poverty Estimates (SAIPE) program within the Census Bureau. Unemployment rates are from the Bureau of Labor Statistics. NOTES: Characteristics for the MCS county group are population-weighted for the year 2013. In addition to the data provided by the counties, the California Department of Justice (DOJ) and the CDCR provided essential data to fill out the state-local picture. Altogether, the newly available data used in this analysis includes demographic characteristics, offender background and criminal history, and recidivism outcomes. The participation of the MCS counties allows us to expand on previous research by assessing outcomes for individuals sentenced to serve time in county correctional agencies, who, because they pass through local systems, are not tracked at the state level. We are also able to examine returns to jail custody within the MCS counties. Finally, sentencing information from these counties allows us to explore the relationship between split sentencing—a new tool under realignment—and recidivism for the first time. PPIC.ORG Realignment and Recidivism in California 8 Measuring Recidivism There are many ways of measuring recidivism.6 The state has adopted reconviction—often considered a more accurate measure of reoffending behavior, relative to rearrest—as its primary recidivism measure. Reconvictions also represent a substantial resource burden and thus may be of particular interest to correctional systems. However, when comparing the reconviction rates of individuals released before and after realignment, it is important to consider how realignment changed the likelihood that criminal justice systems would pursue formal convictions. Before realignment, individuals who violated the conditions of their parole or who were suspected of a new offense could be revoked to prison custody. Under realignment, most offenders can only be sent to prison following conviction on a new, prison-eligible offense—meaning that correctional systems may be shifting away from revocations and toward more formal rearrests and reconvictions. Therefore, if we were to see increases in reconviction rates after realignment, we should not necessarily assume these differences were driven by an increase in reoffending behavior. However, realignment may have also created new incentives for individual offenders. Research on recidivism generally aims to identify how changes in policy or practice affect underlying criminal behavior, but it is often difficult to determine whether changes in recidivism measures reflect changes in criminal behavior or, alternatively, changes in criminal justice processes. In this report, we account for potential changes in criminal justice processes by constructing a measure of reconviction that includes both formal reconvictions and revocations. We use a similar measure for rearrests that includes both formal rearrests and any revocations not preceded by a formal arrest. Using these two measures, we estimate the effects of realignment on recidivism for individuals on PRCS and individuals sentenced under 1170(h).7 Realigned Offenders Released from Prison (PRCS) As defined in realignment law, post-release community supervision refers to a group of offenders who are released from state prison after having been committed for a non-serious, non-violent felony offense.8 The law disqualifies individuals from PRCS who are determined through a CDCR assessment process to be “high-risk sex offenders” or “mentally disordered offenders” (Couzens and Bigelow 2016). Importantly, PRCS does not affect the length of time a person serves in custody, but it does change where one is supervised upon release. People on PRCS are supervised in the community by county probation departments. Prior to realignment, they were supervised by the state parole system.9 In the case of a supervision violation, individuals on PRCS are revoked to jail instead of prison custody. In addition to the change in which agency supervises them, realignment altered other conditions of their supervision as well, including their eligibility for early discharge and new limits on revocations.10 6 Indeed, it is important to note that the best measures of recidivism depend on the purpose of the analysis. For example, if the primary goal is to understand the effect of a policy change on the cost of operating correctional institutions, then the most appropriate metric to examine may be returns to custody. If the goal is to understand how a policy change has affected the demands placed on police departments, then the best recidivism measure may be rearrest rates. No single measure of recidivism fits all purposes. 7 In the Technical Appendix, we also present traditional rearrest and reconviction measures without adjustments for revocations. Note that the 1170(h) population, given their local sentence length, is limited to those likely to have gone to prison prior to realignment, as explained in more detail in the Technical Appendix. 8 PRCS offenders may have prior serious or violent offenses in their criminal history. 9 State parole and county probation have many similarities in terms of the types of supervision practices, legal authority, and surveillance tools available. However, there are some important differences as well. A large portion of parole officers spent at least some portion of their careers working in prison as correctional guards. Probation officers are less likely to have such experiences. Parole officers carry firearms, while probation officers typically do not. Probation is subject to variation by county, given that it is a county government agency, dependent on resources provided by the County Boards of Supervisors. Parole is more centralized and uniformly administered across the state. 