MENU TITLE: National Institute of Justice Journal -Probation in the United States: Practices and Challenges. Series: NIJ Journal Published: September 1997 63 pages 133,269 bytes Figures, charts, forms, and tables are not included in this ASCII plain-text file. To view this document in its entirety, download the Adobe Acrobat graphic file available from this Web site or order a print copy from NCJRS at 800-851-3420. U.S. Department of Justice Office of Justice Programs National Institute of Justice Probation in the United States: Practices and Challenges Also in this Issue: Americans' Views on Crime and Law Enforcement: Survey Findings Who Wrote It? Steps Toward a Science of Authorship Identification ------------------------------ DIRECTOR'S MESSAGE A common thread running through much of this issue's content is its relevance to some of the strategic challenges being addressed by NIJ-- challenges that arise from and build on the evolving body of knowledge generated by research during the past quarter century. For example, Joan Petersilia's lead article on probation practices and challenges pertains, in its larger context, to NIJ's strategic challenge of rethinking justice, which encompasses the task of examining the role of agencies in dispensing justice. Professor Petersilia notes that probation departments are more extensively involved with offenders and their cases than any other justice agency and states that probation officers interact with many criminal justice agencies and affect a wide spectrum of justice-processing decisions. Her analysis of probation practices leads to a number of suggestions to strengthen probation departments so they can effectively execute their justice system role and thereby benefit from the essential ingredient of increased public support. The importance of public support to justice agencies' achieving goals and addressing challenges is an underlying theme of Jean Johnson's article about survey findings on Americans' views on crime and law enforcement. She notes robust public support for police agencies but cautions that such support should not be taken for granted. A critical strategic challenge for NIJ is to help create the tools--especially in the area of technology--that practitioners can use to enhance performance. NIJ Visiting Fellow Carole Chaski--in her article Who Wrote It?--describes her progress toward creating a computer-based system designed to put authorship identification on a scientific footing whether the text is handwritten, typed, or found on a computer disk. Among the many items in this issue's departments is one relating to NIJ's strategic challenge of breaking the cycle of crime and violence: In June of this year, a project--appropriately named Breaking the Cycle and designed by a consortium of Federal agencies, including NIJ--was launched in an effort to break the cycle between drugs and crime. Another department item pertains to rethinking justice--a series of symposiums on restorative justice. Three items relate to "creating the tools": NIJ's Crime Mapping Research Center Visiting Fellowship Program; the upcoming land transportation security technology conference, cosponsored by NIJ and the Department of Transportation, in cooperation with the State Department; and NIJ's first Summer Institute, which focused on making technology viable and effective for law enforcement agencies. Future issues of the National Institute of Justice Journal will continue to report on how the Institute addresses its challenges, such as by transforming its Drug Use Forecasting program into the Arrestee Drug Abuse Monitoring Program and by intensifying research on violence against women. Jeremy Travis Director National Institute of Justice ------------------------------ NIJ Journal Editorial Board Sally T. Hillsman Deputy Director for Office of Research and Evaluation John L. Schwarz Deputy Director for Office of Development and Dissemination David Boyd Director, Office of Science and Technology Cheryl A. Crawford Mary G. Graham Marj Leaming Christy Visher Edwin W. Zedlewski Issue Editor William D. Falcon ------------------------------ The National Institute of Justice Journal is published by the National Institute of Justice, the research arm of the U.S. Department of Justice, to announce the Institute's policy-relevant research results and initiatives. The Attorney General has determined that publication of this periodical is necessary in the transaction of the public business required by law of the Department of Justice. Opinions or points of view expressed in this document are those of the authors and do not necessarily reflect the official position of the U.S. Department of Justice. The National Criminal Justice Reference Service (NCJRS), a centralized national clearinghouse of criminal justice information, is sponsored by the Office of Justice Programs agencies and the Office of National Drug Control Policy. Registered users of NCJRS receive the National Institute of Justice Journal and NCJRS Catalog free. To become a registered user, write NCJRS User Services, Box 6000, Rockville, MD 20849-6000, call 800-851-3420, or e-mail askncjrs@ncjrs.org. The National Institute of Justice is a component of the Office of Justice Programs, which also includes the Bureau of Justice Assistance, Bureau of Justice Statistics, Office of Juvenile Justice and Delin- quency Prevention, and Office for Victims of Crime. ------------------------------ Issue No. 233 Contents Features Probation in the United States: Practices and Challenges Americans' Views on Crime and Law Enforcement: Survey Findings Who Wrote It?: Steps Toward a Science of Authorship Identification Research Preview Civil Protection Orders: Victims' Views on Effectiveness Departments Events NIJ in the Journals Publications NIJ Awards Final Reports Solicitations New & Noteworthy ------------------------------ PROBATION in the UNITED STATES Practices and Challenges by Joan Petersilia This article is adapted from Professor Petersilia's essay in volume 22 of Crime and Justice, edited by Michael Tonry (University of Chicago Press, 1997). Adult probationers in the United States surged to nearly 3.2 million at the end of 1996, up from almost 2 million in 1985 and 1.1 million in 1980.[1] Today they comprise about 58 percent of all adults under correctional supervision.[2] To cope with their workload, probation agencies-- often the target of intense criticism--receive less than 10 percent of State and local government corrections funding.[3] Probation's funding shortfall often results in lax supervision of serious felons, thereby encouraging offender recidivism and reinforcing the public's soft-on- crime image of probation as permissive, uncaring about crime victims, and committed to a rehabilitative ideal that ignores the reality of violent, predatory criminals. This poor public image leaves probation agencies unable to compete effectively for scarce public funds. Although current programs are often seen as inadequate, the concept of probation--begun in 1841 (see "Origin and Evolution of Probation")--has great appeal and much unrealized potential. As one judge noted, "Nothing is wrong with probation. It is the execution of probation that is wrong."[4] Exactly how would one go about reforming probation? Many judges are monitoring probationers more closely, while others are imposing more punitive and meaningful probation sentences. Some jurisdictions have implemented policies and programs designed to overcome the difficult problem of finding jail and prison capacity to punish probation violators. Unfortunately, debating the merits of those and other probation-reform strategies is severely limited because so little is known about current probation practices. Assembling what is known about U.S. probation practices so public policy can be better informed is the main purpose of this article--along with offering suggestions on meeting the challenges facing probation agencies. Probation and modern sentencing practice Probation departments are more extensively involved with offenders and their cases--often starting at arrest--than any other justice agency. Many who are arrested and all who are convicted come into contact with the probation department. Probation officers interact with many criminal justice agencies and significantly affect a wide spectrum of justice processing decisions, including these: o Probation officers, in addition to pretrial service agencies, usually perform personal investigations to determine whether defendants will be released on their own recognizance or bail. o They prepare reports that courts use as the primary source of information to determine whether to divert defendants from formal prosecution. Probation officers supervise diverted offenders and inform courts about whether the diversionary sentence was successfully complied with, thereby influencing the court's decision to proceed or not with formal prosecution. o They prepare presentence reports containing pertinent information about convicted defendants and their crimes. The information is critically important, for research repeatedly indicates that (1) the judge's knowledge of the defendant is usually limited to what is contained in the presentence report, and (2) the probation officer's recommendation for or against prison correlates strongly with the judge's sentence of probation, prison, or a combination thereof. o They supervise offenders sentenced to probation, determine which court-ordered probation conditions[5] to enforce and monitor most closely, decide which violations of conditions to bring to the court's attention, and recommend sanctions. o They affect, through presentence reports, the initial security classification (and eligibility for parole) of offenders sentenced to prison. More than 2,000 probation agencies in the United States[6] carry out those and other responsibilities. The agencies differ in terms of whether they reside within the executive or judicial branch of government, how they fund services, and whether those services are primarily a State or local function. According to one study, 52 percent of staff in the typical probation department are line officers; 48 percent are clerical, support staff, and management.[7] Of line probation officers, only about 17 percent supervise adult felons. The remaining line officers supervise juveniles (half of adult probation agencies have that responsibility) or misdemeanant probationers or prepare presentence reports. Given an estimated 50,000 probation employees in 1994,[8] and given that 23 percent of them (11,500 officers) were supervising about 2.9 million adult probationers, the average caseload that year was 258 adult offenders per line officer. This contrasts with what many believe to be the ideal caseload of 30 adult probationers per line officer. Of course, offenders are not supervised on "average" caseloads. Rather, probation staffs use a variety of risk and needs classification instruments to identify offenders needing more intensive supervision or services. Although risk instruments can identify offenders who are more likely to reoffend, funds are usually insufficient to implement the levels of supervision predicted by classification instruments.[9] Research findings indicate that, across all sites and felony crimes studied, about 20 percent of adult felony probationers were assigned to caseloads requiring no personal contact.[10] Probation funding has long been recognized as woefully inadequate. From the beginning, probation has continually been asked to take on greater numbers of probationers and conduct a greater number of presentence investigations despite stable or declining funding. "Apparently, community supervision has been seen as a kind of elastic resource that could handle whatever numbers of offenders the system required it to."[11] (See "Who Is on Probation?") Does probation work? The most common question asked about probation is, "Does it work?" By "work," most mean whether the person granted probation has refrained from further crime or reduced his or her recidivism--that is, the number of rearrests. Recidivism is currently the primary outcome measure for probation, as it is for all corrections programs. Probationer recidivism. Summaries of probation effectiveness usually report the recidivism rates of felons as if they represented the total adult probation population, instead of 55 percent[12] of it. Failure to make this distinction between felons and misdemeanants is why profoundly different assessments have been offered as to whether probation "works." In reality, there are two stories about probationer recidivism rates. Recidivism rates are low for adults on probation for misdemeanors--data suggest that three-quarters successfully complete their supervision. However, recidivism rates are high for felony probationers, particularly in jurisdictions that use probation extensively, where offenders are serious to begin with, and where supervision is minimal.[13] Recidivism rates vary greatly from place to place, depending on the seriousness of the underlying population characteristics, length of followup, and surveillance provided. A summary of 17 followup studies of adult felony probationers found that felony rearrest rates ranged from 12 to 65 percent.[14] Such wide variation in recidivism is not unexpected, given the wide variability in granting probation and monitoring court-ordered conditions. Despite the desirability of predicting offender recidivism, available data and statistical methods are insufficient to do so very accurately at this time. Other probation outcomes. Another way to examine probation effectiveness is to look at the contribution of those on probation to the overall crime problem. Of all persons arrested and charged with felonies in 1992, 17 percent of them were on probation at the time of their arrest.[15] Practitioners have expressed concern about the use of recidivism as the primary, if not sole, measure of probation's success.[16] The American Probation and Parole Association (APPA), representing U.S. probation officers nationwide, argues that recidivism rates measure just one probation task while ignoring others.[17] APPA has urged its member agencies to collect data on alternative outcomes, such as amount of restitution collected, number of offenders employed, amount of fines/fees collected, hours of community service, number of treatment sessions, percentage of financial obligations collected, enrollment in school, days employed, educational attainment, and number of days drug free. Some probation departments have begun to report such alternative outcome measures to their constituencies and believe this practice is having a positive impact on staff morale, public image, and funding.[18] How can probation be revived? The public has come to understand that not all criminals can be locked up, and so renewed attention is being focused on probation. Policymakers are asking whether probation departments can implement credible and effective community-based sentencing options. No one advocates the abolition of probation, but many call for its reform. But how should that be done? Implement quality programs for appropriate probation target groups. Probation needs first to regain the public's trust as a meaningful, credible sanction. During the past decade, many jurisdictions developed "intermediate sanctions," such as house arrest, electronic monitoring, and intensive supervision, as a response to prison crowding. These programs were designed to be community-based sanctions that were tougher than regular probation but less stringent and expensive than prison.[19] The program models were plausible and could have worked, except for one critical factor: They were usually implemented without creating organizational capacity to ensure compliance with court-ordered conditions. When courts ordered offenders to participate in drug treatment, for example, many probation officers could not ensure compliance because local treatment programs were unavailable.