Jeri Mulrow, former BJS Principal Deputy Director; Jeffrey Sedgwick, former BJS Director; and James Lynch, former BJS Director

With the federal statistical community focused on the controversies of the 2020 Census and relocation of the US Department of Agriculture’s Economic Research Service over the last few years, concern for the Bureau of Justice Statistics (BJS) has largely gone unnoticed. However, a new administration focusing on equity and emphasizing evidence-based policymaking has prompted former BJS leaders to join with the ASA and other BJS stakeholders to draw attention to the agency’s challenges and recommend short- and long-term steps to address them.

One such challenge pertains to delays in report and publication releases. Part of the Department of Justice (DOJ), BJS is the principal federal statistical agency providing wide-ranging statistics on criminal justice topics in the United States. The agency addresses challenging issues, some as fundamental as prison- and jail-related deaths or interactions between the police and public and others as complex as estimating crime unreported to law enforcement. Its data are used by federal, state, and local lawmakers, as well as law enforcement agencies, prisons, jails, local communities, advocacy organizations, and researchers.

BJS stakeholder organizations started expressing their concern about these report and publication delays in 2019. In October of that year, the Consortium of Social Science Associations (COSSA) and Crime & Justice Research Alliance (CJRA) sent a letter to Katherine Sullivan, the DOJ principal deputy assistant attorney general for the Office of Justice Programs (OJP), about several BJS reports that were substantially delayed. Soon after, the Prison Policy Initiative elaborated on and documented the delays.

In February 2020, the ASA also wrote Sullivan, focusing on seven BJS awards that funded research on rape, sexual assault, and victim services—amounting to nearly $17 million—for which, at the time, BJS had yet to release the majority of the data and associated reports. In response, Sullivan organized two calls for the concerned stakeholders with herself and the BJS leadership team in the summer and fall of 2020 to discuss the delays and the agency’s uninformative publication calendar.

In December 2020, the ASA, CJRA, COSSA, and Consortium of Professional Associations on Federal Statistics teamed with former BJS leaders to recommend to the new administration and Congress how to address BJS’s challenges and opportunities, including both immediate data integrity issues and decades-long challenges that have hamstrung the BJS’s ability to meet the nation’s criminal justice information needs. The document emphasizes the need for attention to the BJS budget, budget structure, and strong leadership structure to accomplish the following:

• Modernize and strengthen its data collections and publish data and reports in a timely manner
• Produce new products that are timelier and more specific to communities
• Build partnerships to achieve these needs and relationships with stakeholders

As part of its ongoing State of the Data Infrastructure series—an effort to both communicate the fundamental importance of the federal statistical agencies and highlight their challenges (see sidebar)—the ASA’s Count on Stats team spoke with three former BJS leaders about their perspectives on the state of criminal justice statistics.

The ASA and others have categorized BJS as one of four principal federal statistical agencies facing the most acute threats. Would you agree with that and why?

James Lynch
I think it’s in bad shape, at least in the short run. I can’t remember a time when BJS has been subjected to this kind of external excoriation. If you add up the publications that have come out in the last three years, compared to other years, there has been a substantial reduction in the agency’s productivity. It is not that reports are not being written, but that they are simply not being released.

When you’re getting to the structural issues that are causing cumulative agency strain, the problem I had when I was at BJS was too much demand for our products. Indeed, one of the things that the 2008 National Academy panel pointed out is the mismatch between BJS’s mission and its funding. I’m not talking 20 to 30 percent; I’m talking orders of magnitude.

If you look at England and compare what they spend on crime statistics by the size of their crime problem and population to the United States, it’s ridiculous. There are other kinds of structural and long-term issues to be addressed, but the immediate crisis is to get promised and required products out.

 
Jeri Mulrow
BJS is definitely a gem of a federal statistical agency, but it certainly has more than its fair share of challenges, including being woefully—and historically—underfunded and understaffed. This is a challenging area to measure and to gain cooperation of data providers and the trust of data users. They’ve really strived to do that with a small staff and a small budget. But, as Jim alluded to with the lack of publications coming out and the schedule not available, I think there has been an even further erosion of that trust by data users and data providers.

I also think the lack of adherence to or veering away from the principles and practices of a federal statistical agency needs to be immediately corrected. I’m sure there’s a kernel of that still in the agency, but they need a strong leader to come in and have some foresight to not only manage the current challenges, but to also set a strategic path for the future. These are very challenging data and very challenging measures, so a longer-term plan is also needed.

