Simulation Strategies to Reduce Recidivism

Transcription

1 Simulation Strategies to Reduce Recidivism

2

3 Faye S. Taxman April Pattavina Editors Simulation Strategies to Reduce Recidivism Risk Need Responsivity (RNR) Modeling for the Criminal Justice System

4 Editors Faye S. Taxman Department of Criminology, Law, and Society George Mason University Fairfax, VA, USA April Pattavina School of Criminology and Justice Studies University of Massachusetts Lowell Lowell, MA, USA ISBN ISBN (ebook) DOI / Springer New York Heidelberg Dordrecht London Library of Congress Control Number: Springer Science+Business Media New York 2013 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher s location, in its current version, and permission for use must always be obtained from Springer. Permissions for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecution under the respective Copyright Law. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. While the advice and information in this book are believed to be true and accurate at the date of publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Printed on acid-free paper Springer is part of Springer Science+Business Media (

5 Preface Why Do We Need the RNR Simulation Tools? Over the last decade there has been a growing consensus among academics and practitioners about the importance of using evidence-based practices in the field of corrections. The consensus extends to the need to better align existing practices with known offender and program attributes that will generate better outcomes. Embedded in evidence-based practices is that organizations should: (1) use a valid risk and need assessment tool to identify those factors in an individual that are amendable to change; (2) use cognitive-behavioral programs to address these riskneed profiles; (3) administer programs that are high quality; and (4) focus on recidivism patterns. These are essential elements of the Risk-Needs-Responsivity (RNR) model. RNR is well recognized as a contemporary framework for determining the role of programs in correctional environments (institutional and community). In many ways, this model has revived attention to rehabilitation and the importance of good quality programming. The empirical support for the model along with the lack of progress on recidivism reduction results from punitive or incapactative models of punishment has led to a broad base of support for the RNR framework. The RNR framework is more readily accepted but there are three missing components. First, the literature is based on the RNR philosophy that calls for integrating treatment and rehabilitation into the correctional system. What is lacking is an empirical demonstration designed to explore the effects of adopting RNR model on a large scale. Most findings have referred to single study experiments or metaanalysis but have not illustrated the impact on a large scale. Such a demonstration would assess what would happen to recidivism if we took the RNR model to scale that is, if we expanded the use of risk-needs tools, placed people in programs designed to address risk-need factors, and offered high quality programming what would be the impact. This what if analysis is ripe for simulation models because they provide the mathematical and statistical approaches well suited to illustrate the impact of policy options. In other words, they provide additional empirical support to the notion that it is worth changing sentencing and correctional decisions to v

6 vi Preface incorporate RNR principles. Discrete event simulation model makes the point even more clear since it demonstrates how changes over times and events (different decision points) can affect the outcomes from the correctional system. It provides the empirical evidence that it is beneficial to change systems and practices through thoughtful consideration of the options available. Second, the original RNR framework is built on assumptions regarding the relationships among individual criminal justice risk factors, individual needs, and recidivism. The empirical evidence supporting these assumptions is ever-evolving as more research is produced. Of course, not all studies are the same in terms of methodological rigor, measurement of key variables, and generalizability of the findings. But the emerging literature suggests that the original version of the RNR model may need modification as more scientific information becomes available. For example, the first evidence-based practice principle is that high risk offenders should be placed in correctional programming. This follows from the finding that better results are possible for higher-risk offenders. However, this statement does not consider the degree to which individual needs may trump criminal justice risk factors, the relevance of non-criminogenic factors such as mental illness and housing stability that may affect success in the community, and key demographic key factors (such as age and gender) that affect offending patterns. The RNR model should recognize it could be higher number of needs and clinically relevant factors that increase the need for more structured programming. Third, in the real world the RNR process is more complicated than it appears. The complexity is probably one of the biggest factors that affect the likelihood that the model can and will be used in practice. The complexity has to do with making decisions that integrate complicated information about the individual offender as well as understanding the pros and cons of each program or service. In other disciplines, years of training are provided to build these diagnostic skills in the justice system, this type of skill development is not provided. Rather it is assumed that experiential learning ( on the job training ) will suffice. However, the skills to assess complex human condition such as those factors that influence offending behavior, intergenerational substance abuse or criminal behavior, antisocial personality or values, and so on are not necessarily easy to isolate. Answering the question, what should we do about this? requires yet another set of skills. Understanding the impact of one decision on processing at other decision points requires more complex information. The complexity of the tasks supports the need for simulation models that support the decisions that individuals or jurisdictions need to make about who should go into what program? This book evolved from many discussions among scholars over the years that are devoted to improving criminal justice policy and practice through integrating evidence and evidence-decision criteria to practice. Faye Taxman has been working on many of the concepts in this book over the course of her career. The notion of exploring how to improve the use of assessment tools for identifying the factors that should be addressed during the period of correctional control has been a theme in her work on seamless systems of probation, reengineering probation, and using evidence-based practices. This is supplemented by the attention to more intricate