[20] Over time, what was intended as tougher community corrections in most jurisdictions did not materialize, thereby further tarnishing probation's image. Although most judges still report a willingness to use tougher, community-based programs as alternatives to routine probation or prison, most are skeptical that the programs promised "on paper" will be delivered in practice.[21] As a result, some intermediate sanction programs are beginning to fall into disuse.[22] However, some communities invested adequate resources in intermediate sanctions and made the necessary treatment and work programs available to offenders.[23] In programs where offenders received both surveillance (e.g., drug tests) and participated in relevant treatment, recidivism declined 20 to 30 percent.[24] Solid empirical evidence shows that recidivism is reduced by ordering offenders into treatment and requiring them to participate.[25] So, the first order of business must be to allocate sufficient resources so that the designed programs (incorporating both surveillance and treatment) can be implemented. The resources will be forthcoming only if the public believes that the programs are both effective and punitive. Public opinion is often cited by officials as a reason for supporting expanded prison policies. According to officials, the public's "get tough on crime" demands are synonymous with sending more offenders to prison for longer terms.[26] Recent evidence must be publicized showing that many offenders--whose opinions on such matters are critical for deterrence--judge some intermediate sanctions as more punishing than prison.[27] When, for example, nonviolent offenders in Marion County, Oregon, were given the choice of serving a prison term or returning to the community to participate in intensive supervision probation (ISP) programs--which imposed drug testing, mandatory community services, and frequent visits with the probation officer--about one-third chose prison over ISP.[28] Why should anyone prefer imprisonment to remaining in the community, no matter what the conditions? Some have suggested that prison has lost some of its punitive sting and, hence, its ability to scare and deter. One study found that for drug dealers in California, imprisonment confers a certain elevated "homeboy" status, especially for gang members for whom prison and prison gangs can be an alternative site of loyalty.[29] According to the California Youth Authority, inmates steal State-issued prison clothing for the same reason. Wearing it when they return to the community lets everyone know they have done "hard time."[30] It is important to publicize these results, particularly to policymakers who say they are imprisoning such a large number of offenders because of the public's desire to get tough on crime. But it is no longer necessary to equate criminal punishment solely with prison. The balance of sanctions between probation and prison can be shifted, and at some level of intensity and length, intermediate punishments can be the more dreaded penalty. Once probation's political support and organizational capacity are in place, offender groups need to be targeted on the basis of what is known about the effectiveness of various programs. Targeting drug offenders makes the most sense for a number of reasons. Large-scale imprisonment of drug offenders has only recently taken place, and new evidence suggests that the public seems ready to accept different punishment strategies for low- level drug offenders. The public appears to want tougher sentences for drug traffickers and more treatment for addicts-- what legislators have instead given them are long sentences for everyone. Public receptiveness to treatment for addicts is important, because those familiar with delivering treatment say that is where treatment can make the biggest impact. A report by the Institute of Medicine (IOM) of the National Academy of Sciences notes that about one- fifth of the estimated population needing treatment--and two-fifths of those clearly needing it--are under the supervision of the justice system as parolees or probationers. Because the largest single group of serious drug users in any locality comes through the justice system every day, IOM concludes that the justice system is one of the most important gateways to treatment delivery and should be used more effectively. Research has shown that those under corrections supervision stay in treatment longer, thereby increasing positive treatment outcomes.[31] On the one hand, good-quality treatment is not cheap. On the other hand, it is an investment that pays for itself immediately in terms of crime and health costs averted. Researchers in California[32] concluded that treatment was very cost beneficial: For every dollar spent on drug and alcohol treatment, California saved $7 in reduced crime and health care costs. The study found that each day of treatment paid for itself on the day treatment was received, primarily through an avoidance of crime. The level of criminal activity declined by two- thirds from before treatment to after treatment. The greater the length of time spent in treatment, the greater the reduction in crime. Of course, there is much more to reforming the probation system than simply targeting low-level drug offenders for effective treatment, but this would be a start. There also needs to be serious reconsideration of probation's underlying mission, administrative structure, and funding base. And a program of basic research to address some of probation's most pressing problems should be funded. Make probation a priority research topic. Noted below are a few of the questions that would be highly useful for probation research to address. What purpose is served by monitoring and revoking probation for persons committing technical violations, and is the benefit worth the cost? If technical violations identify offenders who are "going bad" and likely to commit crime, time could be well spent uncovering such violations and incarcerating those persons. But if technical violators are simply troublesome, but not criminally dangerous, devoting scarce prison resources to this population may not be warranted. Despite the policy significance of technical violations, little serious research has focused on this issue. As the cost of monitoring and incarcerating technical violators increases, research must examine its crime control significance. Who is in prison, and is there a group of prisoners who, based on crime and prior criminal records, could safely be supervised in the community? Some contend that many, if not most, prisoners are minor property offenders, low-level drug dealers, or technical violators--ideal candidates for community-based alternatives. Others cite data showing that most prisoners are violent recidivists with few prospects for reform. Research examining the characteristics of inmates in different States (by age, criminal record, and substance-abuse history) is necessary to clarify this important debate. Also critical are better followup studies (ideally, using experimental designs) of offenders who have been sentenced to prison as opposed to various forms of community supervision. By tracking similarly-situated offenders who are sentenced differently, researchers will be able to refine recidivism prediction models and begin to estimate more accurately the crime and cost implications of different sentencing models. How do probation departments and other justice agencies influence one another and, together, influence crime? Decisions made in one justice agency have dramatic workload and cost implications for other agencies and for later decisions (such as probation policy on technical violations). To date, these systemic effects have not been well studied but research examining how various policy initiatives affect criminal justice agencies, individually and collectively, is likely to generate many benefits. Conclusion Several steps may be taken to achieve greater crime control over probationers: o Provide adequate financial resources to deliver treatment programs that have been shown to work. o Combine both treatment and surveillance in probation programs and focus them on appropriate offender subgroups. Current evidence suggests that low-level drug offenders are prime candidates for enhanced probation programs. o Work to garner more public support by convincing citizens that probation sanctions are punitive and in the long run cost-effective. o Convince the judiciary that offenders will be held accountable for their behavior. o Give priority to research addressing probation's most pressing problems. Over time, probation will demonstrate its effectiveness, in terms of both reducing the human toll that imprisonment exacts on those incarcerated and reserving scarce resources to ensure that truly violent offenders remain in prison. ------------------------------ Joan Petersilia is Professor of Criminology, Law, and Society at the School of Social Ecology at the University of California, Irvine. ------------------------------ Notes 1. Bureau of Justice Statistics, Probation and Parole Population Reached Almost 3.9 Million Last Year, Press Release, Washington, D.C.: U.S. Department of Justice, Bureau of Justice Statistics, August 14, 1997. See also Bureau of Justice Statistics, Correctional Populations in the United States, 1995, Washington, D.C.: U.S. Department of Justice, Bureau of Justice Statistics, May 1997. 2. Ibid. 3. Petersilia, Joan, "A Crime Control Rationale for Reinvesting in Community Corrections," Prison Journal, 1995, 75(4):479-496. 4. Judge Burton Roberts, administrative judge of the Bronx Supreme and Criminal Courts. Cited in Klein, Andrew R., Alternative Sentencing, Intermediate Sanctions, and Probation, Cincinnati, Ohio: Anderson, 1997:72. 5. The judge's (and probation officer's) required conditions can include standard conditions (reporting to the probation officer and the like), punitive conditions (house arrest, for example) that reflect the seriousness of the crime and increase the burden of probation, and treatment conditions (such as for substance abuse). 6. Abadinsky, Howard, Probation and Parole: Theory and Practice, Englewood Cliffs, New Jersey: Prentice-Hall, 1997. 7. Cunniff, Mark, and Ilene R. Bergsmann, Managing Felons in the Community: An Administrative Profile of Probation, Washington, D.C.: National Association of Criminal Justice Planners, 1990. 8. Camp, George M., and Camille Camp, The Corrections Yearbook 1995: Probation and Parole, South Salem, New York: Criminal Justice Institute, 1995. 9. Jones, Peter R., "Risk Prediction in Criminal Justice," in Choosing Correctional Options That Work: Defining the Demand and Evaluating the Supply, ed. Alan Harland, Thousand Oaks, California: Sage Publications, Inc., 1996. 10. Langan, Patrick A., and Mark A. Cunniff, Recidivism of Felons on Probation, 1986-89, Washington, D.C.: U.S. Department of Justice, Bureau of Justice Statistics, 1992. 11. Clear, Todd, and Anthony A. Braga, "Community Corrections," in Crime, ed. James Q. Wilson and Joan Petersilia, San Francisco, California: Institute for Contemporary Studies, 1995:423. 12. Bureau of Justice Statistics, Nation's Probation and Parole Population Reached Almost 3.9 Million Last Year. 13. Petersilia, Joan, Susan Turner, James Kahan, and Joyce Peterson, Granting Felons Probation: Public Risks and Alternatives, Santa Monica, California: RAND Corporation, R-3186-NIJ, 1985. 14. Geerken, Michael, and Hennessey D. Hayes, "Probation and Parole: Public Risk and the Future of Incarceration Alternatives," Criminology, 1993, 31(4):549-564. 15. Reaves, Brian A., and Pheny Z. Smith, Felony Defendants in Large Urban Counties, 1992, Washington, D.C.: U.S. Department of Justice, Bureau of Justice Statistics, 1995. 16. Boone, Harry N., and Betsy A. Fulton, Results- Driven Management: Implementing Performance-Based Measures in Community Corrections, Lexington, Kentucky: American Probation and Parole Association, 1995. 17. Ibid. 18. Griffin, Margaret, "Hunt County, Texas, Puts Performance-Based Measures to Work," Perspectives, 1996:9-11. 19. Tonry, Michael, and Mary Lynch, "Intermediate Sanctions," in Crime and Justice: A Review of Research, vol. 20, ed. Michael Tonry, Chicago, Illinois: University of Chicago Press, 1996. 20. Petersilia, Joan, and Susan Turner, "Intensive Probation and Parole," in Crime and Justice: A Review of Research, vol. 17, ed. Michael Tonry, Chicago, Illinois: University of Chicago Press, 1993. 21. Sigler, Robert, and David Lamb, "Community- Based Alternatives to Prison: How the Public and Court Personnel View Them," Federal Probation, 1994, 59(2):3-9. 22. Petersilia, "A Crime Control Rationale for Reinvesting in Community Corrections." 23. Klein, Alternative Sentencing, Intermediate Sanctions, and Probation. 24. Petersilia and Turner, "Intensive Probation and Parole." 25. Gendreau, Paul, "The Principles of Effective Intervention with Offenders," in Choosing Correctional Options That Work: Defining the Demand and Evaluating the Supply, ed. Alan Harland, Thousand Oaks, California: Sage Publications, Inc., 1996. 26. Bell, Griffin B., and William J. Bennett, The State of Violent Crime in America, Washington, D.C.: Council on Crime in America, 1996. 27. Crouch, Ben, "Is Incarceration Really Worse? Analysis of Offenders' Preferences for Prisons over Probation," Justice Quarterly, 1993, 10:67-88; Petersilia, Joan, and Elizabeth Piper Deschenes, "Perceptions of Punishment: Inmates and Staff Rank the Severity of Prison versus Intermediate Sanctions," Prison Journal, 1994, 74:306-328; Spelman, William, "The Severity of Intermediate Sanctions," Journal of Research in Crime and Delinquency, 1995, 32:107-135; and Wood, Peter B., and Harold G. Grasmick, "Inmates Rank the Severity of Ten Intermediate Sanctions Compared to Prison," Journal of the Oklahoma Criminal Justice Research Consortium, 1995, 2:30-42. 28. Petersilia and Deschenes, "Perceptions of Punishment: Inmates and Staff Rank the Severity of Prison versus Intermediate Sanctions." 29. Skolnick, Jerome, "Gangs and Crime Old as Time: But Drugs Change Gang Culture," Crime and Delinquency in California, 1989, Sacramento, California: California Department of Justice, Bureau of Criminal Statistics, 1990. 30. Petersilia, Joan, "California's Prison Policy: Causes, Costs, and Consequences," Prison Journal, 1992, 72(1):8-36. 31. Institute of Medicine, Committee for Substance Abuse Coverage Study, "A Study of the Evolution, Effectiveness, and Financing of Public and Private Drug Treatment Systems," in Treating Drug Problems, vol. 1, ed. D.R. Gerstein and H.J. Harwood, Washington, D.C.: National Academy Press, 1990. 