Jeffrey Sedgwick
BJS has a structural problem in terms of funding and staffing. The agency’s mission keeps growing without concomitant increases and resources to accomplish the mission. So, it becomes increasingly difficult to run BJS—in terms of what you’re supposed to be doing, need to be doing, and are asked to be doing—because of limited resources.

Then there’s a dissemination problem in the sense that it’s hard to get resources for an organization if you’re not doing a good job disseminating because nobody knows you exist. Who’s going to invest in something that you’ve never heard from, or what you hear is mundane and not all that important?

But after you get past the structural and dissemination problems, I would say next in line would be a management/leadership problem. I always thought my main job was translating what statisticians say to lawyers, because you’re a statistical agency embedded in a department where virtually everybody you’re working with has training that doesn’t include a course in statistics. And so, if you can, go to the leadership and say, “Look, here’s what our people are finding, and this is why it’s significant. You might want to pay attention to it.” That requires a skill set that you don’t automatically think is an important qualification for somebody running a statistical agency. There is a translation problem between the way people in a statistical agency talk to each other and the way you can effectively communicate with the outside world. I think that needs to be embraced and understood from the beginning when you’re picking somebody for a leadership position.

The other management issue is that BJS’s culture has, in many cases, operated on a hierarchical model of management, which is not really productive for the agency. BJS is organized internally into units around the different data collections. There’s a victimization unit, the corrections unit, a policing unit, and so on. It makes a lot of sense to take the heads of those units and run the agency like an academic department. You essentially bring people to the table and make decisions by consensus among the unit heads, as opposed to a pyramid in which everything has to go through multiple layers and ultimately the director gets to decide everything. That just paralyzes the organization if you have a front office model where all decisions are made. You have to get your best people, who typically are unit heads, and put them around the table and say we’re going to make a decision as a collegial group. You can get a lot of mileage out of that. Not everything is being funneled through the director. You actually have people who are authorized and empowered to show initiative at the level of the unit, or the data collection, and that helps things in terms of productivity.

The final nail in the coffin is if, hopefully not when, a political appointee decides his job is to squelch things that the boss doesn’t approve of. One cannot effectively lead a statistical agency by serving as a gatekeeper between the agency staff and the external environment.

It’s all of those things stacked up—a structural problem, a dissemination problem, management challenges, and then the politicization of the agency. BJS is in trouble, and it’s not going to get turned around short term. I would bet this is a decade-long project, if not longer.

What do you see as BJS’s strengths and potential?

James Lynch
BJS has two important products that allow ordinary citizens to define the crime problem unfiltered by the authorities. The first is the National Crime Victimization Survey (NCVS). It’s probably the world’s best victimization survey. The survey provides a wealth of information about crimes not reported to the police, as well as the nature of interactions between the police and public. In this period of friction between the police and public, the NCVS should be providing objective data on the interactions between police and citizens and how they are changing over time.

The other is the Survey of Prison Inmates (SPI). This provides a detailed profile of the custody population and allows prisoners to report on their experiences in prison unfiltered by correctional staff. Moreover, SPI gives you an opportunity to do all sorts of disaggregations that aren’t possible in the aggregate statistics reported by state departments of correction.

I think BJS deserves a lot of credit for maintaining these expensive and difficult, but indispensable, data collections.

I think there have been a number of other instances in which BJS has been punching above its weight, such as measuring police killings and rape and sexual assault. Here, the agency has taken on complicated measurement issues in a serious and creative way that is not easy to do for a small statistical agency. I thank Jeff for shaking loose the resources to do some of these things before he left. BJS still has the challenge of preserving the NCVS in an era when the cost of household surveys is increasing dramatically, but there are ways to address this problem without sacrificing the survey. You don’t give up the only chance the citizens have to define the crime problem without a fight.

What is the constructive path forward for BJS, in the short term and long term?

Jeri Mulrow
BJS really needs to modernize pretty much everything it does. For publications, it should be thinking digitally first to release these data. The agency needs to modernize the way it does the data collections, data storage, and data linkages. But if you don’t have the budget and the staff, it’s hard to even think about what you ought to be doing.

