WHAT IS THE VALUE OF PREVENTION AND INTERVENTION FOR TAXPAYERS AND THE PUBLIC?
Chair: Alka Indurkhya
SOCIAL AND MARKET MULTIPLIERS OF THE EFFECTIVENESS OF DRUG PREVENTION. Jonathan Caulkins1, 1Carnegie Mellon University, Pittsburgh, PA United States
There is a large literature evaluating the effectiveness of drug prevention in terms of what might be called intermediate, individual-level outcomes such as, “at three-year follow-up, program participants were 25% less likely to self-report marijuana use as compared to matched controls.” There is less research translating what such findings imply for programs´ role in an overall drug control framework. Previous work (Caulkins et al., 1999, 2002) built bridges between “intermediate” and long-run effectiveness metrics for school-based prevention. The current paper updates and extends some of those results to other forms of prevention.
To clarify, effectiveness findings are typically expressed through metrics such as the percent reduction in self-reported use one or a few years after intervention and for substances commonly used in adolescence, primarily tobacco, alcohol, and marijuana. Four differences between those metrics and others that society also cares about from an economic perspective are: (1) Effects throughout the lifetime, not just over the first few years. (2) Effects on illicit substances that cause the greatest social costs (including cocaine, heron, methamphetamine), not just those used more commonly by adolescents. (3) Indirect spill-over effects on others, as well as effects on people in the program. And (4) Effects on quantities used (days of intoxication, kilograms consumed, etc.), not just prevalence (e.g., number of past-year users).
This paper focuses on spill-overs due to social and market multiplier effects. The social multiplier is estimated by running simulation models of the epidemic spread of a drug in “what if” mode with and without prevented initiation. The change in the (present value of the) subsequent number of initiations is computed. That number (plus one) is the social multiplier.
The market multiplier is the change in consumption per unit change in demand. For a static market model, the multiplier is determined entirely by the shape and slope of the demand and supply curves, but this analysis is embedded within a dynamic framework that reflects the (discounted) accumulation of effects over time both primary (specifically, the difference between long and short-run elasticities of demand) and indirect (dynamic feedback effects).
COST-BENEFIT ANALYSIS OF MULTISYSTEMIC THERAPY WITH SERIOUS AND VIOLENT JUVENILE OFFENDERS. Stephanie Merbler1, Charles Borduin1, Cindy Schaeffer2, 1University of Missouri-Columbia, Columbia, MO United States; 2University of Maryland Baltimore County, Baltimore, MD United States
Treatment of serious and violent juvenile offenders has become an important issue on the nation´s social policy agenda, largely because of the considerable social and economic costs exacted by these youths. Indeed, serious juvenile offenders have the greatest risk of becoming “life-course-persistent” offenders (Moffitt, 1993), costing taxpayers and crime victims $1.5 to $1.8 million over a lifetime of crime (Cohen, 1998). Compounding the staggering costs posed by serious juvenile offenders is the general lack of success that mental health and juvenile justice services have had in ameliorating the serious antisocial behavior of youth (Melton & Pagliocca, 1992; Tate, Reppucci, & Mulvey, 1995). Furthermore, few attempts have been made to incorporate cost-benefit analysis into applied clinical research, which could greatly assist policymakers and program administrators in selecting and implementing mental health programs for serious juvenile offenders (Weisz, Hawley, Pilkonis, Woody, & Follette, 2000). The present study examined the costs and benefits of multisystemic therapy (MST), now widely regarded as a clinically effective treatment for serious antisocial behavior in adolescents (Burns, Hoagwood, & Mrazek, 1999; Farrington & Welsch, 1999; Kazdin, 2000; U.S. Public Health Service, 2001).