32. Gerstein, Dean, R.A. Johnson, H.J. Harwood, D. Fountain, N. Suter, and K. Malloy, Evaluating Recovery Services: The California Drug and Alcohol Treatment Assessment, Sacramento, California: State of California, Department of Alcohol and Drug Programs, 1994. The researchers studied a sample of 1,900 treatment participants, followed them up for as long as 2 years of treatment, and studied participants from all four major treatment modalities (therapeutic communities, social models, outpatient drug-free, and methadone maintenance). ------------------------------ American's Views on Crime and Law Enforcement Survey Findings by Jean Johnson This article updates and summarizes a presentation made at the first of three discussion sessions entitled "Measuring What Matters," held under the auspices of the Policing Research Institute and sponsored jointly by the National Institute of Justice (NIJ) and the Office of Community Oriented Policing Services. Each session brought together about 40 police executives, leading researchers, community leaders, journalists, and government officials to discuss the challenges of assessing police performance. Papers prepared for the sessions will be published in book-length form by NIJ. A synopsis of the first session, "Measuring What Matters--Part One: Measures of Crime, Fear, and Disorder" (NCJ 162205) can be obtained from the National Criminal Justice Reference Service at 800-851-3420. Surveys cited in this article were national in scope, except those conducted by the Empire Foundation and Quinnipiac College (New York City focused). Unless otherwise noted, they were random sample telephone surveys conducted in 1994 or later. Overall, the margin of error of survey findings was about plus or minus 3 percentage points (exceptions noted). This article relied extensively on data from the Roper Center Public Opinion Location Library, operated by the Roper Center for Public Opinion Research at the University of Connecticut. The library, accessible through NEXIS, can provide full-question wording (see exhibits 1, 2, and 3 for some of the questions), complete responses, and, in most cases, demographic breakdowns for the surveys cited in this article. ------------------------------ There are a number of reasons to take a close look at public attitudes about crime and law enforcement. Perhaps foremost is that public safety is a preeminent responsibility of government, involving expenditures by thousands of police agencies for the salaries of hundreds of thousands of sworn and nonsworn personnel. For that reason alone, policymakers seem obligated to take the public's assessment of law enforcement performance seriously. Another reason to examine the public's thinking about criminality and policing is that Americans routinely make decisions that strengthen or hinder the country's ability to fight crime. When citizens choose not to report crimes or press charges or when jurors decide to accept or discount police testimony for any reason other than merit, they profoundly affect the quality of law enforcement and justice. The most obvious action citizens take to affect crime occurs when they elect the governors, mayors, and legislators who shape crime-related policy. Presented below is an analysis of recent public opinion data on crime, the criminal justice system, and the role and effectiveness of the police. In summarizing key findings, this article notes where attitudes can vary sharply between African- Americans and whites. (Unfortunately, most national surveys are not large enough to report with confidence on the views of Hispanics or other minority groups.) An urgent issue even in the best of times Despite falling crime rates and remarkably good news from some of the Nation's large cities, crime remains an urgent issue for most Americans. Crime routinely appears at or near the top of surveys asking Americans to name the most important issues facing the country. Eight in ten Americans say "reducing crime" is a top priority for Congress, with 57 percent giving it the highest possible priority rating (Hart and Teeter Research Companies, December 1996). The public's concerns about crime seem to be somewhat independent of the actual crime rate, a phenomenon that may discourage law enforcement professionals but underscores just how frightening this issue is for most people. Deeply held public fears developed over decades may be slow to dissipate even in the best of circumstances. Public attitudes in New York City, which has experienced dramatic and highly publicized decreases in violent crime, provide a case in point. Polls show a remarkable jump in the New York Police Department's approval rating, which rose from 37 percent in 1992 to 73 percent in 1996 (Empire Foundation, April 1996). Mayor Rudolph Giuliani, former Commissioner William Bratton, and current Commissioner Howard Safir have earned good marks for their efforts in fighting crime (Quinnipiac College, April 1996 and February 1997).[1] Although half of New Yorkers say the city is now safer, 65 percent say they worry about being a victim of crime (Quinnipiac College, February 1997). Many observers have suggested that public fears about crime are driven by media coverage rather than by any real knowledge of crime rates in their area. And 65 percent of Americans themselves say this is true: They get their information about crime from the media (Los Angeles Times, January 1994). However, almost 6 in 10 say their own community has less crime than the country as a whole (Los Angeles Times, January 1994); 8 in 10 say they feel safe in their own community (Los Angeles Times, October 1995). Even in New York City, where 81 percent of residents say crime is a "big problem," only 38 percent say crime is a "big problem" in their own community (Quinnipiac College, February 1997). But people's fears are nevertheless real, and they may be intensified by the conviction of many Americans that the crime problem is getting worse, not better--67 percent think that violent crime in the country is increasing (Louis Harris and Associates, May 1997). Half of Americans think the amount of crime in their own community will be worse in the year 2000 (Yankelovich Partners for Time/CNN, January 1997). Causes of crime: complex and multifaceted Americans identify a wide variety of social, economic, and moral conditions as the causes of crime. Twenty-three percent cite drugs as a chief cause of crime; 22 percent, a lack of parental responsibility or family breakdown; and 11 percent, economic problems and lack of jobs. Other Americans blame declining moral values, TV, movies, or rap music (Los Angeles Times, June 1995). Only 6 percent consider flaws in the criminal justice system or lax law enforcement as actual causes of crime (Los Angeles Times, June 1995), but the public exhibits substantial interest in remedies involving the police and courts. People seem to distinguish between what they see as root causes of crime and what they believe must be done in the near term--catching violent criminals and incarcerating them for long periods of time. Approaches to reducing crime People back a variety of approaches they view as effective ways to reduce crime--some designed to remove dangerous criminals from their neighborhoods, some to prevent youngsters from falling into a life of crime, some to express society's outrage at those who disdain its laws. Sixty-nine percent of Americans want to make owning handguns or assault weapons more difficult. Seventy-one percent want to make greater use of the death penalty (Hart and Teeter Research Companies, December 1996). Public views on reducing crime do not fall neatly into either a liberal or conservative political framework. People consider "mandatory life sentences for three-time felons" and "youth crime prevention programs" equally effective as crime- fighting measures (Los Angeles Times, April 1994). Asked about the best overall approach to reducing crime, 30 percent of surveyed Americans want to emphasize punishment; 18 percent, the causes; and 51 percent, both (Hart and Teeter Research Group, January 1995). Research on incarceration and alternative sentencing by Public Agenda, a nonpartisan research organization, for the Edna McConnell Clark Foundation also strongly suggests that most Americans believe in a mixture of approaches.[2] For youngsters in particular, peoplewant the preventive approach--"stop them before they start, if you can." But for most Americans,the worst possible lesson for young offenders would be to not get caught or to receive the "slap on the wrist" of probation.[3] Indeed, the Public Agenda studies found that the most popular sentence for young offenders is boot camp. Most Americans are convinced that the young person who "gets away with it" is all the more likely to continue a life of crime. Justice not served Opinion research strongly suggests that, for the public, the concept of justice includes both protecting the rights of the accused and redressing wrongs done to victims and society. The vast majority of Americans appears to believe that the balance between these two goals has tipped too far in favor of the accused. Eighty-six percent of Americans say the court system does too much to protect the rights of the accused and not enough to protect the rights of victims (ABC News, February 1994). Only 3 percent of Americans say the courts deal too harshly with criminals; 85 percent say they are not harsh enough (National Opinion Research Center, May 1994). The police: public confidence high but racial differences Putting more police on the streets as an effective way to fight crime is broadly supported. Nine in ten Americans say that increasing the number of police is a very (46 percent) or somewhat (44 percent) effective way to reduce crime (ABC News, November 1994). And, given the general skepticism people feel about many institutions and most of government, Americans voice substantial confidence in law enforcement. Fifty-eight percent say they have a "great deal" or "quite a lot" of confidence in the police; another 30 percent say they have "some" confidence in the police; only a handful (11 percent) express very little or no confidence (The Gallup Organization for CNN/USA Today, April 1995). In a 1995 Gallup survey, only one major American institution rated higher than the police: 64 percent of the public have a great deal or quite a lot of confidence in the military. The police score about as well as "organized religion" (57 percent), and many groups--business corporations, Congress, the news media--do much worse. The police also score significantly higher than the criminal justice system as a whole; only one in five Americans voices strong confidence in it (The Gallup Organization, 1995). (See exhibit 1.) But confidence in law enforcement is an area where African-Americans and white Americans differ dramatically. While 63 percent of whites say they have a great deal or quite a lot of confidence in the police, only 26 percent of African-Americans feel the same way. Perhaps even more important, while only a handful of whites (8 percent) say they have very little or no confidence in the police, 35 percent of blacks make this statement (The Gallup Organization, April 1995). Incidents that shape perceptions Much of the recent opinion research on police bias and brutality has focused on two widely publicized incidents in the last several years: the trial of four Los Angeles police officers in the beating of Rodney King and the role of now-retired Los Angeles detective Mark Fuhrman in the murder trial of O.J. Simpson. Public attitudes about these two incidents suggest the basis for some of the public's thinking about what constitutes appropriate police behavior and the degree to which people believe most officers act professionally most of the time. Surveys conducted during periods of extensive press coverage and heightened public debate can, of course, show levels of concern or anger that recede in quieter times. Mark Fuhrman, for example, has made numerous media appearances in the wake of the civil judgment against O.J. Simpson, and public attitudes about him personally may shift somewhat with time. But the initial public reaction to these two incidents as people understood them at the time was revealing. Surveys of public reaction to the Rodney King beating--undoubtedly shaped by repeated broadcast of a videotape of the incident--showed that the overwhelming majority of Americans did not like what they saw. Just 6 percent of Americans surveyed after the officers' initial acquittal said they thought the verdict was "right" (CBS News/New York Times, May 1992). Only 9 percent said they "sympathize[d]" more with police than the beating victim (Yankelovich Clancy Schulman for Time/CNN, April 1992.) Reactions to the tape-recorded comments of Mark Fuhrman played during the Simpson criminal trial show a similar public recoil against an officer who did not seem to fit commonly held standards for appropriate police behavior. At the time, 87 percent of Americans, with blacks and whites agreeing in roughly equal numbers, said they had an "unfavorable impression" of Fuhrman (The Gallup Organization, October 1995), although Americans were split largely along racial lines about whether he actually planted evidence in the Simpson case (CBS News, September 1995). Regardless of differing perceptions by blacks and whites about what Fuhrman actually did or did not do, only 9 percent of either group said that watching the Simpson trial gave them more confidence that "police officers perform their duties in a professional and ethical manner" (The Gallup Organization for CNN/USA Today, October 1995). The exception or the rule? For many white Americans, these kinds of incidents are mainly viewed as regrettable exceptions to the rule, a belief not shared by a majority or near majority of surveyed blacks. As noted in exhibit 2: o Only 15 percent of white Americans, compared to 53 percent of blacks, think that "the kind of improper behavior by police described on the Fuhrman tapes (racism and falsification of evidence)" is common among their local police (Princeton Survey Research Associates, August 1995). o Twenty-four percent of surveyed whites, compared to 43 percent of blacks, said "police overreaction to crime" is a very serious threat (The Gallup Organization for America's Talking, June 1994). o Seventy-six percent of surveyed whites, compared to 47 percent of blacks, say police in their community "mostly" treat the races equally (The Gallup Organization for CNN/USA Today, September 1995). Concern among African-Americans about their chances of being treated fairly extends beyond law enforcement: 61 percent of whites, compared to 19 percent of African- Americans, say that racial and other minorities receive equal treatment in the criminal justice system (ABC News, February 1997). Common standards, different experiences Interestingly, blacks and whites are in substantial agreement about what constitutes appropriate police behavior. Nine in ten Americans--with no significant differences between blacks and whites-- disapprove of an officer striking a citizen who is being vulgar and obscene. A roughly equal number (92 percent) disapprove of an officer striking a murder suspect during questioning, with no significant differences between blacks and whites. Ninety-three percent say a police officer should be allowed to strike a citizen who is attacking the officer with his fists, with blacks and whites again in agreement (General Social Survey, 1994). However, when the behavioral setting is more problematic, important differences emerge in the views of surveyed blacks and whites (see exhibit 3). Judgments differ widely about what actually happens in most communities regarding police behavior. Middle-class whites generally have only positive interactions with the police, and most experience a sense of relief at seeing police officers out and about. In contrast, a study by the Joint Center for Political and Economic Studies (April 1996) reports that 43 percent of blacks consider "police brutality and harassment of African-Americans a serious problem" in their own community. Although blacks and whites disagree about the prevalence of police bias and brutality, neither group finds such behavior acceptable. Both blacks and whites disapproved of the Rodney King beating, at least as they saw it. Both groups were repulsed by the attitudes and behavior initially associated with Mark Fuhrman. Americans of both races seem dubious that police departments will act forcefully to address problems of racism, dishonesty, or brutality to the extent that they exist in police ranks. Only 14 percent of white Americans and 15 percent of black Americans think it is "very likely" that the controversy surrounding detective Fuhrman will lead to "significant improvement in the way police in this country treat blacks" (The Gallup Organization for CNN/USA Today, October 1995). A fault line in public support In a decade when many Americans seem to think that "government" can do no right, law enforcement is viewed as an essential public service and the police enjoy a robust vote of confidence from most of the public. But support for law enforcement has a fault line. Opinion research surveys suggest that many black Americans are disaffected and suspicious. They are not confident that the police will be fair. They are not confident that the police will be professional. They are not confident that the police will "protect and serve." And while the personal encounters most whites have with police officers may be positive, white Americans have witnessed some graphic, highly publicized examples of police behavior that, in their view, is entirely unacceptable. They may regard these incidents as exceptions but not ones to be glossed over as "the cost of doing business." Public Agenda has looked closely at public attitudes about teachers, another group of government workers whom the public likes. Both teachers and police officers are seen as performing an essential public service and are generally regarded with respect. But Public Agenda research also shows a rising frustration with teachers and their unions for seeming to tolerate and protect the few incompetents among them. Focus groups erupt in anger when discussion turns to teacher tenure. The stories pour out about the one bad teacher the school cannot seem to get rid of. Anger against the few infects attitudes about teachers overall. Law enforcement may now be in a similar position. Police departments that are seen as tolerating racist, brutal, or corrupt officers or police unions perceived as protecting them could slowly and incrementally jeopardize the overall strong support for law enforcement. It is fair to ask how long police departments can tolerate widespread lack of confidence among the black community--an outlook that must daily undermine police effectiveness in fighting crime. Public confidence in law enforcement is, for the country and for law enforcement itself, a priceless asset, but it is neither indestructible nor a cause for complacency. ------------------------------ Jean Johnson is senior vice president at Public Agenda, a nonprofit, nonpartisan research group located in New York City. ------------------------------ Notes 1. Quinnipiac College's April 1996 poll included 741 respondents, with a margin of error of +/- 3.6 percent. The February 1997 poll included 845 respondents, with a margin of error of +/- 3.4 percent. 