Chief Stefan LoBuglio

Pre-Release and Reentry Services Division
Montgomery County (Maryland) Department of Correction and Rehabilitation
November 5, 2009

Statement by

Stefan LoBuglio, Chief
Pre-Release and Reentry Services Division
Montgomery County (Maryland) Department of Correction and Rehabilitation
11651 Nebel Street
Rockville, Maryland 20852
301/240-773-4262

stefan.lobuglio@montgomerycountymd.gov

Hearing before the
Senate Judiciary Committee

Subcommittee on Crime and Drugs

On

The First Line of Defense: Reducing Recidivism at the Local Level

Thursday, November 5, 2009
Dirksen Senate Office Building Room 226
2:00pm

Senator Cardin, Ranking Member Graham, and distinguished members of the Subcommittee,

My name is Stefan LoBuglio and I am Chief of Pre-Release and Reentry Services (PRRS) for the Montgomery County Department of Correction and Rehabilitation (DOCR) in Maryland. In my position, I work with over sixty correctional professionals and oversee a work-release program that transitions prisoners back into our community. Today, PRRS is supervising 175 male and female prisoners who are all within one-year of release, and who will complete their sentences in our community-based program. Our population includes 26 prisoners from the Federal Bureau of Prisons, 5 prisoners from the Maryland State's Division of Correction, and 144 prisoners from the two jails in Montgomery County. These individuals have been convicted for crimes ranging from misdemeanant petty theft and traffic offenses to Part I felony offenses for rape and homicide. They may be serving sentences ranging from 30 days to 30 years. A conviction for escape is the only offense that would disqualify a prisoner outright from the program.

The Pre-Release program requires "residents", which is the term that we prefer to use instead of "prisoners" or "inmates," to follow a customized reentry plan consisting of work, education, treatment, and family engagement. They pay program fees, taxes, child support, and restitution. Most reside in our accredited 177-bed community correctional facility, which is located 15 miles from this hearing room in Rockville Maryland. For almost forty years, the program has served to keep our jail under crowded by managing over 16,000 residents in our community-based pre-release center. In concert with the other operational divisions of the DOCR, PRRS serves as a vital component of the county's investment strategy to effectively and judiciously use jail beds and community-based programs to maximize public safety and to minimize social and economic costs.

Our program is one of many successful models of prisoner reentry that exist across the county, and our field has seen an explosion of interest in this topic over the past decade. However, with all of the accumulated knowledge about "what works" in prisoner reentry, why hasn't reentry penetrated the core of correctional practice for most of the 3,000 jail systems and 50 state prison systems in this country? Perhaps some delay is to be expected, as it takes time to change complex correctional operations. Our country has spent three decades quintupling our incarcerated population and it will take many years before prisoner reentry strategies become pervasive. A second and less examined reason, though, concerns the lack of incentives for correctional agencies to fully embrace a commitment to reentry and to take responsibility for lowering recidivism rates. Providing care, custody, and control in our jails and prisons is challenging to be sure, but fully within the scope and ability of correctional professionals. Reentry, however, requires a different, community focus. Adapting this new orientation is inhibited by the fact that the results are not easily measured, understood, or controlled.

In my testimony today, I want to focus on the incentive problem and suggest two critical roles that the federal government can and should play to expedite the development and adoption of robust reentry strategies in our correctional systems. The first involves providing states and local jurisdictions with incentives to develop an infrastructure of One-Stop Reentry Residential Centers (ORRC) in conjunction with the Federal Bureau of Prisons. These centers would transition individuals leaving federal and state prisons and local jails through a regimen of work, treatment, and family engagement in the last months of their sentence. Too many of our state and federal prisons and local jails are overcrowded or otherwise lack the capacity to place minimum security inmates into such programs - programs that have been shown to successfully manage risk while improving the transition from secure confinement to civilian life. As described in Appendix A, the Montgomery County Pre-Release Center provides a viable model that demonstrates the considerable benefits that these centers provide to the incarcerated population, their families, the community, and the overall functioning of the correctional system.