10 If individuals on PRCS have no violations that result in a custodial sanction, they are eligible for early discharge from supervision after six months. Prior to realignment, eligibility for discharge from parole commenced at 12 months. In addition, realignment imposed limits on the amount of time an individual under PPIC.ORG Realignment and Recidivism in California 9 Characteristics During the first two years under realignment, more than 35,000 individuals were released on PRCS to MCS counties.11 To estimate the effects of realignment on recidivism, we compare the outcomes of those on PRCS after realignment to those of similar individuals who were released from prison in the two years before realignment (between October 2009 and September 2011). A key concern when comparing outcomes of those on PRCS to those of the pre-realignment population is that there are considerable differences between these two groups. Demographically, the pre-realignment population—which includes all individuals released from state prison in the two years preceding realignment—is younger, less likely to be white, more likely to be Latino, and less likely to face mental health challenges than those on PRCS (Table 2). Overall, more than one-quarter of the pre-realignment population is incarcerated for a crime against a person, compared with less than one-sixth of the PRCS population. The pre-realignment population is also more likely to be incarcerated on a conviction for a new offense (rather than a revocation). Those released from prison before realignment also have lower numbers of past arrests and convictions when compared with individuals on PRCS. These differences in characteristics suggest the pre-realignment population includes more serious offenders, but these offenders may have a lower likelihood of recidivism than the PRCS population. We use a matching method to refine the pre-realignment population to a control group that is more similar to the PRCS population. This approach allows us to better isolate the effects of realignment when examining differences in recidivism.12 The demographic characteristics and criminal histories of the pre-realignment control group and the PRCS population are quite similar. However, those on PRCS are still less likely to have been incarcerated for a new conviction and, instead, are more likely to be released after serving a revocation. Our analysis below on recidivism outcomes uses a regression model to adjust for these remaining differences between the prerealignment control group and those on PRCS. TABLE 2 The PRCS population differs considerably from the pre-realignment population of those released from prison Demographic characteristics Age Male African American Asian American Latino Native American White Other race Pre-realignment population 36.2 88.9% 28.4% 1.0% 43.5% 0.6% 23.3% 3.3% Pre-realignment control group PRCS population 37.1 89.3% 28.8% 0.9% 40.0% 0.7% 26.5% 3.2% 37.5 89.9% 29.0% 0.9% 39.8% 0.7% 26.5% 3.1% supervision could be revoked to custody for violations, to a maximum of six months. It also changed the location where those revocations are to be served, from state prison to county jail. 11 Here we refer to individuals. The total number of PRCS releases would be higher than the total number of individuals released over the period. This report focuses on recidivism outcomes for this subset of the statewide PRCS population because the MCS dataset provides information on revocations to jail custody, which is not available for the full, statewide PRCS population. 12 We use a propensity score matching method to select individuals from the larger pre-realignment population that are similar to the PRCS population. This method is described in more detail in the Technical Appendix. PPIC.ORG Realignment and Recidivism in California 10 Pre-realignment population Pre-realignment control group PRCS population Offender background and criminal history Age at first conviction 23.3 22.6 22.4 History of mental illness 15.9% 18.5% 19.0% Length of stay in prison (in days) 636.9 488.4 504.4 New conviction 67.8% 59.6% 56.2% Parole violator with new conviction 19.1% 30.3% 34.5% Parole violator 11.9% 9.3% 9.3% Number of past arrests 15.1 18.1 18.8 Number of past felony arrests 9.4 11.2 11.6 Number of past convictions 5.9 7.1 7.4 Number of past felony convictions 3.2 3.9 4.2 Number of prior serious convictions 0.2 0.3 0.4 Number of prior violent convictions 0.1 0.1 0.1 High-risk for a person offense 11.1% 10.7% 11.5% High-risk for a property offense 9.2% 11.9% 11.6% High-risk for a drug offense 5.0% 7.4% 7.7% Moderate risk for reoffending 21.2% 23.1% 23.0% Low risk for reoffending 16.5% 15.8% 15.1% Second-striker 13.4% 17.1% 18.6% Offense type Murder 0.9% 0.0% 0.0% Manslaughter 0.9% 0.0% 0.1% Robbery 7.1% 0.0% 0.0% Other person offense 14.1% 14.1% 13.7% Rape 0.4% 0.0% 0.0% Molestation 2.3% 0.2% 0.2% Other sex offense 2.2% 2.0% 1.9% Burglary 11.3% 10.7% 10.2% Grand theft 3.6% 4.3% 4.0% Petty theft 4.2% 5.0% 5.1% Vehicle theft 4.4% 6.0% 6.2% Other property offense 7.4% 9.0% 8.9% Controlled substance sales, possession, manufacturing 26.4% 32.2% 32.7% Marijuana substance sales, possession, manufacturing 2.2% 2.1% 2.0% DUI 3.1% 2.8% 2.7% Weapon possession 5.9% 9.2% 10.3% Other offense 3.6% 2.4% 2.3% Number of observations 84,577 26,261 35,218 SOURCES: BSCC–PPIC Multi-County Study data. NOTES: We limit the analysis to first releases in the pre- and post-realignment periods. Therefore, these counts will not