Our economic analysis of MST was based on a cost-benefit model developed by the Washington State Institute for Public Policy (2001) and used arrest and incarceration data from a 13.7-year follow-up (Schaeffer & Borduin, 2004) of a randomized clinical trial of MST with serious juvenile offenders (Borduin et al., 1995). Two types of benefits of MST were evaluated: (1) the monetary value to taxpayers was derived from measures of criminal justice system costs and (2) the value to crime victims was derived in terms of both monetary (e.g., medical and mental health care expenses, property damage and losses, reduction in future earnings) and quality of life (e.g., pain, suffering) benefits. The results indicated that reductions in posttreatment arrests (54% fewer) and days incarcerated (59% fewer) for MST participants were associated with substantial reductions in costs to taxpayers ($40,329) and crime victims ($161,529) almost 14 years following treatment.
The economic benefits of MST, as well as its clinical effectiveness, should be considered by policymakers and the public at large in the selection of interventions for serious juvenile offenders. Future research should assess the costs and benefits of MST with other clinical populations of youths (e.g., youths in psychiatric crisis).
PROSPECTIVE ECONOMIC BENEFITS FROM EFFECTIVE PREVENTION FOR FETAL ALCOHOL SPECTRUM DISORDERS. Rick Harwood1, Chuck Lupton2, 1Lewin Group, McLean, VA United States; 2NGIT, Rockville, MD United States
Fetal alcohol spectrum disorders (FASD) are permanent birth defects that result from exposure of a fetus to alcohol due to maternal drinking. Interventions that lead pregnant women to abstain from alcohol can be expected to yield significant economic benefits in terms of reduced expenditures for services and increased quality-adjusted life years. Quantification and valuation of such benefits, combined with sound evidence of the effectiveness of interventions, would be very useful to deliberations about public policy with respect to FASD preventive services.
This analysis will present new estimates of the economic value of preventing one case (an actuarially "expected" case) of an FASD, applying the methodology of the Panel on Cost-Effectiveness in Health and Medicine convened by the U.S. Public Health Service. The analysis will present estimates of expected lifetime savings of direct expenditures and improvements in quality adjusted life years that are expected to result from effective preventive intervention.
FASD is an umbrella term describing the range of effects that can occur in an individual whose mother drank alcohol during pregnancy. These effects may include physical, mental, behavioral, and/or learning disabilities with possible lifelong implications. The term FASD includes conditions such as fetal alcohol syndrome, the leading preventable cause of mental retardation, and fetal alcohol effects. These conditions require a variety of medical, educational, and social services across the expected lifetime of an individual with an FASD. Previous studies have developed estimates of the annual and lifetime cost of providing appropriate services and the loss of productivity value by the human capital method only for persons with fetal alcohol syndrome (a minority of those with FASD). However, the state of knowledge about the nature, severity, and incidence of FASD has significantly increased since the most recent cost estimates were developed. Areas of particular importance include recognition of the high prevalence among the FASD population of severe mental disorders and involvement with the criminal justice system, including incarceration and community supervision. Also, the U.S. Public Health Service Panel recommended that the benefits of health interventions be measured using impacts on quality-adjusted life years. This study will present estimates of expected impacts using this metric.
The results of this study will have particular value to public policy discussions about proposals to increase or improve public health interventions intended to reduce the prevalence of FASD. It is also expected that this study will increase the motivation for public agencies to implement rigorous evaluations of FASD preventive interventions.