The second federal role involves developing robust data systems and analytical capabilities that would allow jurisdictions at all levels to measure key reentry performance measures in real time, and to readjust resources and policies as needed. The COMPSTAT model of informational analysis and resource deployment that transformed the New York City police department in the 1990s and that has fueled the growth of community policing nationwide provides the example for what is needed to spur the development of reentry strategies. Unfortunately, the myopic focus on recidivism rates as the single measure of success of reentry programs often obscures other key measures of community well-being and public safety. Also, recidivism proves surprisingly difficult to measure and interpret. For a recent study in Montgomery County, we encountered significant challenges in measuring our own recidivism outcomes, due to outmoded criminal justice record systems, replete with missing data, which must be accessed individually, and which require customized approaches to interpret and utilize the data. After this work, we found that the total three-year recidivism rates appeared high, but recidivism involving serious offenses was only one-third to one-fourth of the total rate. Preliminary results from that study are presented in Appendix B.

I will leave it to the other panelist to describe why jails "matter" in the current discussion of prisoner reentry, because of their size, scope, and the cost and complexity of their operations. In previous testimony that I submitted to the House Judiciary Committee in March 2007 on behalf of the Second Chance Act of 2007, I also addressed these issues at length: http://judiciary.house.gov/hearings/March2007/LoBuglio070320.pdf

One-Stop Reentry Residential Centers
Federal incentives are needed to spur state and local correctional agencies to collaborate with the Federal Bureau of Prisons (FBOP) to develop the nation's infrastructure of community-based One-Stop Reentry Residential Centers (ORRC). These centers would offer services and monitoring with much greater structure than the typical halfway house and would transition soon-to-be released prisoners from the federal, state, and local systems. They could be privately or publicly run, must be located near transportation networks and jobs, and would meet minimum standards of facility operations and services along the lines developed by the FBOP. The ORRC model reflects the facts that prisoners return from all three correctional systems at different frequencies and times, but inmates from all of these systems present similar public safety risks and have similar transitional requirements. Within a community, the ORRC would serve as the nexus for social services including housing, substance abuse, and mental health treatment, and serve as a coordinating mechanism with corrections, probation and parole, and local law enforcement to ensure that those returning are monitored carefully and appropriately.

Many studies of federal, state, and local correctional systems find that inmates are over-classified - that is, they are occupying prison and jail beds at security levels higher than warranted - often due to the lack of available community correctional beds. This is a costly policy problem. If one considers a medium or maximum security prison cell, with its accompanying high staffing ratios, as a scarce resource, good correctional practice would reserve these beds for the truly dangerous. Increasing the number of community correctional pre-release beds will make all of our prisons safer for staff and inmates by providing much greater incentives for inmates to comply while in custody in order to have a greater chance of being "stepped down" to a community program. Research amply demonstrates the ability of these programs to reduce recidivism and improve institutional operations. Therefore, the development of ORRC capacity will lead to improvements within correctional systems beyond the improvement of transitional services.

The Federal Bureau of Prisons currently oversees and funds the largest number of ORRC-type programs. Under mandate from the Second Chance Act of 2007 and with the support of its leadership, the FBOP has accelerated its efforts to place federal prisoners in ORRCs when they are within one-year of release. They have developed an excellent system of oversight to ensure that locally-run programs meet the Bureau's strict set of performance and correctional standards. However, the FBOP is constrained by the absence or limited availability of community correctional beds in some jurisdictions such as Northern Virginia. As a result, soon-to-be released prisoners are either placed in ORRCs that are not proximate to their returning communities and/or spend much less time in the limited community beds that need to be rationed. Both of these responses to limited capacity undermine the model of inmate release underlying the program in the first place.

Although there is a chronic shortage of community correction beds in our country, curiously many go unused on a daily basis due to poorly coordinated practices between the different correctional agencies that contract for them. While the number of prisons vastly expanded in the past 30 years, there has been no proportional increase in community correction facilities. Many non-profits and religious organizations that had operated these centers lacked the financial capital required to bring them up to higher building codes and correctional accreditation standards, and some beds and facilities have been taken offline.

In some jurisdictions, the agency or organization running the community correctional facility may prefer to contract with the FBOP due to its ability to pay higher per diem costs than most state and local correctional agencies. Community correctional beds are not necessarily cheaper than institutional beds, and in tight budget times, state and local correctional agencies often cut these programs first. However, probably the larger reason for the low utilization rates concerns the lack of incentives that state and local correctional agencies have to fully engage in a reentry mission rather than retreat to the traditional goals of running clean, safe, and orderly institutions that meet correctional and constitutional standards.