correspond perfectly with the total number of prison releases, as some individuals are incarcerated for multiple prison spells in a given time period. PPIC.ORG Realignment and Recidivism in California 11 Recidivism Outcomes Our findings suggest that realignment increased recidivism by a small amount for people on PRCS. This increase could be driven by a rise in offending behavior related to a perceived reduction in sanctions under realignment. It is also possible that the PRCS population was monitored more closely while under probation supervision than their pre-realignment counterparts under parole supervision. If this were the case, the higher rates of recidivism may be due to a greater likelihood of detecting misconduct under realignment or because the conditions of supervision created more opportunities for violations that led to revocations. Conditions of supervision might include drug testing, frequent reporting, required programs, or required meetings with probation officers, among others. We focus on two key measures of recidivism: rearrest and reconviction outcomes after release. As described earlier, we use a measure of rearrest that includes any revocations that are not preceded by a formal arrest, and a measure of reconviction that also includes revocations.13 After adjusting for differences between these groups, we estimate that those on PRCS have a one-year rearrest rate of 61.7 percent, or a notable 5.2 percentage points higher than their pre-realignment counterparts (Figure 2). When we examine two-year rearrest rates, however, the difference shrinks considerably. The two-year rearrest rate for individuals on PRCS is 71.9 percent, only 2.6 percentage points higher than that of similar individuals prior to realignment. Those on PRCS also have slightly higher reconviction rates when compared with similar individuals released before realignment: 40.1 percent (1.8 points higher than before realignment) of individuals on PRCS are reconvicted within one year of release and 56.4 percent (2.4 points higher) are reconvicted within two years. Although we have used regression models to adjust for differences in the characteristics available in the data, it is possible that people on PRCS differ from the pre-realignment control group in ways that we cannot account for and that would lead to higher rates of recidivism. It is also possible that the outcomes of realigned offenders may improve over time as counties become more experienced in managing the PRCS population and better able to provide evidence-based, recidivism-reduction interventions. Indeed, the findings presented may reflect some improvement over time, as the difference in the two-year rearrest rate is much lower than the difference in the one-year rearrest rate. Share of population (%) FIGURE 2 Individuals on PRCS have slightly higher recidivism rates than similar individuals before realignment 80 61.7 60 56.5 69.3 71.9 54.0 56.4 Pre-realignment PRCS 38.3 40.1 40 20 0 One-year rearrest Two-year rearrest One-year reconviction Two-year reconviction SOURCE: BSCC–PPIC Multi-County Study data. NOTE: Chart shows regression-adjusted rates, which should be interpreted as causal findings. All differences in recidivism rates between the pre-realignment control group and the PRCS population are statistically significant. Please see the Technical Appendix for more details. 13 The first- and second-year release cohorts are included in the one-year rearrest rate, but only the first-year release cohort is included in the two-year rearrest rate due to limitations in the date range over which we have data. PPIC.ORG Realignment and Recidivism in California 12 County Variation In California, there is considerable variation in recidivism outcomes across counties both before and after realignment. Across the state, counties have vastly different demographic compositions, economic resources, and geographic environments. For these reasons, counties should not be directly compared to each other or judged to be performing better or worse under realignment on the basis of this analysis. However, it will be important for counties to assess their own outcomes over time as they make efforts to improve. It is important to note that, while the recidivism rates presented below account for changes in the composition of county populations over time, they do not adjust for differences in population or other characteristics across counties. Therefore, these estimates should be compared within a given county over time, but not across counties. As shown in Figure 3, in some counties, such as Contra Costa, Humboldt, and Kern, those on PRCS have oneyear rearrest rates that are lower than the rates of their pre-realignment counterparts.14 In addition, in San Francisco County, we find similar rearrest rates for the PRCS population and the pre-realignment population. Overall, however, those on PRCS in most of the MCS counties have higher rearrest rates compared with similar individuals released to parole before realignment. FIGURE 3 In three out of twelve counties, those on PRCS have lower rearrest rates than similar individuals before realignment Alameda Contra Costa Fresno Humboldt Kern Los Angeles Orange Sacramento San Bernardino San Francisco Shasta Stanislaus Pre-realignment PRCS 0 20 40 60 80 One-year rearrest rates (%) SOURCE: BSCC–PPIC Multi-County Study data. NOTE: These are regression-adjusted recidivism rates. This analysis draws on small sample sizes in some counties and, therefore, we caution against causal interpretations. In addition, regression adjustments are made for changes in the composition of county