2. Public Agenda has conducted three studies on public attitudes about incarceration and alternative sentencing in Pennsylvania (1993), Delaware (1991), and Alabama (1989). The research was sponsored by the Edna McConnell Clark Foundation. 3. Readers are referred to Joan Petersilia's article in this issue for data showing that many probationers are largely unsupervised. She notes that when given the choice between incarceration and closely supervised probation, many offenders choose incarceration because they view closely supervised probation as too restrictive. In addition, in some communities "doing hard time" carries status rather than stigma. ------------------------------ Who Wrote It? Steps Toward a Science of Authorship Identification by Carole E. Chaski A forged check, a ransom note, a farewell letter printed at a university computer lab, a threatening letter, a diary of crimes, motel receipts, suicide notes on computer disks--such documents create paper trails leading to suspects. The authorship of documents has played an important role in the investigation of such recent high-profile cases as the Una-bomber, the Oklahoma City bombing, the World Trade Center bombing, and the murder of JonBenet Ramsey. From an investigative standpoint, tools eliminating or furnishing suspects on the basis of authorship are invaluable (see "Investigative Uses of Language-Based Author Identification"). Once a case goes to court, an investigator or expert witness must be prepared to show that the paper trail from document to author was constructed in a manner that justifies its admissibility as testimony and that enhances trust and acceptance by judge and jury. But how can one be certain about a document's origin? Is the document authentic or forged? This article outlines methods used to address those questions and offers preliminary insights into what approaches and techniques must be developed to keep pace (or catch up) with authorship-related technology (e.g., computer disks) and with court- developed criteria affecting the admissibility of, and weight given to, testimony pertaining to authorship identification. From eyewitness to questioned document examiner Legal proceedings have long required some kind of authentication of documentary evidence. Until the 19th century, a document was typically authenticated by an eyewitness to its creation or signing (much as notaries do)--or by someone, such as a spouse, other close relative, or banker, who knew and recognized the writing. This kind of document authentication is still admissible testimony under the Federal Rules of Evidence (specifically Rule 901, "Requirement of Authentication or Identification"). But what if no such person is available or willing to testify in that regard? In the early 19th century, another kind of document authentication came on the scene: comparison by an expert of the questioned document with known writing samples. Through the 1800s and into the early 1900s, most State courts allowed such expert testimony on the basis of court decision or State statute. In 1913, the United States Code permitted the admissibility of handwriting identification by expert testimony. Despite the U.S. Supreme Court's 1923 ruling in Frye v. United States that scientific testimony should be limited by its "general acceptance" within the scientific community, questioned document examination (QDE) was, in a sense, immune from such scrutiny since any lawyer seeking to introduce it in court could argue that it was directly admissible by statute. When the Federal Rules of Evidence (FRE) were enacted by Congress in 1975, testimony based on "comparison by trier or expert witness"--that is, handwriting identification by a questioned document examiner--continued to be admissible through Rule 901. In 1993, 70 years after Frye, the U.S. Supreme Court heard Daubert v. Merrell Dow Pharmaceuticals, Inc. (113 S. Ct. 2786), which resulted in major changes in the way that expert testimony is admitted as scientific evidence. The Court ruled that testimony, if it is to be considered "scientific," must be demonstrated to have characteristics shared by established sciences, like biology or chemistry. The Court listed some of those characteristics: empirical testing, known or potential rate of error, standard procedures for performing a technique, peer review and publication, as well as general acceptance in the scientific community. Key criteria distinguishing scientific endeavors from others are stepwise procedures, identifiable (or discrete) units, measurement, replication, and predictability. The laboratory experiments conducted in high school embody those characteristics. So does cooking: o A recipe is a procedure with steps that must be performed in a stated order. o The ingredients in a recipe are separate, distinguishable items. o The ingredients or units are measured using standard measuring devices. o Anyone can repeat these procedures using the standard measures and standard tools. o Anyone repeating the recipe in the same way with the same units, measures, and tools will get the same result. In a Daubert hearing--one held to determine whether evidence can be admitted as scientific--Professor Barry Scheck attacked questioned document examination on the basis of its scientific foundation (United States v. Starzecpyzel, 1995). District Judge McKenna agreed: "The Daubert hearing established that forensic document examination, which clothes itself with the trappings of science, does not rest on carefully articulated postulates, does not employ rigorous methodology, and has not convincingly documented the accuracy of its determinations." Judge McKenna ruled that handwriting identification testimony could not be considered scientific and so would not fall under Daubert criteria for admitting scientific evidence. This ruling, however, would still admit handwriting identification testimony as long as it was perceived as "technical or other specialized knowledge" (as specified admissible by FRE Rule 702, "Testimony by Experts") rather than as science. Judge McKenna's ruling initiated a series of judicial decisions from Federal, State, and military courts challenging the admissibility of handwriting identification by expert comparison in various ways. For example, at the trial of Timothy McVeigh for the Oklahoma City bombing, after the defense requested a Daubert hearing on the proffered questioned document testimony, prosecutors withdrew such testimony and relied on other means of authenticating motel receipts, identification cards, and other documentary evidence. The choices available appear to be not presenting handwriting identification testimony at all, presenting QDE as "technical" rather than scientific knowledge, or building a foundation for handwriting identification that meets Daubert criteria for scientific testimony. Thus, questioned document examination is facing a steep legal and intellectual challenge to create an authentic science of document authentication via handwriting. Toward a science of authorship: handwriting identification In July 1996, the National Institute of Justice sponsored a workshop on developing a research agenda for questioned document examination. The workshop included questioned document examiners working in local, State, and Federal law enforcement agencies; forensic linguists; attorneys; and experts in voice identification, computer engineering, neural networks, neuroscience, and statistics. Workshop participants discussed this question: What would be needed to argue successfully in a Daubert hearing that QDE testimony is accurate, reliable, and based on sound, empirically tested principles? The fundamental requirement, workshop participants agreed, is empirical evidence to support the two central principles underlying handwriting identification: o Each person's handwriting differs from any other's. Because this requires a comparison between documents from different people, the difference between individuals' writing is called interwriter variation. o Each person's handwriting contains some variation. Because this requires a comparison between different documents from the same person, the differences within one person's handwriting samples are called intrawriter variation. The workshop reached a consensus that proof of the foregoing principles is the highest priority for research. For handwriting identification to be within the realm of the possible, a distinct difference must be detectable between interwriter variation and intrawriter variation so that, even if one's handwriting changes from document to document, it is still identifiably different from another's. Writing samples from one individual may differ, but an individual writing pattern exists in all of them. If these principles are not proved true, the current approach of QDE should stop because testimony could be based on a false belief about the human behavior of writing and have very serious legal consequences. Thus, a prime scientific challenge to QDE is to demonstrate that its principles and methods meet or exceed the Daubert criteria for scientific evidence, such as by taking the following steps: o Create a database by collecting samples from writers. o Determine how to measure the samples. What characteristics should be measured (slant, height, etc.)? Which units of measure should be counted? What procedures should be established for measuring characteristics? o Analyze statistically the results of quantifying handwriting samples. Replication would test the hypothesis that handwriting is individually identifiable and validate the predictive ability of QDE to identify and differentiate handwriting samples. The next step would be to determine how to apply these principles of interwriter and intrawriter variation--if proved true--to actual cases. Applying scientific criteria (stepwise procedures, identifiable units, measurement, replication, and predictability) would create a standard protocol for performing QDE, much like existing protocols for conducting DNA analyses. But it would not be sufficient just to have a standard protocol that everyone agrees on (which itself may be difficult to obtain). The standard protocol must be tested to determine if it actually works and the rate at which it works correctly. The development and testing of a standard protocol is essential to demonstrating that handwriting identification is reliable. Reliability means that using the protocol on the same set of documents at different times under different circumstances--even by different people--will yield the same results. Once a standard protocol for performing handwriting identification is tested and shown reliable, it can be taught as part of training for QDE. At this point, yet another scientific challenge to QDE arises: the question of proficiency among examiners trained to use the standard protocol. Until a standard protocol is developed and practiced by questioned document examiners, proficiency tests may measure luck, visual acuity, persistence, or an odd talent, but not the ability to apply a scientific method to solve a problem in handwriting identification. (See "Recent Developments in Handwriting Identification.") From pen and pencil to electronic texts Meanwhile, what kind of authentication is available if a document is not handwritten? Society is rapidly moving beyond pen and pencil and producing more and more electronic documents. Documents composed on the computer, printed over networks, faxed over telephone lines, or simply stored in electronic memory defy traditional handwriting identification techniques. When authorship of an electronically produced document is disputed, analysis of handwriting and typing does not apply. In the case of networked printers--to which thousands of potential users have access--even ink, paper, and printer identification cannot narrow the range of suspects or produce a solitary identification. The language of a document, however, is independent of whether a document is handwritten or printed or faxed or stored electronically. In the past 20 years, techniques of language-based authorship identification have been developing within university departments--classics, English, and applied linguistics--in Great Britain, Germany, Australia, and the United States. But if these methods of language-based document authentication were offered as scientific evidence in a criminal trial, one would expect a Daubert hearing to conclude in much the same way Judge McKenna ruled with regard to handwriting identification.