Simply put, correctional agencies bear the costs and risks of reentry while the benefits accrue to individuals and the general community in ways that are hard to measure. Commissioner A.T. Wall of Rhode Island relates two closely-spaced incidents in his career that illustrate this paradox. A prisoner from a maximum security institution escaped but fortunately was apprehended. The escape triggered a public outcry for Wall's resignation, which Wall said he could understand. As commissioner he was ultimately responsible for safeguarding the prisoners in his system. By way of contrast, however, when a former prisoner committed a vicious killing just days after release, no one sought to put blame on him for poorly preparing this individual for release. The lesson, he said, was that correctional agencies will be held responsibility for the actions of individuals while in custody, but not after release. By definition, reentry extends the reach of corrections into the community and beyond the safe confines of the prison walls which makes it feel risky to many correctional practitioners.

The most notorious example of the risk averseness in community corrections was the decision by the Massachusetts Department of Correction to discontinue the use of halfway house beds in 1988 in response to the highly publicized and politicized case of Willie Horton. Horton escaped from a furlough program in Massachusetts and committed a heinous crime in Maryland before his apprehension. Many other correctional systems also retrenched their commitment to community corrections as a result of this case. Under the leadership of a new commissioner, the Massachusetts DOC finally resumed using halfway house beds in March of 2009 after a 21 year hiatus.

To address the concerns about risk, federal incentives could coalesce a wide spectrum of stakeholders from law enforcement to human service providers who have come to understand the public safety and human dimensions of prisoner reentry. Such a coalition could potentially mitigate the risk to correctional agencies in engaging in reentry by taking some ownership and responsibility in the program themselves. The coalition could also prove helpful in overcoming some of the resistance to site ORRC facilities near jobs and transportation. In some jurisdictions, the political will and correctional leadership may prove insufficient, and communities may decide not to allow the development of these community correctional centers. However, other jurisdictions may argue persuasively that ORRCs will prove more effective and safer in reentering local, state, and federal prisoners upon release to communities than the alternative of having them live transiently on the streets, shelters, and on the couches of friends.

Federal incentives can change the landscape of corrections. In 1994, the Truth-in-Sentencing legislation tied federal subsidies for corrections to sentencing reforms and helped spur a boom in the construction of prisons. Under this proposal, the federal government could use the same strategy by offering assistance to build and operate One-Stop Reentry Residential Centers for those state and local jurisdictions that agree to an integrated prisoner reentry strategy in coordination with the Federal Bureau of Prisons.

Develop Reentry COMPSTAT-type Informational Systems
As more jurisdictions develop reentry strategies, there is a federal role to help state and local authorities develop modern informational systems that can provide real-time analysis about their operations and effectiveness. Many of these systems were designed in the 1970s to handle the custodial and administrative duties, and are not easily adapted to track key reentry performance measures. Without federal intervention, the prospects for evaluating performance and being smart about the use of resources to further reentry goals are dim. At a recent conference of the National Association of Sentencing Commissions, Professor Anne Piehl of Rutgers University rued that in 2009 the difficulty of understanding and using criminal history data to compute recidivism rates and explore other key performance measures is just as challenging as it was twenty years ago, despite the revolution in information technology that has transformed so many other aspects of our society.

As evidence of this problem, I would like to mention a recidivism study that we recently conducted in Montgomery County in collaboration with Justice & Security Strategies, a consulting firm, and with funding from the Maryland Governor's Office of Crime Control and Prevention. The preliminary analytical results are presented in Appendix B. One of the main findings of the research project was just how difficult it is to conduct a comprehensive study. Even with four doctorate-level researchers and experts in interpreting criminal records from the local, state, and federal record systems, we had to access seven databases, contact half-a-dozen agencies, establish data sharing arrangements, print and review tens of thousands of pages of RAP sheets, develop algorithms to translate data from paper and from electronic sources consistently into newly created databases, and to double check repeatedly for inconsistencies and missing data. For instance, we found that the overlap in capturing Part I crimes in the FBI/NCIC databases and the Maryland State Record of Arrest and Prosecution was highly imperfect, and records were needed from each of the systems to complete comprehensive criminal records for the study sample.