populations over time, within county. Adjustments are not made for differences in population composition across counties and, therefore, these estimates should be compared within county over time, but not across counties. Figure 4 presents one-year reconviction rates for the 12 MCS counties. In nine of the twelve counties, individuals on PRCS have lower reconviction rates than similar individuals before realignment. This demonstrates that the overall finding of higher reconviction rates for the PRCS population (see Figure 2 above) is driven by higher rates in a few counties. It is important to note that about one-third of individuals on PRCS statewide and nearly one- 14 Our county-level analysis uses one-year rearrest and reconviction rates because two-year rates are based only on the first-year release cohort and the size of the 1170(h) populations (i.e., the treatment groups) are quite small for some counties. Though the PRCS populations (i.e., the treatment groups) are larger, we use the same approach for both groups of offenders for consistency. PPIC.ORG Realignment and Recidivism in California 13 half of individuals on PRCS in the MCS counties were released to Los Angeles County. Because Los Angeles is such a large county, it is a driving factor in the overall recidivism outcomes of the MCS group.15 FIGURE 4 In nine out of twelve counties, those on PRCS have lower reconviction rates than similar individuals before realignment Alameda Contra Costa Fresno Humboldt Kern Los Angeles Orange Sacramento San Bernardino San Francisco Shasta Stanislaus Pre-realignment PRCS 0 10 20 30 40 50 60 One-year reconviction rates (%) SOURCE: BSCC–PPIC Multi-County Study data. NOTE: These are regression-adjusted recidivism rates. This analysis draws on small sample sizes in some counties and, therefore, we caution against causal interpretations. In addition, regression adjustments are made for observable changes in the composition of county populations over time, within county. Adjustments are not made for differences in population composition across counties and, therefore, these estimates should be compared within county over time, but not across counties. Realigned Offenders Sentenced under 1170(h) Realignment law also defined a group of felony offenders under California Penal Code §1170(h) who now are ineligible for state prison and must serve their sentences in county correctional systems. In general, the 1170(h) population is made up of non-serious, non-violent, and non-sexual offenders who are sentenced to county jail under realignment.16 Those offenders who have a serious or violent offense in their past conviction history in California or another state are ineligible to be sentenced under 1170(h). Registered sex offenders are also excluded.17 Judges may sentence qualifying individuals to a straight sentence or a split sentence.18 Those who receive a straight sentence serve the entirety of their term in jail and, upon release, have no further supervision associated with the sentence. In contrast, those who receive a split sentence serve a portion of their term in jail and a portion under supervision by county probation. During this period of supervision, these individuals may be revoked to jail 15 Although overall reconviction rates for the PRCS population among MCS counties are higher than for similar individuals before realignment, when Los Angeles County is removed from the analysis we find the one-year reconviction rate for the PRCS population is 5.7 percentage points lower and the two-year reconviction rate is 2.2 percentage points lower than for those released before realignment. Los Angeles County is not alone among the MCS counties in having higher reconviction rates for those on PRCS, but it does have a disproportionate effect on the overall results because of its size. Please see the Technical Appendix for more details. 16 This includes over 1,200 separate sections of the California Penal Code. 17 In addition, offenses included in over 70 separate sections of the California Penal Code were deemed ineligible by the legislature, even though these offenses are not serious, violent, or sexual in nature. 18 For individuals sentenced after January 1, 2015, split sentencing became the presumptive sentence under California Penal Code §1170(h). PPIC.ORG Realignment and Recidivism in California 14 to serve out the remainder of their sentence. Realignment law did not change the length of sentences, so a split sentence is the same amount of time as a straight sentence for the same conviction offense.19 Characteristics The 1170(h) population includes nearly 13,500 individuals sentenced under 1170(h) and released from MCS county jails during the first two years of realignment.20 In contrast to the number of individuals on PRCS, which reached a peak in the first year of realignment and then leveled out, the number of people sentenced under 1170(h) more than doubled in the second year of realignment. This is to be expected, as 1170(h) convictions began only after the implementation of realignment, and those sentenced under 1170(h) must first complete their jail terms before they enter a release cohort. In addition, as realignment implementation rolls out over time, fewer individuals sent to prison will be eligible for PRCS, as non-serious, non-violent, non-sexual offenders are increasingly serving sentences in county correctional systems. Table 3 summarizes the characteristics of those released after serving 1170(h) sentences. As was the case with the PRCS population, the full pre-realignment population is demographically different from the 1170(h) population. In