Amstat News invited William Sabol, director of the Bureau of Justice Statistics, to respond to the following questions so readers could learn more about him and the agency he leads.

William Sabol has more than 25 years of research experience on criminal justice and sentencing policy, formal and informal social control in communities, disparities in criminal justice outcomes, and statistical methods. A Fulbright Scholar, Sabol completed his PhD from the University of Pittsburgh Graduate School of Public and International Affairs.

What about this position appealed to you?

The challenge of building and enhancing statistical programs that can address the breadth of BJS’s statutory mission while working within the budgetary and staff constraints within which BJS has to operate. There are increasing demands for timely and relevant criminal justice statistics; the position gives me an opportunity to contribute to meeting these demands, to work with talented staff in setting priorities and implementing changes and improvements to BJS’s statistical programs, and to participate in the broader efforts of the federal statistical system to address some of the challenges presented by declining response rates in household surveys and the use of administrative data and Big Data for statistical purposes.

(I note that I came to BJS in late 2006 to head up one of its statistical units; I served as its deputy director and then acting director beginning in January 2013, when Jim Lynch, the former director, took the position as the department chair of criminology at the University of Maryland.)

Describe the top 2–3 priorities you have for the Bureau of Justice Statistics.

Maintaining the integrity of BJS statistical programs and the data it releases. I believe BJS has a solid record of producing credible, if sometimes less than timely, statistics, and, as a result, BJS has a reputation as an honest broker, even when this means it will release statistics that could be perceived to be unfavorable to an administration either in terms of content or timing. This reputation must be preserved.

Enhancing the relevance of BJS’s statistical programs. As with all the federal statistical agencies, the information demands—content and timeliness—on BJS are increasing. To meet these demands, we may need to retire or modify some programs fairly extensively, but not at the expense of losing the continuity of key indicators; we will need to continue to innovate through greater use of a core-supplement model to expand the content of surveys; we will have to continue to expand our use of administrative records while focusing on the value that can be through record linkage, whether this involves linkages internal to record systems to create longitudinal systems or linkage with external data systems to expand content.

Dissemination. We need to modify our dissemination strategies to address challenges in the marketplace, in which entities repackage or use official statistics and become sources that users rely upon. We need to ensure that users and stakeholders view BJS as an authoritative source, less they fail to see the need to continue to fund the statistical system. I think developing products targeted to specific audiences and stakeholders will help.

What do you see as your biggest challenge(s) for the BJS?

Maintaining high-quality statistical programs within a challenging budget environment. The demands on BJS to produce new series and take on additional institutional research responsibilities continue to increase. To meet these, we will have to make difficult decisions about programs that may affect their scope and periodicity, as well as the nature of statistical releases. We will have to figure out how to expand our coverage of criminal justice issues without undermining the coherent statistical infrastructure in place.

Recruiting, training, and retaining staff. BJS staff members are dedicated and committed to their craft and have deep knowledge about how criminal justice systems operate. To meet emerging demands and respond to some of the challenges facing federal statistical agencies—declining response rates, use of administrative data, and Big Data—BJS will need to retool and develop new skills through recruitment and training. A special challenge for a statistical agency with a small staff is how to accomplish the retraining while keeping the trains running.

Demonstrating the relevance of statistical indicators in the era of evidence-based programming. The demands for evidence about what works could provide a basis for a relative shift in funding away from statistical systems toward experimental evidence, but building new statistical systems and upgrading existing ones provides fundamental information to program administrators about the magnitude of problems to consider. It also provides important information to evaluators to use in determining what to focus on to determine what works. In addition, the criminal justice system statistical infrastructure can be used to support large-scale evaluations based on quasi-experimental methods that a small number of experimental trials cannot support on a project-by-project basis.

What kind of support from the statistical community do you look for?

The federal statistical system has been immensely helpful to BJS by establishing both standards that statistical agencies must achieve and protections they should enjoy to preserve the quality of our statistics and the independence necessary to maintain public confidence in these data. Foundation documents like Principles and Practices and the Holdren Memo on Scientific Integrity are valuable resources for any statistical agency.

The statistical community should continue to support institutional arrangements that benefit all statistical agencies.

BJS serves on the Interagency Council for Statistical Policy (ICSP), and this provides a great opportunity to learn from the experiences of the other heads of the principal statistical agencies and the Office for Management and Budget staff. The ICSP is a very helpful group for sharing experiences related to common problems such as staffing and succession planning, budgeting and planning under extreme uncertainty, and addressing challenges confronting the statistical agencies such as Big Data and dissemination and stakeholder engagement issues, to name a few. OMB’s Federal Committee on Statistical Methodology has produced a number of important products that BJS benefits from, notably the FCSM efforts related to administrative data such as its model Memorandum of Understanding for data sharing and its guidance on data quality for administrative records.

BJS has a committee of the American Statistical Association, the ASA Committee on Law and Justice, that provides information to it on statistical and organizational issues such as data quality and statistical reporting practices, designing and using statistical collections for research and evaluation purposes, and training and staff development. BJS also benefits from an Office for Justice Program Science Advisory Board (SAB) with a subcommittee that focuses on statistical issues. The members of this subcommittee are knowledgeable about BJS’s programs and report to the full SAB.

BJS is in the process of opening itself up to the broader statistical community through its visiting fellowship program, its use of intergovernmental personnel act transfers, its analytic research centers, a small young scholars program, and a graduate research fellowship program. A central aim of these efforts is to engage the broader statistical community in mutually beneficial projects that address methodological problems and substantive issues. Through these arrangements, researchers get access to data behind BJS’ firewall in return for useful products. A challenge that BJS faces in opening up to the outside is dealing with the sense of entitlement that some on the outside bring to the enterprise. In these cases, there is a view that BJS can do things putatively in a cost-less manner, such as building customized data systems.

Prior to your tenure, what do you see as the biggest recent accomplishment of the agency?

It’s difficult to single out the biggest accomplishment because the agency has done a lot with a relatively small number of staff. For example, it has entered the world of Big Data by developing an automated system for linking, parsing, and standardizing into statistical research databases on criminal histories the variable and unstructured text files that are known as “records of arrest and prosecution” (or RAP) sheets maintained by states’ criminal history repositories. It has used administrative data from state departments of correction to develop systems for creating longitudinal histories of persons’ prison terms and used this to track the prison histories of individuals. It has a project that will test linking corrections records to other sources of administrative data. It has improved the precision of the National Crime Victimization Survey, is developing an NCVS subnational estimates program, and is expanding the content to address important issues related to police service.

BJS and its data-collection agents received the 2014 American Association for Public Opinion Research Policy Impact Award for the decade of work it did in response to the requirements of the Prison Rape Elimination Act (PREA) of 2003 (P.L. 208-79), which required BJS to measure annually the incidence and effects of prison rape and conduct surveys in not less than 10% of all federal, state, and county jails and of current and former inmates. Along with these, BJS’s more recent efforts to establish a national system of incident-based crime statistics drawn from the operational data systems of local police departments, a project known as the National Crime Statistics-Exchange, and its efforts to build statistical systems about the victims’ services field stand out as models of collaborative efforts to build new statistical infrastructure that fill gaps in our understanding of crime and the broader justice system’s response to it.