The new director is going to have to figure out how to balance getting the current situation back on track and bringing the agency into the modern age. They have a huge mandate. And that’s just measuring criminal justice. That doesn’t even go beyond thinking about how to integrate it with other data to be talking about what the drivers of crime are and how we prevent them. It’s all the big-picture stuff that everyone wants, plus produce data at the granular level. There are all sorts of challenges, for sure.

The new director needs to come in and restore the trust of the communities that need to be involved. That includes the federal statistics community and the criminal justice communities that range from victimization all the way through the whole pipeline of the criminal justice system. I think the director needs to focus outwardly on that and get a really strong deputy director who can focus on managing and helping to restructure the inside of the organization and get it moving forward. Those two together need to create a strategic plan for the whole organization to meet the current challenges and move into a new era.

James Lynch
It was a big mistake to remove the BJS director as a Senate-confirmed presidential appointee. Restoring that aspect would help in the recruitment process to attract talent.

You’re also going to have to get out of the structural hole BJS is in. The agency is required to build new statistical collections to report on parts of the justice system currently not included in its holdings at the same time it maintains ongoing data collections. The annual budget, which has been in steady state or worse for decades, does not allow for developing new statistical collections. The Office of Management and Budget (OMB), BJS, and Justice Department should agree to a series of revolving five-year plans in which BJS has a capital budget and maintenance budget. The former should be used to build new collections, while the latter is used to maintain current collections. This is the only way BJS will make progress in providing data on the full range of decisions made in the criminal justice system.

BJS can abet this process of building new collections through collaboration with agencies like Office for Victims of Crime (OVC) or other client agencies in OJP [Office of Justice Programs]. Cooperative agreements with these agencies can be used to build data collections of particular substantive interest to those organizations. These collaborations give the mission agencies the information they need and allow BJS to fill gaps in the coverage of the current statistical collections.

Unless BJS’s budget takes account of the need for both a capital and maintenance budget, the structural problems of the agency will never be fixed.

Jeri Mulrow
I want to build on that and talk about strategic partnerships. When I was there, there was a really nice partnership with BJS and OVC to support some of the NCVS redesign work, and there was the partnership with the FBI to access the National Incident-Based Reporting System (NIBRS) data and push it out. There are opportunities for some of those types of strategic partnerships that can help augment what Jim was talking about in terms of putting together this five-year, two-phase plan. Maybe it’s a five-year, three-phase plan in which you also engage external partners to help provide some of the funding, expertise, and knowledge resources. I think there are many opportunities, but it would need some creative thinking and it will need somebody who can barter those partnerships.

Jeffrey Sedgwick
I love the idea of a capital budget to look at the long-term needs of the agency infrastructure and its core funding. And then there’s the operating budget—I love that idea. I think that’s sellable if you’re prepared to make this kind of a commitment of resources that can be depended on and not vary wildly from year to year, depending on the mood on The Hill. Then, you can begin to plan. But if you’re going to plan to take advantage of that stable commitment of resources, can you execute unless you’ve got a similar kind of stability in the leadership of the organization?

I’m a big fan of a fixed term for the director of BJS with Senate confirmation and a term that’s staggered across presidential administrations. If a BJS director’s term is coterminous with a president, then the dynamic is every president deserves their own director. But if you’re looking at somebody who’s up for confirmation, and they’re going to serve this president and potentially the next president, then the question is, is this person the best person for the people of the United States? Because they’re not tied to a particular president or administration. They’re serving the country. It also gives the director some tenure. Turnover in the leadership of the organization is a good thing. But too rapid turnover undermines the ability to address some of these long-term structural challenges.

The Foundations for Evidence-Based Policymaking Act is changing the way the federal government uses data. How do you see BJS responding to these changes?

Jeri Mulrow
I think they can’t do anything until they address the other challenges that we talked about. Right now, they’re just not able to fully participate in the activities surrounding the Evidence Act. At this point, they just don’t have the resources, and they don’t have the brain space. And they are so far behind in their current data. But this is part of what needs to be their long-term strategy. How do they get to this? How do they get to fully realize all of the opportunities they really have? I feel a little discouraged that they’re not participating and not able to participate very well right now.

Further Reading
Consortium of Social Science Associations and Crime & Justice Research Alliance letter to Katherine Sullivan

Prison Policy Initiative. Since You Asked: Is It Me, or Is the Government Releasing Less Data About the Criminal Justice System?

Bureau of Justice Statistics: Data to Keep Our Communities Safe—Priorities for the 117th Congress and 2021–2025 Administration