this case, the pre-realignment population is older, more likely to be male, and more likely to be African American than those sentenced under 1170(h). In terms of criminal history, the pre-realignment population has a higher number of prior serious convictions and a greater share are incarcerated for a crime against a person. But those sentenced under 1170(h) have higher rates of past arrests and convictions. These differences suggest that while offenders in the pre-realignment population may have more serious criminal histories, they may also be less frequent offenders than the 1170(h) population. As a result, we might expect the pre-realignment population to have a lower overall likelihood of recidivism. We again create a matched pre-realignment control group of similar individuals to serve as a better comparison for those sentenced under 1170(h). The characteristics of this comparison group are much more similar to those of the 1170(h) population. However, the pre-realignment control group is still slightly more likely to be African American and has slightly lower rates of past arrest and conviction. Therefore, below we use regression analysis, which adjusts for these remaining differences, to estimate the effects of realignment on recidivism for the 1170(h) population. TABLE 3 Individuals sentenced under 1170(h) differ in notable ways from those who were released from prison before realignment Demographics Age Pre-realignment population 36.2 Pre-realignment control group 34.8 1170(h) population 34.6 Male 88.9% 84.2% 84.1% African American Asian American Native American Latino White 28.4% 1.0% 0.6% 43.5% 23.3% 23.5% 1.4% 0.3% 43.6% 29.0% 22.4% 1.8% 0.2% 43.8% 29.7% 19 Moreover, in most cases a judge may elect to sentence an individual convicted for an 1170(h) offense to felony probation in lieu of a straight or split sentence. This option existed prior to realignment and remains an option today for both 1170(h)-eligible and prison-eligible offenses. 20 The Technical Appendix provides details on the construction of the MCS 1170(h) population. As explained therein, this population will not match total counts of 1170(h) convictions. PPIC.ORG Realignment and Recidivism in California 15 Pre-realignment population Pre-realignment control group 1170(h) population Other race 3.3% 2.3% 2.1% Offender background and history Age at first conviction 23.3 23.2 22.9 Length of stay in jail or prison (in days) 636.9 243.2 233.2 Number of past arrests 15.1 18.3 19.1 Number of past felony arrests 9.4 11.0 11.3 Number of past convictions 5.9 7.0 7.3 Number of past felony convictions 3.2 3.8 4.0 Number of prior serious convictions 0.2 0.1 0.1 Number of prior violent convictions 0.1 0.1 0.1 Offense type Crime against persons 27.8% 2.8% 2.6% Burglary 11.3% 16.5% 16.0% Grand theft 3.6% 5.0% 4.7% Petty theft 4.2% 4.6% 4.3% Vehicle theft 4.4% 7.9% 8.7% Other property offense 7.4% 15.3% 16.3% Controlled substance sales, possession, manufacturing 26.4% 35.4% 36.2% Marijuana substance sales, possession, manufacturing 2.2% 2.8% 2.7% DUI 3.1% 3.6% 3.3% Weapon possession 5.9% 3.6% 3.0% Other offense 3.6% 2.6% 2.3% Number of observations 84,577 10,869 13,435 SOURCES: BSCC–PPIC Multi-County Study data. NOTES: We limit the analysis to first releases in the pre- and post-realignment periods. Therefore, these counts will not correspond perfectly with the number of prison releases in any period or the counts of 1170(h) convictions, as some individuals are incarcerated for multiple prison or jail spells. In addition, for the analysis of the effects of realignment on recidivism, we further limit the 1170(h) population to those likely to have been sent to prison if they had been convicted prior to realignment. This limitation is described in more detail in the Technical Appendix. Recidivism Outcomes Overall, we find that realignment did not have a consistent effect on recidivism among the 1170(h) population. After adjusting for differences in the characteristics of the pre-realignment control group and the 1170(h) population, we find no differences in one-year rearrest rates (Figure 5). Those sentenced under 1170(h) have a two-year rearrest rate of 74.5 percent, slightly higher (by 2.3 percentage points) than that of similar individuals before realignment.21 However, when we examine reconviction rates, the 1170(h) population has a one-year reconviction rate of 39.1 percent, slightly lower (by 3.8 points) than that of the pre-realignment control group. We also estimate that those sentenced under 1170(h) have a two-year reconviction rate of 54.9 percent, 2.0 percentage points lower than that of their pre-realignment counterparts. 21 One-year recidivism rates are based on the first- and second-year release cohorts under realignment. Two-year rates are based only on the first-year release cohort due to limitations in the follow-up time period for which we have data. PPIC.ORG Realignment and Recidivism in California 16 These findings suggest the 1170(h) population had better outcomes under realignment—at least as measured by reconviction rates—than the PRCS population. However, the 1170(h) population includes both those with straight jail sentences and split sentences (a jail term followed by probation). When we conduct a separate analysis for these two groups, we find that those who received split sentences have higher one-year rearrest rates (by 7.2 percentage points) but lower one-year reconviction rates (by 5.4 percentage points) when compared to similar individuals released before realignment. The two-year rearrest rates for those who received split sentences are higher