[1] Even without such a ruling, language-based author identification needs to develop a sound scientific method if it is to serve justice and truth. Foundation for scientific linguistic identification A good starting point for developing a scientific method of linguistic identification is to ask whether a theoretical foundation for language-based author identification exists within linguistic science. Theoretical foundations for language-based author identification derive from various branches of linguistics. The following concepts are well founded and uncontroversial among linguists: dialect and idiolect, language processing, and metalinguistic awareness. Dialect and idiolect. Dialect is a variation within a language, while an idiolect is an individual variation within a dialect. For instance, a speaker of the Southern American English dialect may say "John might should check his parking meter," while a speaker of Northern American English dialect would say "Maybe John should check his parking meter." Because dialect is a group phenomenon, one might predict that Southerners who use "might should" are a fairly homogeneous group. But this is not necessarily so. One Southerner may use "might should" in statements only, while another may produce "might should" in statements as well as in questions such as "Should you might check the meter?" Even though Southerners can be identified by their Southern American English dialect, they can be individuated by their idiosyncratic uses of the dialect, that is, by idiolect. The theoretical notions of language, dialect, and idiolect suggest that individual identity in language is feasible. But can we control and manipulate idiolectal features, or is language use unconscious enough to be reliably indicative of a specific person? To answer, from a theoretical perspective, one must consider language processing. Language processing. In normal language processing, we communicate so quickly that we can finish each other's sentences. Typically we do not even remember the exact words we used but do remember the gist of a conversation long after it has concluded. The form (or syntax) is disposable, while the message (or semantics) is durable. In normal language processing, the construction of the form is automatic or unconscious. Automatic or unconscious control of language enables us to do all we are doing while we are speaking (selecting and retrieving words, combining words into phrases and larger units, attaching phrases to other phrases and larger units) while we as communicators focus consciously only on the meaning of our message. From the perspective of automatized processing, linguistic production (especially syntactic structures) would appear to be very difficult to control. The more automatic a behavior, the more reliably it indicates a personal identity. Fingerprints are reliable indicators of individuality because, normally, we do not control them. Likewise, syntactic structures may be so automatic as to be reliable indicators of individuality. Metalinguistic awareness. Although unconscious control of language is normal behavior, we can distance ourselves from language and make conscious commentary about it. This ability to think consciously and talk about language itself (rather than the message language conveys) is called metalinguistic awareness. The possibility of metalinguistic awareness raises the question of linguistic disguise. Suppose someone is so sensitive to language, so metalinguistically aware, that such a person can overcome the automatization of language processing and actually change natural patterns to such a degree as to imitate the idiolectal patterns of another speaker and suppress one's own idiolectal patterns. Research has already shown that adults vary in their metalinguistic abilities.[2] Therefore, until more is known about metalinguistic awareness in adults, the theoretical position should be taken that linguistic disguise is possible depending on the author's particular level of metalinguistic awareness. So, because the notion of individual identity in language is credible, language-based author identification is theoretically feasible. Because some kinds of linguistic production, especially syntactic processing, are unconscious, detection of authorship through reliable linguistic patterns is also feasible. Because individual awareness of and sensitivity to language varies, an individual may be able to manipulate linguistic patterns; thus, disguising authorship is also theoretically possible. Basic steps toward a science of linguistic identification Since language-based author identification seems theoretically feasible, a scientific method that results in an identification would require analytical approaches and technologies--some taken directly from linguistic theory, others to be developed for this application of the theory-- enabling repetitions of a stepwise procedure to yield consistent results that could be tested statistically. Standard procedures for analyzing a document into syntactic structures--noted above as difficult to manipulate or disguise--are already available from theoretical linguistics. Having the analytical method at the ready leads to the question, What sort of documents should be assembled on which to apply the standard analytical procedure? A database of documents should be assembled with the principles of linguistic performance in mind. If we compare two different document types, say a business letter and a diary entry, some differences in the linguistic patterns will be at least in part due to the differing social context and communicative goals of business letters and diary entries. A person's idiolect will vary depending on the document type being produced. Further, since this database is being constructed for forensic application, the document types should be similar to the document types found in actual cases-- similar, rather than same, because one would not request, for example, actual suicide notes or actual threatening letters or actual ransom notes from human subjects. So the database should contain several writing samples from each writer, and these writing samples should be similar to document types found in actual forensic cases. Once the document database is assembled and each document is analyzed into its syntactic structures, the next step is to examine writing samples for idiolectal markers--those fragments of syntactic structures that serve to identify idiolects, or individual language patterns. DNA typing provides an interesting analogy. We share an enormous amount of DNA; only a minute fraction distinguishes us as individuals. So, analogous to DNA, the bulk of linguistic patterns are shared, and the minor quantitative variation is where we need to look for idiolectal markers. There is no list of idiolectal markers available from linguistic theory. Thus, we need standard operating procedures for how to determine idiolectal markers in documents of varying size, type, and authorship. As with QDE, when evidence of the idiolect and standard operating procedures are developed for performing language-based author identification, proficiency testing of forensic linguists can be designed and conducted on a regular basis. Progress in scientific linguistic identification An NIJ Visiting Fellowship provided the opportunity for developing a computer system--Automated Linguistic Authentication System (ALAS), which has two main components (see exhibit 1). The first is a database of documents (see "The Writing Sample Database"). The second embodies natural language parsing programs that "process" documents in the writing sample database by assigning syntactic labels to the words, phrases, and larger units of each text, which can then be quantified and used statistically to categorize texts into authorship clusters. Currently, ALAS is being used to analyze writing samples from a small subset of subjects to search for idiolectal markers--that is, syntactic structures or combinations of syntactic features that can both discriminate between documents authored by different writers and group together documents written by the same person. Texts of varying lengths were examined to determine which markers are feasible to use depending on the amount of text available, taking into account the special problem of very short texts typical of forensic cases. In a pilot study, ALAS parsed and computed the syntactic distribution in writing samples from two subjects. These subjects, known as 016 and 080, are both women, white, in their 40s, with 2 to 3 years of college education and similar dialectal backgrounds. Since these sociological features may affect linguistic performance, one would expect that the two women might be similar--perhaps indistinguishable--linguistically. Indeed, in response to the first writing task, both discussed their fear of dying and leaving their children motherless. ALAS analyzed three samples of 343, 557, and 405 words from subject 016, and one sample of 240 words from subject 080. The first task was to discover whether any linguistic features were constant through different samples of one subject's writing so that such features could be used to identify the subject's writing. One linguistic feature exemplifying an "identifying ability" is the number of combinations per sentence. The combinations tested here were both clauses per sentence and phrases per sentence. On average, writer 016 produced 5.4 clauses per sentence in text 1, 5.13 clauses per sentence in text 2, and 3.75 clauses per sentence in text 5-- and 23.5 phrases per sentence in text 1, 24.85 phrases per sentence in text 2, and 22.33 phrases per sentence in text 5. When the measures of these combinations for the three writing samples are examined statistically, one may conclude that the three texts were written by the same person. The second task was to find out whether any linguistic features could discriminate between different writers. For this purpose, three writing samples were compared from 016 with the 240-word sample from subject 080. One linguistic feature exemplifying a "differentiating ability" is the variant structures of prepositional phrases. The most frequent form of the prepositional phrase is a preposition followed by a noun phrase. Variant forms of the prepositional phrase include a "stranded" preposition, as in "what are you up to?," or prepositions followed by verb phrases, as in "tired of living a lie." When computing the ratio between the most frequent form of the prepositional phrase and the variant forms, one finds that writer 016's ratios are 18:1, 38:7, and 38:3 for her three samples. Writer 080's ratio is 11:5 for her one sample. Statistically, the chance of these documents coming from the same source is 2 percent; put another way, the chance that these documents were written by different authors is 98 percent, which, of course, is the case. But an idiolectal marker must serve both to identify and to differentiate. So a third task is to test whether any linguistic features had both "identifying" and "differentiating" ability and so could serve as idiolectal markers. For this purpose, measures of the feature differentiating subject 080 from subject 016--ratio of types of prepositional phrases--were compared with the measures of this same feature within the writing samples of subject 016. The feature differentiated between 016 and 080. Would it also have an identifying ability for 016? Statistical testing of the three samples of subject 016 did not detect statistically significant differences among the three samples' ratios of prepositional phrase types. This suggests that the samples were written by the same person. Thus, the ratio of prepositional phrase types serves both to discriminate between the writing samples of 080 and 016, and to cluster together the writing samples of 016. An exciting finding of this pilot study is that idiolectal markers were found in writing samples that are well under 1,000 words in length. Because such small writing samples can be used for syntactic analysis, the method is forensically applicable--in contrast to other techniques, which require much longer texts. The pilot study indicates that language-based author identification may be possible with samples in sizes actually found in kidnaping, homicide, and libel cases. Much more work, however, is needed. At this point, the essential conclusion this and other pilot studies demonstrate is that a syntactic method of analysis, which is grounded in linguistic theory and implemented within a computer program, may be the route to an authentic science of language-based author identification. ------------------------------ Carole E. Chaski, Ph.D., is an NIJ Visiting Fellow. After earning her doctorate in linguistics at Brown University, she taught syntax and computational linguistics at the University of South Carolina and North Carolina State University. She has consulted for law enforcement agencies since 1992. ------------------------------ Notes 1. In technical report 95-IJ-CX-0012-01, A Daubert- Inspired Assessment of Currently Available Language-Based Methods of Authorship Identification, I show that previous methods of language-based identification either violate well- established principles of linguistic theory or do not flow from linguistics but from literary studies and thus lack a scientific method. In their reviews, Crystal and Goutsos found similar problems with McMenamin's work (see Crystal, David, "Review of Forensic Stylistics by Gerald R. McMenamin," Language, 71 (1995):2, 381-385; and Goutsos, Dionysis, "Review Article: Forensic Stylistics," Forensic Linguistics, 2(1995):1, 99-113). On the issue of replicability, Tiersma and Finegan report problems in previous methods, including lack of publication, lack of peer review, and nonreplicated results (see Tiersma, Peter M., "Linguistic Issues in the Law," Language, 69(1993): 1, 113-137; and Finegan, Edward, "Variation in Linguists' Analyses of Author Identification," American Speech, Winter 1990:334-340.) Hardcastle was unable to replicate results using Morton's CUSUM method (see Hardcastle, R.A., "Forensic Linguistics: An Assessment of the CUSUM Method for the Determination of Authorship," Journal of the Forensic Science Society, 33(1993):2, 95-106). At this point, language-based authorship identification would fail a Daubert hearing. 2. Metalinguistic ability in adults is related to literacy level (Chaski, Carole E., and Randall Engle, Cognitive and Metalinguistic Characteristics of Adult Illiterates, Technical Report, State of South Carolina Commission on Higher Education, 1990; Chaski, C.E., "Segmental Manipulation and Metalinguistic Ability in Adult Literates and Pre- literates," Linguistic Society of America Annual Winter Meeting, Philadelphia, Pennsylvania: 1991) and also to training in disciplines related to language (Davis, Hayley, "Ordinary People's Philosophy: Comparing Lay and Professional Metalinguistic Knowledge," Language Sciences, 19(1997):1, 33-46). ------------------------------ Research Preview Civil Protection Orders: Victims' Views on Effectiveness Summary of a Research Study by Susan L. Keilitz, Courtenay Davis, Hillery S. Efkeman, Carol Flango, and Paula L. Hannaford of the National Center for State Courts Domestic violence has moved into the spotlight in public debate in this country, particularly with the 1994 passage of the Violence Against Women Act. After years of considering domestic violence a "family matter," the criminal justice, legal, and medical communities are now collaborating to protect women and children from abusers. Previous research has shown that the effectiveness of civil protection orders for victims of family violence depends on how specific and comprehensive the orders are and how well they are enforced. Recent National Institute of Justice (NIJ)-sponsored research, conducted by the National Center for State Courts (NCSC) and involving interviews with women who filed protection orders, concluded that victims' views on the effectiveness of protection orders vary with how accessible the courts are for victims and how well established the links are between public and private services and support resources for victims. In addition, violations of the protection order increase and reported effectiveness decreases as the criminal record of the abuser becomes more serious. In the majority of cases, victims felt that civil protection orders protected them against repeated incidents of physical and psychological abuse and were valuable in helping them regain a sense of well-being. A protection order alone, however, was not as likely to be effective against abusers with a history of violent offenses; women in these cases were more likely to report a greater number of problems with violations of the protection order. The researchers noted that criminal prosecution of these individuals may be required to curb such behavior. The study confirmed previous research showing a strong correlation between the severity and duration of abuse--the longer women experience abuse, the more intense the behavior is likely to become and the more likely women are to be severely injured by their abusers. These findings led researchers to suggest: o Safety planning is of paramount importance at the earliest point of contact with the victim. o The criminal record of the abuser should be considered in fashioning the protection order. In addition, researchers called for further research on the interactive aspects of