PEER INFLUENCE IN ADOLESCENT SUBSTANCE USE. Kenneth Dodge1, 1Duke University, Durham, NC United States
For many adolescents, using drugs is a social event. Their exposure to drugs comes from peers, and their decision-making occurs with reference to peer norms. Association with drug-using peers is the single strongest predictor of drug-use during adolescence. Unfortunately, current prevention programs do not benefit sufficiently from basic-science knowledge of processes in social decision-making, peer influence, and social diffusion. The goal of the research reported in this symposium is to translate knowledge about peer-influence processes from basic social science disciplines to innovative intervention hypotheses. The studies are part of the Duke Transdisciplinary Prevention Research Center (TPRC), funded by NIDA. Three programs of research will be described, corresponding to intrapersonal, interpersonal, and institutional levels of peer influence. The first program addresses processes of intrapersonal social cognition as they relate to drug use. Findings will be reported regarding adolescents' reasons for initiating substance use and the social contexts of their initial use. The second program addresses interpersonal processes of peer influence in small-group interactions and drug-use contexts. Studies will be described that address these questions: 1) What individual characteristics contribute to leadership and influence within adolescent peer groups? and 2) How can adolescent deviant peer-group leaders be utilized to influence other adolescents not to use drugs? The third program addresses institutional peer effects. Institutional design policies at the macro-level, such as academic tracking and the timing of transition to middle school may indirectly influence adolescent drug use by altering opportunities for drug-use diffusion. Studies will be described that utilize administrative data sets with academic and behavioral records of public school children to test the influence of various institutionally-imposed peer contexts on students' deviant behavior, including suspensions from school for possession of illicit substances A final aim of the symposium is to cross-fertilize the products of these three programs of research. This goal will drive the discussion that will conclude this symposium.
THE SOCIAL CONTEXTS OF ADOLESCENT SUBSTANCE USE. Janis Kupersmidt1, Ann Brewster2, Joseph Lewandowski1, 1Innovation Research and Training, Durham, NC United States; 2University of North Carolina at Chapel Hill, Durham, NC United States
The relationship between child and adolescent social functioning and substance use is well documented; however, less is known about the mechanisms by which social contexts affect the onset versus the maintenance of substance use and abuse. The purpose of this study was to collect qualitative data regarding common social contexts that are associated with both initial and ongoing use of alcohol, tobacco, and other drugs. We hypothesized that social contexts associated with peer social acceptance would emerge as most commonly associated with drug use among children and adolescents. We also hypothesized that situations associated with stress relief and self-medication would characterize later use. Six focus groups ranging in size from four to eight participants were conducted (n = 37) including substance abuse counselors, school teachers and health educators, parents of adolescents, young adults who were recovering addicts, middle schoolers and high schoolers. Groups were led by one of two trained adult moderators and each moderator used a Moderator´s Guide that was designed to provide detailed instructions regarding the topics and questions to be followed. Content analysis of focus group transcripts revealed that both parents and peers comprise the majority of the social contexts in which children and adolescents use substances. Somewhat surprisingly, members of all groups repeatedly reported about the strong and influential effect of parents on initial drug use rather than peers. Parents appeared to exert a direct influence on initial use and both a direct and indirect influence on ongoing usage. Members of all focus groups described the role of peers as being important as well, but they appeared to exert influence on the maintenance of adolescent substance use. Interestingly, multiple mechanisms were described associated with the transmission of peer influence as being both through conventional peer pressure (i.e., a pushing effect) as well as through a “rite of passage” to adulthood (i.e., a pulling effect). Analyses of common themes across groups confirmed our hypotheses that social acceptance situations characterized the motive for the majority of initial substance use and that stress relief situations characterized the motive for ongoing use for a subgroup of adolescents. Influences by type of drug, race, ethnicity, gender, socioeconomic status, and developmental level were also examined. Implications of these findings include constructing a developmentally-appropriate taxonomy of substance-use problem contexts or situations for youth for the purpose of developing more responsive prevention interventions related to child and adolescent substance use and abuse.