than those of the pre-realignment group (by 7.8 percentage points) and the two-year reconviction rates are lower than those of the pre-realignment control group (by 3.4 percentage points). In contrast, we find that those who received straight sentences have lower one-year rearrest rates (by 1.4 percentage points) and lower oneyear reconviction rates (by 4.1 percentage points) when compared to their pre-realignment counterparts. The twoyear rearrest rates for those given straight sentences are about the same as for their pre-realignment counterparts. However, the two-year reconviction rates are 3.0 percentage points lower than those of the pre-realignment comparison group. Therefore, lower rearrest rates among the full 1170(h) population are driven by lower rates for those receiving straight sentences. (Additional details are in Table A4 in the Technical Appendix.) FIGURE 5 The effects of realignment on recidivism vary among those sentenced under 1170(h) 80 72.2 74.5 61.6 61.7 60 56.9 54.9 42.9 39.1 40 Pre-realignment 1170(h) Share of population (%) 20 0 One-year rearrest Two-year rearrest One-year reconviction Two-year reconviction SOURCE: BSCC–PPIC Multi-County Study data. NOTE: Chart shows regression-adjusted rates, which should be interpreted as causal findings. The small difference in the one-year rearrest rate is not statistically significant. All other differences in recidivism rates between the pre-realignment control group and the PRCS population are statistically significant. Please see the Technical Appendix for more details. County Variation As is the case for individuals on PRCS, recidivism rates vary considerably across counties for those released after serving 1170(h) sentences. Here we see that more than half of the MCS counties have lower one-year rearrest rates among those released for 1170(h) sentences, when compared to a similar group of offenders released prior to realignment (Figure 6). However, as shown above (see Figure 5), we find no evidence of an overall difference in PPIC.ORG Realignment and Recidivism in California 17 one-year rearrest rates for the 1170(h) population after realignment—suggesting higher rearrest rates in a minority of counties outweigh lower rearrest rates in the majority of counties.22 FIGURE 6 Over half of counties have lower rearrest rates among 1170(h) offenders than similar individuals before realignment Alameda Contra Costa Fresno Humboldt Kern Los Angeles Orange Sacramento San Bernardino San Francisco Shasta Stanislaus 0 Pre-realignment 1170(h) 10 20 30 40 50 60 70 80 90 One-year rearrest rates (%) SOURCE: BSCC–PPIC Multi-County Study data. NOTE: These are regression-adjusted recidivism rates. This analysis draws on very small sample sizes in some counties and, therefore, we caution against causal interpretations. In addition, regression adjustments are made for changes in the composition of county populations over time, within county. Adjustments are not made for differences in population composition across counties and, therefore, these estimates should be compared within county over time, but not across counties. In the case of reconviction, more than half of the MCS counties have lower one-year reconviction rates, consistent with the overall findings (Figure 7). The variation across counties in the recidivism rates before and after realignment suggest there is a great deal to learn from the various county experiences under realignment. At the same time, the variation across counties in the pre-realignment period suggests that counties had different starting points and have considerably different populations to manage. 22 For example, when we remove Los Angeles from the analysis, we find that the one-year rearrest rate among 1170(h) offenders is 1.4 percentage points lower and the two-year rearrest rate is 0.7 percentage points lower, compared with those released before realignment. Please see the Technical Appendix for more details. PPIC.ORG Realignment and Recidivism in California 18 FIGURE 7 Over half of counties have lower reconviction rates among 1170(h) offenders than similar individuals before realignment Alameda Contra Costa Fresno Humboldt Kern Los Angeles Orange Pre-realignment 1170(h) Sacramento San Bernardino San Francisco Shasta Stanislaus 0 10 20 30 40 50 60 70 One-year reconviction rates (%) SOURCE: BSCC–PPIC Multi-County Study data. NOTE: These are regression-adjusted recidivism rates. This analysis draws on very small sample sizes in some counties and, therefore, we caution against causal interpretations. In addition, regression adjustments are made for changes in the composition of county populations over time, within county. Adjustments are not made for differences in population composition across counties and, therefore, these estimates should be compared within county over time, but not across counties. Conclusions and Policy Implications Public safety realignment marked a new era for corrections and rehabilitation—and raised questions about the reform’s effects on crime and recidivism. Our analysis examines recidivism among the two realigned groups of offenders using data from the BSCC–PPIC Multi-County Study, an integrated state-local collaboration with the Board of State and Community Corrections and 12 counties representative of the state. We find that individuals released under post-release community supervision during the first two years of realignment have slightly higher rates of rearrest and reconviction than their pre-realignment counterparts. Recidivism among individuals sentenced under 1170(h) is more mixed. The 1170(h) population has a higher two-year