domestic violence, such as the: o Use of criminal history information in crafting orders and counseling victims. o Effects and enforcement of specific terms of protection orders. o Actions of police and prosecutors. Research design Initiated in 1994, after a wave of reform across the country had expanded the availability and scope of relief provided in civil protection orders, the project selected for the study three jurisdictions using disparate processes and service models for providing protection orders: the Family Court in Wilmington, Delaware; the County Court in Denver, Colorado; and the District of Columbia Superior Court. Two primary measures of effectiveness were applied. First was self-reported improvement in quality of life after obtaining the order. Second were the extent and types of problems related to the protection order reported by the women, including repeated physical or psychological abuse and continued attempts by the abuser to contact the women at work or home. Four data sources were used in the study: telephone interviews conducted with 285 women petitioners for protection orders approximately 1 month after they received a protection order (temporary or permanent), followup interviews with 177 of these same women 6 months later, the civil case records of these women, and criminal history records of men named in the orders. Key findings Victims. Before receiving a protection order, study participants experienced abuse ranging from intimidation to injury with a weapon. Researchers found that 37 percent of the women had been threatened or injured with a weapon; more than half had been beaten or choked; and 99 percent had been intimidated through threats, stalking, and harassment. More than 40 percent experienced severe physical abuse at least every few months, and nearly one-quarter had suffered abusive behavior for more than 5 years. Abusers. Among men named in the protection orders filed by participants, 65 percent had an arrest history. Researchers noted that many of these men appeared to be career criminals, with more than half having four or more arrests. Charges included violent crimes, drug- and alcohol-related crimes, and property, traffic, and miscellaneous offenses. Of the 129 abusers with any history of violent crime, 43 percent had 3 or more prior arrests for violent crimes other than domestic violence. Effects of protection orders. The act of applying for a civil protection order was associated with helping participants to improve their sense of well-being. In the initial interviews, 72 percent of participants reported that their lives had improved. During followup interviews, the proportion reporting life improvement increased to 85 percent, more than 90 percent reported feeling better about themselves, and 80 percent felt safer. Seventy-two percent of participants in the initial interviews and 65 percent in the followup interviews reported no continuing problems. In several areas, however, the proportion reporting problems rose between the two interviews: calls from the abuser to the participant at home or work (16 percent in the initial interview and 17 in the followup), stalking the victim (4 percent and 7 percent), repeated physical abuse (3 percent and 8 percent), and repeated psychological abuse (4 percent and 13 percent). Victim services. The study also looked at the use of services by participants before and after obtaining a protection order. These were grouped into eight categories: private legal services, medical assistance, police protection, assistance from government services, counseling services, moral support and guidance from friends or relatives, support groups, and assistance from private community organizations. Overall, 78 percent of participants reported they had used at least one type of service. Assistance from friends and relatives was most frequently used, with 46 percent of participants seeking help from people they knew. Next were private community services, such as battered women's shelters and victim advocacy services provided by universities and private agencies (32 percent). Researchers felt that more could be done to ensure that victims are provided with user-friendly information about available services as well as information regarding protection orders and their enforcement through the contempt process. They suggested that judges and police both make this a priority when dealing with domestic violence victims. In addition, a more centralized court process and direct assistance to petitioners make it more likely that victims will develop safety plans and seek services. In conclusion, the researchers noted that the Violence Against Women Act offers a pivotal opportunity through changes in current practice to increase awareness of and access to protection orders and to enhance enforcement strategies. They also emphasized, however, that civil protection orders are only one part of the fight against domestic violence. ------------------------------ Susan L. Keilitz, Project Director; Courtenay Davis; Hillery S. Efkeman; Carol Flango; and Paula L. Hannaford conducted this study at the National Center for State Courts. This research was supported by NIJ grant number 93--IJ--CX--0035. ------------------------------ Points of view in this document do not necessarily reflect the official position of the U.S. Department of Justice. FS 000191 ------------------------------ Events The Challenge of Crime in a Free Society-- 30 years later Responding to the high level of public concern about crime in the 1960s, President Lyndon B. Johnson ordered the establishment of the President's Commission on Law Enforcement and Administration of Justice. The Commission was to examine "every facet of crime and law enforcement in America." The results of that examination were published in 1967 as The Challenge of Crime in a Free Society, a landmark report that called for a "revolution in the way America thinks about crime." The Commission's work laid the foundation for the current Federal role in assisting State and local law enforcement and justice administration. The U.S. Department of Justice's Office of Justice Programs (OJP) evolved from predecessor organizations created as a result of the Commission. To commemorate publication of the report, OJP and its components--including the National Institute of Justice (NIJ)--and the Office of Community Oriented Policing Services sponsored a retrospective symposium held in Washington, D.C., in June 1997. Attending the symposium, whose theme was "Looking Backward, Looking Forward," were prominent criminologists; professionals and practitioners from law enforcement, the courts, and corrections; Federal and State officials; and members of the Commission staff. Participants examined changes in the nature of crime and the criminal justice system, the use of research and statistics, and the societal response to crime and the criminal justice system. Focusing on the results of the Commission's recommendations, attendees assessed the reach of change that has occurred since the report was issued. A publication based on symposium presentations is planned by NIJ. Details of its availability will be announced in the NCJRS Catalog. Record attendance at annual research and evaluation conference Sponsored by NIJ and other components of the Office of Justice Programs, the annual conference on criminal justice research and evaluation (July 20-23, 1997) attracted about 850 participants. They attended a wide range of presentations--some tailored to researchers, others to practitioners, and still others of interest to both. Topics encompassed community policing, drug testing and drug treatment, juvenile and violent crime, violence against women, correctional programs, community restorative justice, place- based crime prevention, DNA databases, and evaluation methodology and issues, among other areas. Addressing the conference, held in Washington, D.C., were William Bratton, former New York City police commissioner and now president of First Security Consulting, and Dr. Alan Leshner, Director of the National Institute on Drug Abuse. Next year's conference will be held in Washington, D.C., July 26-29. Under way: Breaking the cycle of drug use and crime NIJ Director Jeremy Travis participated in launching Breaking the Cycle (BTC) in a ceremony at the University of Alabama at Birmingham on June 10, 1997. Developed by a consortium of Federal agencies--including NIJ and the Office of National Drug Control Policy (ONDCP)--BTC is a research demonstration project to test the effectiveness of a systemwide criminal justice intervention with drug-addicted offenders. The goal of BTC in Birmingham is to provide drug testing, graduated sanctions, judicial supervision, and drug treatment to each felony drug-using defendant regardless of charge and detention status. NIJ is responsible for administering and evaluating BTC demonstration projects. The Birmingham project is funded by a $1 million grant from ONDCP through NIJ. The project is a collaboration between the university's Treatment Alternatives to Street Crime (TASC) project, the Jefferson County court system, the district attorney's office, and the Jefferson County sheriff's department. ONDCP has committed $9 million in fiscal year 1997 to fund an expansion of the BTC concept. Of that sum, $4 million has been allocated for the expansion of additional adult sites. NIJ expects to fund a minimum of two additional sites and will continue to fund rigorous evaluation. NIJ also plans to hold a strategic planning meeting in mid-November 1997 to design a juvenile BTC project. Three million dollars has been allocated for creation of juvenile sites. NIJ expects to fund two juvenile sites and an evaluation. Summer Institute for law enforcement technology Nineteen law enforcement officers from 15 States participated in NIJ's first annual Summer Institute, held in Washington, D.C., August 18-22, 1997, and hosted by NIJ's Office of Science and Technology. The Institute's goal is to facilitate the sharing of technology information. Officers were briefed on NIJ's efforts in support of law enforcement, with particular focus on how participants can make technologies viable and effective in their own agencies. In addition to NIJ program and technology briefings, participants toured the Department of Justice; the Pentagon; the Alcohol, Tobacco and Firearms Laboratory; the Federal Bureau of Investigation (FBI) Crime Laboratory; the Drug Enforcement Administration Training Center; the FBI Hostage Rescue Center in Quantico, Virginia; and the National Law Enforcement and Corrections Technology Center in Rockville, Maryland. Look for more details about the results of the Summer Institute in a forthcoming NIJ Journal. Continuation of perspectives lecture series NIJ Director Jeremy Travis announced the continuation of the Institute's successful policy lecture series, "Perspectives on Crime and Justice," which will resume in December and run through May 1998. Nationally prominent scholars will address such issues as gun markets and the U.S. approach to substance abuse. A speaker schedule will appear in a forthcoming NIJ Journal. Concluding the recent five-lecture series were Cathy Spatz Widom, State University of New York at Albany, in April 1997, and Norval Morris, University of Chicago Law School, in May. Their topics were, respectively, "Child Victims: In Search of Opportunities for Breaking the Cycle of Violence" and "Crime, the Media, and Our Public Discourse." Previous speakers were James Q. Wilson, University of California; Peter Reuter, University of Maryland; and Mark H. Moore, Harvard University. Upcoming land transportation security technology conference Scheduled for Atlanta, Georgia, in April 1998, an international land transportation security technology conference will feature presentations by experts in the field and exhibits of new technologies. Cosponsored by NIJ and the Department of Transportation (DOT), in cooperation with the Department of State, the conference will focus on counterterrorism technology. Among the planned conference topics are terrorism vulnerability assessments, multijurisdictional command structure, weapons detection technology, night vision equipment, analytical tools, information-technology sharing, and lessons learned from previous land transportation security incidents. The conference evolved from a 1996 meeting that Attorney General Janet Reno held with transportation officials regarding terrorism- related concerns. Two subsequent international conferences, organized by DOT and the State Department, focused on sharing counterterrorism technologies related to land-based transportation targets. NIJ, through its Office of Science and Technology, agreed to support DOT's efforts by gathering information from first-responders, while the American Public Transit Association (APTA) obtained information from transit law enforcement. Topics that first-responders indicated as important included detection equipment, contingency planning, toxicology, computer modeling, hazard prediction analysis, standards for vulnerability assessments, procedures for dealing with chemical/biological attacks, and a unified command structure. The APTA group added the following topics: light, portable, and effective detection devices; closed- circuit television; silent alarms; enhanced and secure radio communications; encrypted digital computer systems; electronic information networks; bomb-resistant garbage receptacles and windows; access-control systems; new facility and environmental designs; and specialized personal protective equipment for first-responders. To assist the Department of Transportation in synthesizing position papers regarding technology needs, NIJ cosponsored a focus group in March 1997. Participants were asked to provide information about their counterterrorism technology and training needs. The upcoming Atlanta conference is designed to build upon, synthesize, and disseminate the information obtained as the result of the foregoing efforts. Restorative justice regional symposiums Among the most promising new approaches to criminal justice are those focusing on victim and community involvement in the system. Through various programs and initiatives, the principles of restorative justice help to repair the harm caused by crime and provide a more substantive role for victims and the community. A series of regional symposiums is being held between June 1997 and January 1998 to provide policymakers and practitioners with the opportunity to discuss restorative justice philosophy, practices, issues, and roadblocks with national experts and regional representatives. The symposiums are sponsored by NIJ, other components of the Office of Justice Programs, and the National Institute of Corrections. The symposiums follow on the heels of a national symposium held in January 1996 in Washington, D.C. The event proved so successful that many participants recommended that sponsors build on the momentum generated by the first symposium and hold regional events. The first regional symposium was held for the Northeast region in Burlington, Vermont, early this summer. Following are the dates and regions for future symposiums: o North Central Region. Sept. 28-30, 1997, in Milwaukee, Wisconsin. Invited States: Illinois, Indiana, Iowa, Kentucky, Michigan, Minnesota, Missouri, Ohio, West Virginia, and Wisconsin. o Southwest Region. October 26-28, 1997, in Santa Fe, New Mexico. Invited States: Arizona, California, Colorado, Hawaii, Kansas, Nevada, New Mexico, Oklahoma, and Utah. o Northwest Region. December 11-13, 1997, in Portland, Oregon. Invited States: Alaska, Idaho, Nebraska, North Dakota, Montana, Oregon, South Dakota, Washington, and Wyoming. o Southeast Region. January 1998, in Austin, Texas. Invited States: Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Texas, and Virginia. Because of limited space, attendance is determined by competitive application process. Potential participants must form a five-person team representing a jurisdiction