FACTORS THAT DIFFERENTIATE DEVIANT AND CONVENTIONAL LEADERS OF ADOLESCENT PEER GROUPS:IMPLICATIONS FOR DRUG USE PREVENTION EFFORTS. Philip Costanzo1, Nicole Polanichka1, Jennifer Lansford1, Rebecca Chu1, Kim Arrington1, 1Duke University, Durham, NC United States
Emerging research has indicated that deviant leaders of adolescent groups have a disproportionate influence over their peer associates. This influence is most pernicious when it increases the probability of high-risk behavior, including substance abuse. This program of research is devoted to understanding the dynamic structure of adolescent peer groups and the functions and processes of leadership within them. We have sought to find those factors which characterize deviant adolescent leaders and which distinguish them from more conventional leaders of adolescent groups. This study is based on data collected in a multi-layered survey of 153 seventh grade students. Students were also administered sociometric items for which they nominated their 7th grade peers for a variety of roles, including those of "leader who is good to have in charge" ( conventional leadership) and " someone who can lead others to break rules" ( deviant leadership). With regard to substance use, level of deviant leadership was significantly correlated with earlier drug initiation, drug use, friend's drug activity, and alcohol use. Deviant leadership nominations were also significantly correlated with lower grades as well as indices of reciprocal susceptibility to the negative behavioral influences of best friends. Finally, deviant leadership was correlated with self and peer ratings of being "cool", attractive, strong, able to stand up for self, able to scare others, and able to get others to do things they don't want to do. The only categories for which there were shared significant relations for conventional and deviant leadership was social centrality and the ability to get others to do things they do not want to do. In contrast to deviant leadership, level of conventional leadership was correlated positively with grades, cheerfulness, smartness, sociability, bonding with best friends and the values of ambition and obedience. Conventional leadership was also negatively correlated with the bad behavior of best friends. A follow-up poisson multiple regression indicated that the strongest discriminants of conventional and deviant leaders were patterns of social bonding, the greater susceptibility and social power of deviant leaders, and their clear proclivity to be engaged in substance use. This pattern of findings will be discussed in terms of the importance of incorporating deviant leaders as allies in school-based prevention efforts.
DOES THE TIMING OF THE TRANSITION TO MIDDLE SCHOOL AFFECT SUBSTANCE ABUSE TRAJECTORIES?. Jacob L. Vigdor1, Josh Kinsler1, Elizabeth Glennie1, Peter Arcidiacono1, Edward Norton2, 1Duke University, Durham, NC United States; 2University of North Carolina at Chapel Hill, Chapel Hill, NC United States
Are children exposed to middle school environments at an earlier age more likely to initiate substance use, or exhibit problem behaviors predictive of later substance use? Such an impact might occur, for example, if younger adolescents are more susceptible to potentially deviant influence of older peers. We analyze this question using a comprehensive administrative dataset on school enrollment and reports of school disciplinary infractions for the State of North Carolina. The transition to middle school occurs after fifth grade in some school districts and after sixth grade in others. This variation forms the basis for a "natural experiment" identifying the impact of sixth graders' exposure to older peers. We show that the incidence of problem behavior is significantly higher among sixth graders attending middle schools, relative to rates among sixth graders attending elementary schools. The magnitude of the differential is sizable: students attending middle school in sixth grade have a seven percentage point higher probability of appearing in our disciplinary database. For comparison, the male-female differential is fifteen percentage points. This disparity in rates of problem behavior persists beyond the sixth grade. Seventh and eighth graders who attended middle school in sixth grade display higher rates of problem behavior than their counterparts attending elementary school in sixth grade. We exploit the longitudinal nature of our database to examine further impacts through the high school years. The behavior differential does narrow over time, particularly for certain subgroups of the population. Among non-Hispanic white students, rates of problem behavior nearly converge by eighth grade. Among African-American students, however, effects are highly persistent over time. We also show that fourth and fifth grade students destined to attend different types of school in sixth grade show no significant differences in rates of problem behavior. This finding addresses concerns that sixth grade school assignment policies correlate with uncontrolled student-level risk factors. We also examine the behavior of students who switch schools after sixth grade, to address the concern that incident reporting patterns vary systematically across schools.