rearrest rate than their pre-realignment counterparts, but lower one- and two-year reconviction rates. Notably, among 1170(h) offenders, we find higher rearrest rates for those who received probation supervision following their jail term (known as a split sentence), but lower reconviction rates. Those who served a jail term with no probation after release (known as a straight sentence) have the same or lower recidivism rates when compared with their pre-realignment counterparts—the only group for which there is evidence of consistently better outcomes after realignment. Realignment has created a unique opportunity for counties to learn from each other and for the state to leverage the diversity of practices to identify the most effective recidivism-reduction strategies. We find substantial variation in both overall recidivism rates across counties and changes in those rates within counties over the PPIC.ORG Realignment and Recidivism in California 19 realignment period. While some of this variation is likely linked to differences in the demographic, economic, and geographic characteristics of these counties, it may also be explained by differences in county capacity and experience with evidence-based interventions before realignment. However, if some of this variation is due to different intervention strategies, counties could potentially improve their outcomes by learning from each other over time. Taken together, these findings suggest three key takeaways. First, the effects of realignment appear to vary across offender groups and counties. Second, as more data becomes available we may observe the effects of realignment changing over time. The majority of the realigned population will shift from PRCS offenders to 1170(h) offenders, and in the coming years, counties will continue to build capacity and experience with evidence-based and datadriven practices. And third, we need to carefully consider the relationship between supervision and higher recidivism rates. Our findings indicate higher rates of recidivism for the groups supervised following release: those on PRCS and those who received split sentences under 1170(h). While it may simply be easier to detect reoffending when an individual is under supervision, it is also possible that supervision conditions create more opportunities for non-criminal violations. Additional research is needed to further investigate the likely complex relationship between supervision and recidivism. The next steps in the Multi-County Study will be to gather and analyze the data necessary to help counties identify the program, service, and sanction interventions that offer promise in reducing recidivism. A greater understanding of these effects will be essential to help policymakers and practitioners identify strategies that will reduce recidivism and maintain public safety. PPIC.ORG Realignment and Recidivism in California 20 REFERENCES Bird, Mia, and Ryken Grattet. 2015. “Policy Change and Recidivism: The Effects of California’s Realignment and Local Implementation Strategies on Rearrest and Reconviction.” Criminal Justice Policy Review 28 (6): 601–623. Bird, Mia, and Ryken Grattet. 2016. “Realignment and Recidivism.” The Annals of the American Academy of Political and Social Science 664 (1): 176–195. California Department of Corrections and Rehabilitation. 2017a. 2016 Outcome Evaluation Report: An Examination of Offenders Released in Fiscal Year 2011–12. California Department of Corrections and Rehabilitation, Office of Research, October. California Department of Corrections and Rehabilitation. 2017b. 2017 Outcome Evaluation Report: An Examination of Offenders Released in Fiscal Year 2012–13. California Department of Corrections and Rehabilitation, Office of Research, October. Couzens, J. Richard, and Tricia A. Bigelow. 2016. Felony Sentencing after Realignment. California Courts, May. Lofstrom, Magnus, Mia Bird, and Brandon Martin. 2016. California’s Historic Corrections Reforms. Public Policy Institute of California. Lofstrom, Magnus, and Steven Raphael. 2015. Realignment, Incarceration, and Crime Trends in California. Public Policy Institute of California. Lofstrom, Magnus, and Steven Raphael. 2013. Public Safety Realignment and Crimes Rates in California. Public Policy Institute of California. Lofstrom, Magnus, Steven Raphael, and Ryken Grattet. 2014. Is Public Safety Realignment Reducing Recidivism in California? Public Policy Institute of California. Lofstrom, Magnus, and Brandon Martin. 2015. Public Safety Realignment: Impacts So Far. Public Policy Institute of California. Martin, Brandon, and Magnus Lofstrom. 2014. Key Factors in California’s Jail Construction Needs. Public Policy Institute of California. Misczynski, Dean. 2012. Corrections Realignment: One Year Later. Public Policy Institute of California. Santos, Michael. 2013. “California's Realignment: Real Prison Reform or Shell Game?” The Blog, HuffPost, March 11. Weisberg, Robert. 