that has some experience with restorative justice. Each team should have a balance of juvenile and adult perspectives and should draw from among the following: policymaker, prosecutor, defense bar, representative of victim and community organizations, law enforcement, courts, corrections, and system research or administration. For more information or to submit an application for attendance, contact the Institute for Law and Justice, 1018 Duke Street, Alexandria, VA 22314; 703-684-5300. ------------------------------ NIJ IN THE JOURNALS The following articles, based on research funded by NIJ, are available from the National Criminal Justice Reference Service (NCJRS). For information on ordering copies, call NCJRS at 800-851-3420. "Accuracy of Adult Recollections of Childhood Victimization: Part 1. Childhood Physical Abuse," Psychological Assessment, 8(1996):4, by C.S. Widom and S. Morris, grant numbers 86-IJ-CX-0033, 89- IJ-CX-0007, and 93-IJ-CX-0031, accession number (ACCN) 166613. This article discusses a study assessing the accuracy of recollections of adults who were physically abused as children. Retrospective self-reports of early childhood physical abuse were compared with official cases of physical abuse documented and substantiated through court records. "Accuracy of Adult Recollections of Childhood Victimization: Part 2. Childhood Sexual Abuse," Psychological Assessment, 9(1997):1, by C.S. Widom and S. Morris, grant numbers 86-IJ-CX-0033 and 89-IJ-CX-0007, ACCN 166614. The article discusses a study assessing the accuracy of recollections of adults who were sexually and physically abused or neglected as children. Findings indicate gender differences in self- reporting and accuracy and substantial underreporting by sexually abused respondents in general. Self-reported measures of childhood sexual abuse are significant predictors of alcohol abuse, depression, and suicide attempts among women. "Childhood Victimization and Subsequent Risk for Promiscuity, Prostitution, and Teenage Pregnancy: A Prospective Study," American Journal of Public Health, 86(1996):11, by C.S. Widom and J.B. Kuhns, grant numbers 86-IJ-CX-0033 and 89-IJ-CX-0007, ACCN 166615. Among the findings of the study described in this article is that early childhood abuse and/or neglect was a significant predictor of prostitution for females but was not associated with increased risk for promiscuity or teenage pregnancy. "Drug Policy and Community Context: The Case of Small Cities and Towns," Crime and Delinquency, 42(1996):2, by M.J. McDermott and J. Garofalo, grant number 91-DD-CX-K049, ACCN 163484. This article reports the findings of a national assessment of drug problems and antidrug initiatives in small jurisdictions. Findings indicate that school officials and community leaders consider alcohol a top concern, while law enforcement and government officials give that distinction to crack and other forms of cocaine. Marijuana use is seen as a problem in most small jurisdictions. Community leaders favor drug education and prevention measures over law enforcement. "Further Exploration of the Flight From Discretion: The Role of Risk/Need Instruments in Probation Supervision Decisions," Journal of Criminal Justice, 24(1996):2, by A.L. Schneider, L. Ervin, and Z. Snyder-Joy, ACCN 162937. This article discusses the findings of a study of the implementation and role of risk/need assessment instruments in probation and parole decisions in Oklahoma. Findings indicate that corrections officials have generally negative or neutral attitudes toward quantitative risk/need instruments but, paradoxically, found it difficult to envision a probation system without them. A small majority believe the probation system is better off with the instruments than with discretionary decisions. "Citizen Involvement in the Coproduction of Police Outputs," Journal of Crime and Justice, 19(1996):2, by J. Frank, S.G. Brandl, R.E. Worden, and T.S. Bynum, grant number 89-DD-CX-0049, ACCN 163880. This article examines how citizens' attitudes toward the police affect their willingness to help the police. Findings of a study indicate little correlation between citizens' attitudes toward police and their "coproductive behaviors" with police, especially to maintain order. Citizen attitudes toward police performance in drug law enforcement, however, do appear to affect whether they will provide police (or community groups) with drug-related information. "Moral Reconation Therapy and Problem Behavior in the Oklahoma Department of Corrections," Journal of the Oklahoma Criminal Justice Research Consortium, volume 2 (August 1995), by D.R. MacKenzie and R. Brame, grant number 94-IJ-CX-0064, ACCN 163402. The authors discuss an evaluation of Oklahoma's use of moral reconation therapy (MRT), a treatment program designed to alter offenders' moral reasoning skills. Findings indicate that individuals who participated in MRT showed a moderate but statistically significant drop in misconduct and recidivism. "Inmates' Attitude Change During Incarceration: A Comparison of Boot Camp With Traditional Prison," Justice Quarterly, 12(1995):2, by D.L. MacKenzie and C. Souryal, grant number 90-DD-CX-0061, ACCN 158211. This article reports the findings of a study on the impact of a military-type regime on inmates' attitudes in six State-level shock incarceration programs. Findings suggest that boot camp inmates develop more positive and less oppositional and antisocial attitudes than inmates in traditional settings. "Less-Than-Lethal Force Weaponry: Law Enforcement and Correctional Agency Civil Law Liability for the Use of Excessive Force," Creighton Law Review, 28(1995):3, by N. Miller, grant number 91-IJ-CX- K017, ACCN 161863. The author reviews the legal principles applicable to officer use of less-than- lethal (LTL) force and makes recommendations to policymakers and agency heads on limiting their exposure to liability claims. The liability principles that govern use of LTL force are similar to those that apply to the use of conventional and deadly force. Recommendations for avoiding liability include adopting policies, providing training, and requiring incident reporting and internal affairs reviews of excessive-force incidents. "Published Findings From the Spouse Assault Replication Program: A Critical Review," Journal of Quantitative Criminology, 11(1995):1, by J. Garner, J. Fagan, and C. Maxwell, grant number 93--IJ-CX- 0021, ACCN 153919. This article reviews the published findings of the spouse assault replication program (SARP), which addressed whether arrest effectively deters misdemeanor spouse assault. Findings indicate that while many methodological approaches used to assess experiments in six jurisdictions were sound, not one was used consistently. The authors conclude that available information is inadequate to support a definitive statement about the results of the experiments. ------------------------------ Publications The following recent publications disseminated by NIJ are available from NCJRS in both online and hard-copy formats. For ordering information, call NCJRS at 800-851-3420. Annual Report to Congress 1996, U.S. Department of Justice, National Institute of Justice, August 1997, 104 pp., NCJ 166585. NIJ offers the tools of research, evaluation, and technology development to expand knowledge and understanding of how public policies can control crime and achieve justice. This report reviews how NIJ applied those tools and its $99 million in total expenditures in 1996, a year in which the Institute's research portfolio increased multifold, spurred in large part by creative collaborations with partners at the Federal, State, and local levels. After reviewing the year's highlights, the report presents essays that focus on NIJ research and development projects pertaining to the underlying issues of violence, criminal justice responses to drugs and crime, community crime control and prevention, trends and emerging concepts in adjudication and corrections, and science and technology. Youth Afterschool Programs and Law Enforcement, Research Preview, summary of a presentation by Marcia Chaiken, U.S. Department of Justice, National Institute of Justice, August 1997, 4 pp., FS 000169. This publication is based on a presentation describing the results of a survey of youth-serving organizations to identify the nature of the crime problem affecting them during nonschool hours and the approaches they are using to prevent crime. Jointly sponsored by NIJ and the Carnegie Corporation of New York, the survey found that local affiliates of national organizations are serving many high-risk juveniles. To the extent that the local affiliates are themselves imperiled by crime and violence, they are likely to enlist police assistance in implementing prevention programs. Findings indicate that young people prefer programs that provide a range of choices-- sports and recreation activities and those that bolster educational and social skills, offer help in coping with peer pressure, and provide instruction in computer and technical subjects. A 60-minute VHS videotape (NCJ 163057), "Youth Afterschool Programs and the Role of Law Enforcement," is also available. Crack's Decline: Some Surprises Across U.S. Cities, Research in Brief, by Andrew Lang Golub and Bruce D. Johnson, U.S. Department of Justice, National Institute of Justice, July 1997, 16 pp., NCJ 165707. Research shows that drug epidemics, like their epidemiological counterparts, follow a natural course, from incubation to decline. For 10 years, NIJ has been gathering information on the course of illicit drug use through its Drug Use Forecasting (DUF) program. This publication examines the progress of the crack cocaine drug epidemic at 24 DUF locations from as early as 1987 through 1996. DUF data show that crack use has followed a distinct four-phase pattern: incubation, expansion, plateau, and decline. Pinpointing crack's current phase in a locality helps criminal justice and health officials to develop better strategies and deploy resources more effectively. The authors state that crack use appears to be in the decline phase in DUF cities on the east and west coasts; at some DUF sites in the interior regions of the country, the drug is at the plateau stage. Child Sexual Molestation: Research Issues, Research Report, by Robert A. Prentky, Raymond A. Knight, and Austin F.S. Lee, U.S. Department of Justice, National Institute of Justice, June 1997, 24 pp., NCJ 163390. The information included in this publication has been distilled from several interrelated reports and studies sponsored by NIJ to strengthen the efficacy of intervention and prevention strategies and ultimately reduce child sexual victimization rates. The publication discusses the frequency of child sexual molestation and factors leading to sexual deviancy in individual offenders, describes classification models for typing and diagnosing child molesters, notes treatment approaches and strategies for community-based maintenance and control, addresses reoffense risk as it relates to criminal justice decisions, and discusses predictors of sexual recidivism. To illustrate the variability of recidivism among child molesters, the authors present findings of a 25-year followup study of 115 released offenders. Reorienting Crime Prevention Research and Policy: From the Causes of Criminality to the Context of Crime, Research Report, by David Weisburd, U.S. Department of Justice, National Institute of Justice, June 1997, 28 pp., NCJ 165041. Crime prevention research and policy have traditionally been concerned with offenders or potential offenders. This publication focuses not on criminals but on the context in which crime occurs. This approach--often associated with situational crime prevention--seeks to develop a greater understanding of crime and prevention strategies through studying the physical, organizational, and social environments that make crime possible. The author reviews factors that have either hindered or contributed to the development of a situational approach to crime prevention research and policy, compares the relative strengths of this approach with more traditional approaches to crime prevention, and identifies areas where situational crime prevention has generated new insights into the crime problem and potential responses to it. Guns in America: National Survey on Private Ownership and Use of Firearms, Research in Brief, by Philip J. Cook and Jens Ludwig, U.S. Department of Justice, National Institute of Justice, May 1997, 12 pp., NCJ 165476. Survey findings indicate that, in 1994, 44 million Americans owned 192 million firearms, of which 65 million were handguns. About 74 percent of gun owners possessed two or more firearms. Gun ownership was highest among middle-aged, college- educated people living in rural small-town America. Whites were substantially more likely to own guns than African-Americans. Survey results indicate that the proportion of American households that keeps firearms appears to be declining. The authors also discuss findings pertaining to the methods of, and reasons for, acquiring firearms; storage of firearms; and the frequency with which guns are used against criminal attackers. Solving Crime Problems in Residential Neighborhoods: Comprehensive Changes in Design, Management, and Use, Issues and Practices, by Judith D. Feins, Joel C. Epstein, and Rebecca Widom, U.S. Department of Justice, National Institute of Justice, April 1997, 116 pp., NCJ 164488. This publication discusses place-specific crime prevention in urban and suburban neighborhoods. Place-specific crime prevention builds on crime prevention through environmental design and draws on results of research on active crime prevention tactics (such as community policing) to emphasize modification of design, use, and management of a specific place to prevent and reduce crime. The authors stress that selection of place-specific crime prevention strategies and tactics should be made in close collaboration with the community. Physical design changes and management changes can be combined to combat criminal activity, reduce disorder, improve safety, and enhance the quality of life in a variety of settings. The publication emphasizes the need for ongoing monitoring and evaluation of the place-specific strategies selected. Preventing Crime: What Works, What Doesn't, What's Promising, Research Report, by the University of Maryland, Department of Criminology and Criminal Justice, for the U.S. Department of Justice, Office of Justice Programs, February 1997, 536 pp., NCJ 165366. This state-of-the science publication responds to the latest in the long line of congressional initiatives to ensure that its local assistance funding is effective in preventing crime. The authors report on what is known--and what is not--about the effectiveness of local crime prevention programs and practices. Chapters focus on communities and crime prevention, family-based crime prevention, school-based crime prevention, labor markets and crime risk factors, crime prevention in specific places, the role of police in crime prevention, and the role of the rest of the criminal justice system. ------------------------------ NIJ AWARDS Three NIJ publications--Research in Brief titles issued in July 1997--summarize awards made by the Institute in fiscal year 1996. The publications provide the following information for each award: identification number of the grant, contract, or other award; project title; name of entity that received the award; name of principal investigator(s); award amount; and a brief description of the award. Awards made under the Crime Act (Violent Crime Control and Law Enforcement Act of 1994) represented more than half of all NIJ awards and more than half of the Institute's spending for fiscal year 1996. Of the two publications listing those awards, one focuses on science and technology awards: NIJ Science and Technology Awards Under the Crime Act: Fiscal Year 1996 (NCJ 165586); the second presents all other awards under the