CLASSROOMS AS A CONTEXT FOR PREVENTION. Bridget Hamre1, Sara Rimm-Kaufman1, Laura Justice1, 1University of Virginia, Charlottesville, VA United States
Children´s academic trajectories become relatively stable as early as third grade, indicating a need to intervene during preschool and early elementary school years to ensure later academic success. The purpose of this symposium is to introduce classrooms as an important context for prevention. Three classroom-level interventions within a preschool or elementary school context are described, all of which focus on the promotion of literacy, language, and/or social-emotional development and the prevention of school failure.
The first two interventions are rooted in a conceptual framework that emphasizes teachers´ roles in creating classroom environments that are conducive to learning. The first presentation elaborates this conceptual framework and follows with a description of an observational instrument that provides a statistically valid and reliable measure of the quality of the classroom interactions between teachers and children. An intervention called MyTeachingPartner will then be described, which seeks to improve the quality of the instructional and emotional contexts in preschool classrooms through internet-based staff development among pre-kindergarten teachers of children at risk of school failure.
The second presentation describes the Responsive Classroom (RC) approach, an intervention to promote children´s success in school by helping teachers integrate social and academic learning, address discipline problems, and enhance children´s self-regulation. The RC intervention was instituted within Kindergarten through 4th grade classes at three schools, and the effects of the intervention were assessed by comparing academic and social outcomes among children from three comparison schools. Results indicate positive effects of the RC approach on a number of outcomes, as well as moderated effects that suggest this approach has differential effects among children who experience social and economic risks.
Early deficits in language and literacy skills are difficult to reverse and highly associated with later language deficits and reading disabilities. The final presentation describes a theoretical basis for early interventions to prevent children´s language and literacy deficits, and follows with a description of a language-focused curriculum that attempts to address these deficits among four-year-olds who are at risk of school failure. Preliminary findings of the effects of this curriculum on the language and literacy functioning of four-year-olds will be presented by comparing the development of children from classrooms that adopted this curriculum with classrooms that did not.
DAILY INTERACTIONS IN THE CLASSROOM AS A PREVENTION TOOL. Bridget Hamre1, Jason Downer1, Andrew Mashburn1, 1University of Virginia, Charlottesville, VA United States
The goals of this presentation are to: 1) provide a conceptual framework for understanding daily interactions between teachers and students as an important context for prevention efforts; 2) provide reliability and validity data on a classroom observation system which helps identify aspects of the classroom environment that are important to children´s development and amenable to change; and 3) introduce a study currently underway that is testing the efficacy of using these observations as a tool for working with teachers to improve the emotional and instructional quality of their classrooms.
This work is based on the assumption that it is the daily interactions between teachers and their students that are most proximal to students´ academic and social development. These interactions are the mechanism through which teachers are able (or unable) to support children at risk of school failure and thus are an important target for school-based prevention efforts.
The Classroom Assessment Scoring System
The CLASS is a procedure in which an observer watches the interactions of teachers and children and rates the classroom on 9 different dimensions. The CLASS dimensions are based entirely on interactions of teachers and children in the classroom; scoring for any scale is not determined by the presence of materials, the physical environment or safety, or the adoption of a specific curriculum. In several large, national studies, these dimensions have consistently factored into two broad areas, emotional support and instructional supports, which are uniquely associated with children´s social and academic development.
Classroom Observations as a Prevention Tool
MyTeachingPartner (MTP) is a project that seeks to validate the use of classroom observations as a mechanism for improving the quality of classroom interactions between teachers and their students. It is designed to provide, and evaluate the effects of, two types of web-based support for teachers in pre-kindergarten programs. One group of teachers has access to a dynamic website which offers teaching tips and guidance about how to improve classroom practice. Another group has access to this website and also works with MTP consultants who provide direct, individualized, regular, and systematic feedback to teachers in an ongoing partnership aimed at enhancing teachers´ classroom practices, based largely on the CLASS constructs. This consultancy process involves ongoing review of videotape taken in the classroom, teachers´ response to reflective questions about their practice, and i-chats between teachers and consultants every two weeks. Within this presentation, we will provide an overview of this project, highlighting its relevance to broad, school-based prevention efforts.