Not surprisingly, the more places we looked for criminal records and the more sophisticated our data culling strategies became, the higher became our computed rates of recidivism. This gives credence to concerns that programs that receive funding based on promised reductions in recidivism will have few incentives to conduct exhaustive analyses. In our study, we found that measured recidivism rates were one-third higher than would have been measured by the "usual" technique of looking at Maryland state data only. Without access to effective reentry information systems, the reasonable expectations that evaluation will determine which programs are working for whom and to what extent is folly.

With federal support, we could develop information systems that help us use recidivism rates to better understand offender flow within criminal justice and social services systems for policy analysis, rather than as a crude measurement of program success. Recidivism is affected by many factors including changes in police, probation, and parole practices, and requires a nuanced understanding. For instance, in our study, we were surprised that adding measures of Violations of Probation as recidivist events to arrests did not markedly increase recidivism rates, which is contrary to the experience of other jurisdictions. This finding is forming the basis for discussions within our jurisdiction about the role of graduated sanctions within the overall criminal justice system.

Our study found relatively high rates of recidivism by arrest and by conviction within three years of release which were on the order of those measured by national studies. However, a careful analysis of the recidivist offense types found that rates of recidivism based on offenses classified as serious (according to Maryland sentencing law) was one-third to one-fourth of the overall rate, which again provides useful information for policy development. Through our use of survival analysis and hazard modeling, we also tracked the timing of recidivist events, which would allow us to explore and determine whether programs might both reduce and delay recidivism.

Despite the obvious value of the study to the County, and the possible influence of its findings on other agencies as well, the sheer effort it took and the accompanying time until results are known, mean that this type of careful research will be done sporadically, not routinely. And it is the routine, rapid feedback loop that is the cornerstone of the COMPSTAT and related innovations that have improved law enforcement performance in other arenas.

Conclusion
Prisoner Reentry requires leadership as much as it requires resources. Fortunately, as evidenced by the work of the other panel members in today's hearing, our field has no shortage of excellent practitioners and researchers who have advanced knowledge and understanding about promising strategies to transition soon-to-be released individuals from the country's jails and prisons. We also have many leaders from other government agencies and community institutions ranging from the police and faith-based organization who now strongly support reentry and who have declared a willingness to share the work, costs, and risks. However, as described in this testimony, I believe there is a need for federal leadership to accelerate the adoption of reentry strategies in many more correctional systems. There are two areas in which federal government intervention is necessary to solve longstanding deficiencies: by adding to the capacity of community correction beds and by helping usher in new information technology systems that can measure reentry outcomes. Through the careful use of incentives, the federal government can coordinate and manage the sometimes competing and sometimes concurrent needs of the Federal Bureau of Prisons, the state prison systems, and the local jails systems to develop an integrative strategy of prisoner reentry for communities. Without this leadership, much of the promise of the prisoner reentry movement will remain unfulfilled.

Appendix A: Montgomery County (Maryland) Reentry Programs

The Pre-Release and Reentry Services (PRRS) Division of the Montgomery County Department of Correction and Rehabilitation is a work-release residential program that began in 1969 and has served over 16,000 individuals

PRRS operates out of a 177-bed two-story residential community correctional facility that is located close to jobs and transit in Rockville Maryland that has been continuously accredited since 1983 by the American Correctional Association. It also oversees a home confinement program that allows carefully selected individuals - most of whom have first lived and participated successfully in the program at the Center - to complete their sentences in their pre-approved and pre-inspected homes under electronic monitoring. On a given day, the program serves approximately 180 individuals, which comprise almost 30% of the population sentenced to our local correctional system to a sentence of 18 months or less.

The Pre-Release and Reentry Services (PRRS) Division provides residential and non-residential reentry services to convicted and sentenced individuals who are within 12 months of release and who have been incarcerated in the county's correctional system. Additionally, the Division is contracted by the Maryland State Division of Correction and the Federal Bureau of Prisons to serve prisoners in state and federal custody who are within six months of release and who are returning to Montgomery County and the Greater Washington Metropolitan area. The program carefully screens and accepts only those individuals that it assesses can be safely managed in a community setting regardless of offense type with one exception. While the program will accept sex offenders and individuals convicted of violent offenses including homicide, it excludes individuals convicted of prior escape. The Division advances the Department of Correction and Rehabilitation's mission to improve public safety and reduce victimization, and relies on a considerable body of research that demonstrates the cost-benefit advantages of releasing incarcerated individuals through a highly-structured community-based program.