2011. “California’s De Facto Sentencing Commissions.” Stanford Law Review 64 (1): 1–7. PPIC.ORG Realignment and Recidivism in California 21 ABOUT THE AUTHORS Mia Bird is a research fellow in the areas of corrections and health and human services at the Public Policy Institute of California. She also serves on the faculty of the Goldman School of Public Policy at the University of California, Berkeley. At PPIC, she cofounded and directs a collaborative project between the California Board of State and Community Corrections (BSCC), PPIC, and 12 California counties, known as the BSCC–PPIC Multi-County Study. This data collection and evaluation effort is designed to estimate the effects of realignment on recidivism outcomes and identify best practices for recidivism reduction at the local level. She also leads a project focused on the impact of the Affordable Care Act on enrollment and recidivism outcomes for the criminal justice population. Her past work has covered topics such as the allocation of realignment funding, healthcare for the correctional population, and the use of data to improve policymaking. She holds a PhD in public policy, an MA in demography, and an MPP from the University of California, Berkeley. Ryken Grattet is an adjunct fellow at the Public Policy Institute of California and professor and chair of sociology at the University of California, Davis. Previously, he served as assistant secretary of research in the California Department of Corrections and Rehabilitation. He is a cofounder of a collaborative project between the California Board of State and Community Corrections (BSCC), PPIC, and 12 partner counties, known as the BSCC–PPIC Multi-County Study. This data collection and evaluation effort is designed to estimate the effects of realignment on recidivism outcomes and identify best practices for recidivism reduction at the local level. He is the co-author of Making Hate a Crime: From Social Movement to Law Enforcement, Parole Violations and Revocations in California, and numerous articles in professional and policy publications. He holds a PhD in sociology from the University of California, Santa Barbara. Viet Nguyen is a research associate at the Public Policy Institute of California. He conducts research on corrections policy, including the effects of realignment. He is also the data manager for a collaborative project between the California Board of State and Community Corrections (BSCC), PPIC, and 12 partner counties, known as the BSCC–PPIC Multi-County Study. Before joining PPIC, he was a survey specialist and operations analyst at NORC at the University of Chicago. He holds a BA in political science, with a minor in public policy, from the University of California, Los Angeles. ACKNOWLEDGMENTS This report benefited from valuable feedback from our external reviewers, Alison Anderson and Scott MacDonald. We also appreciate the helpful feedback of our internal reviewers, Magnus Lofstrom and Shannon McConville; insights from our director of government affairs, Deborah Gonzalez; and editorial support from Vicki Hsieh and Lynette Ubois. This work would not be possible without the support of the Board of State and Community Corrections and data provided by the California Department of Corrections and Rehabilitation and the California Department of Justice. This report also draws on data provided by the 12 counties participating in the BSCC–PPIC Multi-County Study. We are grateful to these partner counties for their willingness to participate in this effort and their interest in data-driven practice. This kind of sustained support for research among California’s justice agencies enables efforts to investigate the effects of major criminal justice reforms and to identify promising practices for the future. Finally, we acknowledge our colleague, Sonya Tafoya, who is an original cofounder of the MCS and greatly contributed to the project while she was a research associate at PPIC. This project was supported by Award No. 2013-R2-CX-0010, awarded by the National Institute of Justice, Office of Justice Programs, US Department of Justice. The opinions, findings, and conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect those of the Department of Justice. PPIC.ORG Realignment and Recidivism in California 22 PUBLIC POLICY INSTITUTE OF CALIFORNIA Board of Directors Mas Masumoto, Chair Author and Farmer Mark Baldassare President and CEO Public Policy Institute of California Ruben Barrales President and CEO, GROW Elect María Blanco Executive Director University of California Immigrant Legal Services Center Louise Henry Bryson Chair Emerita, Board of Trustees J. Paul Getty Trust A. Marisa Chun Partner, McDermott Will & Emery LLP Chet Hewitt President and CEO Sierra Health Foundation Phil Isenberg Former Chair Delta Stewardship Council Donna Lucas Chief Executive Officer Lucas Public Affairs Steven A. Merksamer Senior Partner Nielsen, Merksamer, Parrinello, Gross & Leoni, LLP Leon E. Panetta Chairman The Panetta Institute for Public Policy Gerald L. Parsky Chairman, Aurora Capital Group Kim Polese Chairman, ClearStreet, Inc. Gaddi H. Vasquez Senior Vice President, Government Affairs Edison International Southern California Edison The Public Policy Institute of California is dedicated to informing and improving public policy in California through independent, objective, nonpartisan research. Public Policy Institute of California 500 Washington Street, Suite 600 San Francisco, CA 94111 T: 415.291.4400 F: 415.291.4401 PPIC.ORG PPIC Sacramento Center Senator Office Building 1121 L Street, Suite 801 Sacramento, CA 95814 T: 916.440.1120 F: 916.440.1121" ["post_date_gmt"]=> string(19) "2017-12-12 03:32:41" ["comment_status"]=> string(4) "open" ["ping_status"]=> string(6) "closed" ["post_password"]=> string(0) "" ["post_name"]=> string(9) "r_1217mbr" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2017-12-11 19:33:28" ["post_modified_gmt"]=> string(19) "2017-12-12 03:33:28" ["post_content_filtered"]=> string(0) "" ["guid"]=> string(52) "http://www.ppic.org/wp-content/uploads/r_1217mbr.pdf" ["menu_order"]=> int(0) ["post_mime_type"]=> string(15) "application/pdf" ["comment_count"]=> string(1) "0" ["filter"]=> string(3) "raw" ["status"]=> string(7) "inherit" ["attachment_authors"]=> bool(false) }