7 Preface vii decisions such as how to respond to negative behaviors and how to increase compliance with the conditions of correctional control. The growing emphasis on implementation processes (see Taxman & Belenko, 2012 on implementation models) has led many in the corrections field to promote the use of assessment tools and models to drive decisions regarding placement in appropriate programs and services. Energized by the evidence in support of the RNR model and interested in the possibilities for widespread adoption, the Bureau of Justice Assistance sought to learn more about the correctional impacts if the model were implemented on a larger scale. With their support, several related projects described in the chapters were undertaken with the purpose of determining the ways in which we could (1) build RNR-based tools to support RNR programming for agencies and (2) to understand how the application of these tools would collectively serve to reduce recidivism at the national level. After the framework for the RNR program simulation tool was developed, the challenge was to determine how the use of this tool across agencies might contribute to significant reductions in recidivism on a national scale. April Pattavina s work in this area was informed by recent developments in computer simulation that would allow for investigation of this issue. Simulation models have recently been used for operations research in health care and criminal justice applications and were appropriate for our work. We seek to first give the reader an understanding of how these techniques may be applied in operational contexts and then use the techniques in our investigation into the recidivism reduction effects of the RNR model. In conjunction with the evolution of more effective models and tools available for assessing offender risks and needs has been the advancement of information and computing technology that has created opportunities to realize the possibilities for these models beyond a single study or agency. Throughout much of this book, authors were able to take advantage of publically available datasets, such as the Survey of Inmates in State Correctional Facilities, the National Corrections Reporting Program, and the Bureau of Justice Statistics classic recidivism study for our work in this book. Using the RNR model as a framework, it was possible to use the data sources to measure key RNR concepts, map the processes for treating offenders, and finally investigate the impact of a large-scale implementation of RNR programming on recidivism. We were also fortunate to have partnership with state and local correctional and substance abuse agencies to validate the tool and assumptions. Using the most current simulation techniques, it was possible to assess the impact over time and found that the RNR model holds considerable promise for reducing recidivism. Simulation models are important tools that are underutilized in research and policy. The attraction of using different types of simulation models was that the issues are often too complicated to design in experiments. This book outlines how to put together a simulation tool, and then use the tool to assess various problems. The how to notion of this book is to help others consider the various steps to develop a simulation model. In the course of developing the simulation tools, we learned how to handle a number of challenging data, methods, and theory issues. These challenges are presented to foster a greater understanding of the mystic

8 viii Preface involved in creating a simulation model, and how to use data from one model to another. We hope this book helps foster a number of simulation models in the field of justice policies. A few years from now we will know whether we were successful in this goal. This book was possible through the support from many people including the contributors. We are indebted to the authors of chapters for their hard work including James Byrne, Stephanie Ainsworth, Erin Crites, Michael Caudy, Joseph Durso, Avi Bhati, Andrew Greasely, Matthew Concannon, and David Hughes. Ed Banks from the Bureau of Justice Assistance offered endless assistance and his passion for helping improve the knowledge integration process inspired us. Thanks are also given to our colleagues that have moved this field forward including Edward Latessa, Todd Clear, and Redonna Chandler. Fairfax, VA, USA Lowell, MA, USA Faye S. Taxman April Pattavina Reference Taxman, F. S., & Belenko, S. (2012). Implementation of evidence-based practices in community corrections and addiction treatment. New York: Springer.

9 Contents Part I Advances in Simulation Modeling for Criminal Justice Planning and Management 1 Planning for the Future of the US Correctional System... 3 April Pattavina and Faye S. Taxman 2 A Case Study in Gaps in Services for Drug- Involved Offenders Faye S. Taxman and Matthew L. Perdoni 3 The Simulation Modelling Process Andrew Greasley 4 The Empirical Basis for the RNR Model with an Updated RNR Conceptual Framework Faye S. Taxman, April Pattavina, Michael S. Caudy, James Byrne, and Joseph Durso Part II Simulation Inputs for an RNR Model 5 Creating Simulation Parameter Inputs with Existing Data Sources: Estimating Offender Risks, Needs, and Recidivism Stephanie A. Ainsworth and Faye S. Taxman 6 The Responsivity Principle: Determining the Appropriate Program and Dosage to Match Risk and Needs Erin L. Crites and Faye S. Taxman 7 Reducing Recidivism Through Correctional Programming: Using Meta-Analysis to Inform the RNR Simulation Tool Michael S. Caudy, Liansheng Tang, Stephanie A. Ainsworth, Jennifer Lerch, and Faye S. Taxman ix

10 x Contents Part III Simulation Applications 8 RNR Simulation Tool: A Synthetic Datasets and Its Uses for Policy Simulations Avinash Bhati, Erin L. Crites, and Faye S. Taxman 9 A Simulation Modeling Approach for Planning and Costing Jail Diversion Programs for Persons with Mental Illness David Hughes 10 Using Discrete-Event Simulation Modeling to Estimate the Impact of RNR Program Implementation on Recidivism Levels April Pattavina and Faye S. Taxman Part IV Conclusion 11 Risk-Need-Responsivity (RNR): Leading Towards Another Generation of the Model Faye S. Taxman, Michael S. Caudy, and April Pattavina Index

11 Contributors Stephanie A. Ainsworth Department of Criminology, Law and Society, George Mason University, Fairfax, VA, USA Avinash Bhati Maxarth, LLC, Gaithersburg, MD, USA James Byrne School of Criminology and Justice Studies, University of Massachusetts Lowell, Lowell, MA, USA Michael S. Caudy Department of Criminology, Law and Society, George Mason University, Fairfax VA, USA Erin L. Crites Department of Criminology, Law and Society, George Mason University, Fairfax, VA, USA Joseph Durso Department of Criminology, Law and Society, George Mason University, Fairfax, VA, USA Andrew Greasley Aston Business School, Birmingham, UK David Hughes Human Services Research Institute, Cambridge, MA, USA Jennifer Lerch Department of Criminology, Law and Society, George Mason University, Fairfax, VA, USA April Pattavina School of Criminology and Justice Studies, University of Massachusetts Lowell, Lowell, MA, USA Matthew L. Perdoni Department of Criminology, Law and Society, George Mason University, Fairfax, VA, USA Liansheng Tang Department of Statistics, George Mason University, Fairfax, VA, USA Faye S. Taxman Department of Criminology, Law, and Society, George Mason University, Fairfax, VA, USA xi