PRRS requires program participants to work, pay room and board, file state and federal taxes, and address restitution and child support obligations. Each client works with several staff members including a case manager to develop an individualized reentry plan that addresses their specific transitional needs including employment, housing, treatment, and medical services. Whenever possible, family members of clients are encouraged to participate in the development of the plan. Additionally, the program holds clients accountable for their location at all times, and clients only access the community with pre-approval. Through the use of the latest technologies in electronic monitoring, substance abuse testing, and by utilizing mobile teams of staff, clients are held to high standards of conduct and compliance. There is a zero- tolerance policy with regard to engaging in criminal activity, using drugs and alcohol, and accessing the community at locations and times that have not been approved. Individuals found in violation of such policies are immediately returned to secure detention. In a given year, the program serves over 700 clients and 85% successfully complete the program. In 2008, they earned collectively $1.9 million, paid over $332,000 in county, state, and federal taxes, paid over $300,000 in program fees, contributed almost $200,000 in child support, and over $10,000 in restitution.

PRRS is but one of four operational divisions of the Montgomery County Department of Correction and Rehabilitation -- one pre-trial division, one central processing unit (CPU) jail, and one longer-term holding detention center -- which work in concert, to carry-out our forty year mission to efficiently and effectively use jail beds and community-based programs to maximize public safety and minimize societal costs. In all aspects of our operations, we are guided by best practices and evidenced-based research.

Working with the courts, prosecutors, and the defense bar, the pre-trial division diverts over 2,300 individuals arrested from our jail back into the community before adjudication. They either remain on pre-trial supervision status during which they are closely monitored and drug tested, or they are sentenced to an alternative drug treatment and community services program that will expunge the offense from their criminal record if they successfully complete the program. In our main jail called the Montgomery County Correctional Facility, we fold library services, educational programs, drug treatment programs, job training, and case management partnerships with community providers into an overall reentry strategy for inmates who are within 90 days of release. Sixty-day ID passes are issued as temporary forms of identification and also serve to allow released individuals to ride the county buses and access the county library system. Additionally, the County's Department of Economic Development and the County's Workforce Investment Board opened up the nation's first One-Stop Career Center located within the secure perimeter of the correctional facility, and staff from the county's workforce agency has worked steadily to place soon-to-be released inmates into jobs.

One of the chief characteristics and greatest strengths of our system is our community and interagency partnerships, and the DOCR has long recognized the vital necessity of reaching out and welcoming other government agencies from law enforcement to human services, to community provides, to faith-based organizations to help us deliver services and assess the risk and needs of our population. In our system, we find that our efforts to develop reentry services add significantly to the community well-being, the improvement of our clients, and well as assisting us with the operational basics of running safe, clean, and orderly correctional institutions for the benefit of our staff. More information about the Montgomery County (Maryland) Department of Correction and Rehabilitation are available on our website: www.montgomerycountymd.gov/cor .

Appendix B: Montgomery County DOCR Recidivism Analysis - Initial Results

Recidivism Rates for DOCR Sentenced Inmates Released 2003/2004
Category One-Year Three-Year
Male Female Male Female
Any Conviction/Arrest all data 21% / 41% 14% / 32% 49% / 66% 35% / 54%
Any Conviction using only MD data 16% 9.2% 36 % 22%
Serious Conviction/Arrest all data 7.4% / 6.4% 6.4% / 5.7% 19% / 17% 15% / 12%
N=294 male inmates & N=282 female inmates released from MCDC, MCCF, or PRRS

Figure 1 - Any Conviction Figure 2 - Serious Conviction

Figure 3 - Any Conviction Figure 4 - Serious Conviction


Figures 1 & 2 show the survival curves for sample of male inmates, which are the cumulative proportion surviving over time from release. At the moment of release from custody, 100% are out of custody. All those who have not been rearrested as of a given time are considered to have "survived." The curve shows what proportion remain arrest free at any given length of time from release. Survival curves provide more information than recidivism rates for a specified time period (i.e. 1year or 3 year) because the curve shows how rapidly the recidivism occurs. In comparing curves for different populations, one compares how much and when the recidivism occurs. Figures 3 & 4 provide an alternate way to view the same phenomenon by plotting hazard rates for the male inmates, which show the rate at which people are recidivating. Hazard rates for recidivism generally rise sharply soon after release and then decline to relatively low levels.