Crime Act: NIJ Awards Under the Crime Act: Fiscal Year 1996 (NCJ 165700). NIJ Awards in Fiscal Year 1996 (NCJ 165701) lists all non-Crime Act awards for 1996 and categorizes them as follows: criminal behavior, crime control and prevention, criminal justice system, technology research and development, and information dissemination and technical support. For information about ordering copies of the foregoing publications or accessing them online, please call the National Criminal Justice Reference Service at 800-851-3420. ------------------------------ FINAL REPORTS The following final reports--in manuscript form as submitted by authors--pertain to completed NIJ- sponsored research projects. The reports are available from NCJRS through interlibrary loan and as photocopies. For information about applicable fees, call NCJRS at 800-851-3420. "Divorce Mediation and Domestic Violence," by J. Pearson, NCJ 164658, 1997, 234 pp., grant number 93-NIJ-CX-0036. The divorce mediation and spousal violence project used several information collection procedures to examine how divorce mediation programs address the problem of domestic violence. Findings revealed that domestic violence is a frequent problem in divorce mediation and that most of the surveyed mediation programs have revised their procedures to enhance victim safety during and after mediation. The report states that domestic violence victims need a variety of community services and dispute resolution forums. "How Portland Does It: Community Prosecution," by B. Boland, NCJ 165182, 1996, 21 pp., grant number 94-IJ-CX-0004. This report describes the genesis, activities, and nature of a community prosecution experiment in the Multnomah County (Portland, Oregon) district attorney's office. Community prosecution is an organizational response to grassroots public safety demands of neighborhoods. Portland's experiment focuses predominantly on quality-of-life and low-level disorder crimes. Other prosecutors' offices are devising surprisingly similar organizational responses to deal with serious violent crime. "Evaluation of the Reasoning and Rehabilitation Cognitive Skills Development Program as Implemented in Juvenile ISP in Colorado," by K. English, NCJ 165183, 1996, grant number 93-IJ-CX-K017. This report presents findings from the Division of Criminal Justice's evaluation of the reasoning and rehabilitation (R&R) cognitive skills development program as it is delivered to juveniles placed on juvenile intensive supervision probation (JISP) in Colorado. The R&R program is mandatory for all JISP clients unless they are deemed by the probation officer to be too disruptive or have characteristics that would prohibit them from benefiting from it. The report indicates that JISP could do more to meet the standards of R&R program developers and to prepare for program delivery. "Prosecutor and Criminal Court Use of Juvenile Court Records: A National Study," by N. Miller and T. McEwen, NCJ 165184, 1996, 105 pp., grant number 93-IJ-CX-0020. This report examines how prosecutors and judges use juvenile records of defendants charged with violent crimes in court. One indicator of a violent repeat criminal is the offender's juvenile record, and the use of this identifier can lead to both priority prosecution and increased court sanctioning. This study was conducted in two phases by the Institute for Law and Justice (ILJ). In phase I, ILJ reviewed the legal and programmatic status of adult courts' juvenile record use in the 50 States. In phase II, ILJ examined use of juvenile records by court decisionmakers in Wichita, Kansas, and Montgomery County, Maryland. ------------------------------ Solicitations Visiting Fellowship Program NIJ's Visiting Fellowship Program supports research and development on high-priority topics that enhance the capabilities of the criminal justice system to combat crime, violence, and substance abuse. Visiting Fellows, while in residence for 6 to 18 months, study topics of mutual interest to Fellows and the Institute. NIJ seeks research-oriented practitioners at the middle and upper levels of the justice profession as well as persons with extensive experience in criminal justice research. Concept papers may be submitted at any time. Applicants should anticipate a decision timeframe of 6 to 9 months from receipt of concept paper to award. For application procedures, selection criteria, eligibility requirements, and other information, call NCJRS at 800-851-3420 and ask for brochure NCJ 165588. Graduate research fellowship NIJ's Graduate Research Fellowship Program provides dissertation research support to outstanding doctoral students undertaking independent research on issues in the criminal justice field. Students from any discipline may apply. Research must focus on a topic relevant to national criminal justice policy or related to concerns of operating criminal justice agencies. Fellowship awards of as much as $35,000 are for periods of up to 24 months. Application deadlines are January 15 and May 15, 1998. For application information, call NCJRS at 800-851-3420 and ask for brochure NCJ 166367. Crime mapping fellowship Through its Crime Mapping Research Center (CMRC), NIJ is supporting research and development pertaining to computerized crime mapping. CMRC's Visiting Fellowship Program offers research opportunities to individuals interested in criminal justice applications of mapping. Award periods range from 3 to 18 months, during which Fellows are in residence at NIJ. Applications may be submitted at any time. Applicants should anticipate a period of 3 to 9 months between proposal receipt and award decision. For more information, call NCJRS at 800-851-3420 and ask for brochure NCJ 166375. Investigator-initiated research NIJ continues to seek proposals for investigator- initiated criminal justice research. Investigators are invited to submit proposals to explore any topic relevant to State or local criminal justice policy. The deadline for receipt of proposals is December 16, 1997. Call NCJRS at 800-851-3420 to receive a copy of the Solicitation for Investigator-Initiated Research (SL 000201). ------------------------------ NEW & NOTEWORTHY Crime and justice Volume 22, the most recent book in the NIJ- sponsored, 20-year Crime and Justice series, reviews research on hate crimes, homicide, probation, sentencing, and other topics. In the preface, editor Michael Tonry pays special tribute to those at NIJ who created and nourished the series throughout the years. To order volume 22, contact The University of Chicago Press, Journals Division, P.O. Box 37005, Chicago, IL 60637; 773-753-3347. ------------------------------ Origin and Evolution of Probation Probation: "A court-ordered disposition alternative through which an adjudicated offender is placed under the control, supervision, and care of a probation field staff member in lieu of imprisonment, so long as the probationer meets certain standards of conduct."--American Correctional Association, Probation and Parole Directory, 1995-1997. Probation in the United States began in 1841 with the innovative work of John Augustus, a Boston bootmaker, who was the first to post bail for a man charged with being a common drunk. Thanks to Augustus's persistence, a Boston court gradually accepted the notion that not all offenders required incarceration. Virtually every basic practice of probation was conceived by Augustus. He developed the ideas of presentence investigation, supervision conditions, social casework, reports to the court, and revocation of probation. By 1956, all States had adopted adult and juvenile probation laws. Between the 1950s and the 1970s, U.S. probation evolved in relative obscurity. But a number of reports issued in the 1970s brought national attention to the inadequacy of probation services and their organization. In recent years, probation agencies have struggled- -with continued meager resources--to upgrade services and supervision. Important developments have included the widespread adoption of case classification systems and various types of intermediate sanctions (e.g., electronic monitoring and intensive supervision). Those programs have had varied success in reducing recidivism, but evaluations of them have been instructive in terms of future program design. ------------------------------ Who Is on Probation? According to a Bureau of Justice Statistics study of correctional populations in the United States in 1996:[1] o About 55 percent of all offenders on probation had been convicted of a felony, 26 percent of a misdemeanor. About 17 percent had been convicted of driving while intoxicated, which can be considered either a felony or misdemeanor, and 2 percent for other offenses. o Women comprised 21 percent of the Nation's probationers. o About 64 percent of adult probationers were white, 35 percent black. Hispanics, who may be of any race, comprised 15 percent of the probation population. o Southern States generally had the highest per capita ratio of adult probationers. Texas had the largest probation population, followed by California. Data from one study suggest that many offenders who are granted felony probation are indistinguishable in terms of their crimes or criminal record from those who are imprisoned (or vice versa).[2] Another analysis found that 50 percent of probationers did not comply with court-ordered terms of their probation; 50 percent of known violators went to jail or prison for their noncompliance.[3] A more recent analysis indicates that 33 percent of those exiting probation failed to successfully meet the conditions of their supervision.[4] A study of a national sample of felons placed on probation found that, on any given day, about 10 to 20 percent of probationers were on abscond status, their whereabouts unknown; no agency actively invested time finding those offenders.[5] Notes 1. Bureau of Justice Statistics, Nation's Probation and Parole Population Reached Almost 3.9 Million Last Year, Press Release, Washington, D.C.: U.S. Department of Justice, Bureau of Justice Statistics, August 14, 1997. 2. Petersilia, Joan, and Susan Turner, Prison versus Probation in California: Implications for Crime and Offender Recidivism, Santa Monica, California: RAND Corporation, 1986. 3. Langan, Patrick, "Between Prison and Probation: Intermediate Sanctions," Science, 1994, 264:791- 793. 4. Bureau of Justice Statistics, Nation's Probation and Parole Population Reached Almost 3.9 Million Last Year. 5. Taxman, Faye S., and James Byrne, "Locating Absconders: Results from a Randomized Field Experiment," Federal Probation, 1994, 58(1):13-23. ------------------------------ Investigative Uses of Language-Based Author Identification Even though language-based author identification has a long way to go before achieving status as scientific testimony, it certainly can be used effectively now as an investigative tool. On April 29, 1992, Michael Hunter died of a lethal injection of lidocaine, Benadryl, and Vistaril. One of his two roommates, Joseph Mannino, reported the death. The other roommate, Gary Walston, was out of town. A fourth-year medical student, Mannino was only weeks away from his medical degree. He admitted giving antihistamines to Hunter for a migraine headache. After an autopsy report showed the presence of lidocaine, Mannino denied injecting Hunter with the drug, which can induce central nervous system collapse. He declared that Hunter must have injected the drug himself. Mannino produced computer disks containing suicide notes from Hunter. Detective W. Allison Blackman of the Raleigh, North Carolina, Police Department contacted Dr. Carole E. Chaski about the possibility of determining the authorship of the suicide notes on the basis of the language used. Detective Blackman gathered almost 10,000 running words of documents spontaneously written by Hunter, Mannino, and Walston. Dr. Chaski's analysis of the syntax, vocabulary, and punctuation patterns of the documents, with statistical testing, showed that the suicide notes were most likely not written by Hunter or by Walston but by Mannino. Soon after, Mannino was charged with first-degree homicide. Through his attorney, Mannino admitted to writing the phony suicide notes. Due to ambiguous test results concerning the level of lidocaine in Hunter's blood, Mannino was convicted of the lesser charge of involuntary manslaughter. In the sentencing phase of the trial, Judge Stephens said, "It's terrible that Michael Hunter died. It's terrible that the defendant unlawfully caused his death. But to give the impression that Michael Hunter took his own life, I find that extremely aggravating in this case" (Raleigh News and Observer, July 28,1994). Mannino was sentenced to 7 years in prison. ------------------------------ Recent Developments in Handwriting Identification Electronic approaches to verifying interwriter variation are under way. The Secret Service uses the German-based FISH computer program, which measures the pixel positions of handwriting scanned into digital format. Funded by the Federal Bureau of Investigation (FBI), Professor Kam of Drexel University and his students are developing another automated system for measuring pixels in digitized samples of handwriting from 400 bank robbery notes. The FBI has formed the Technical Working Group on Documents (TWGDOC). A TWGDOC subcommittee has already begun work on documenting standard procedures for questioned document examination. Proficiency testing of document examiners has been undertaken by Professor Kam. The most recent test revealed that trained document examiners and untrained laypersons with comparable educational backgrounds match handwriting samples correctly at approximately the same level, with each group making correct matches 87 percent of the time. However, trained document examiners made false positive matches 7 percent of the time, compared to 38 percent for untrained laypersons. ------------------------------ The Writing Sample Database The Writing Sample Database--a major component of the Automated Linguistic Authentication System--is designed to take into account general statistical sampling issues and linguistic performance. Decisions about selecting the types of subjects for inclusion were based on a variety of factors, such as a subject's availability, the prominence of writing in the subject's normal lifestyle, dialect similarity or dialect grouping, generally equivalent educational level, and representation of both genders and several ethnicities. Factors considered in selecting topics for writing samples from the subjects included document type and similarity to actual types of questioned documents. Writing tasks assigned to subjects included narrative essays describing traumatic events and personal influences, business letters, personal letters, and threatening letters. Data were collected from two groups: criminal justice majors at a community college and business and nursing majors at a private 4-year college. Subjects wrote, at their leisure, on 10 topics. At present, the Writing Sample Database includes samples from 98 persons, almost evenly divided between males and females, ranging in age from 18 to 49. Almost three-quarters of subjects are white; the rest are black or multiracial. Because the writing samples were collected in a community college environment, writing is part of the subjects' lifestyles, and the subjects generally share equivalent educational levels and dialect backgrounds. The texts produced by the subjects contain approximately 100,000 words. Some subjects contributed as few as 50 words while others produced several thousand. ------------------------------ U.S. Department of Justice Office of Justice Programs 810 Seventh Street N.W. Washington, DC 20531 Janet Reno Attorney General U.S. Department of Justice John C. Dwyer Acting Associate Attorney General Laurie Robinson Assistant Attorney General Jeremy Travis Director, National Institute of Justice ------------------------------ Justice Information Center World Wide Web Site http://www.ncjrs.org