U.S. flag

An official website of the United States government, Department of Justice.

The Data-Informed Jail

A data-informed approach to managing jails can yield benefits in key functional areas.
Date Published
January 29, 2021

Jails are the entry point to the correctional system in the United States. As a critical component of the criminal justice system, jails interface regularly with law enforcement, the courts, and the probation, parole, and prison systems.

There are 3,500 local jails and detention centers in the United States,[1] and in 2018, jails held roughly 740,000 individuals on any given day.[2] Recent data suggests that approximately 4.9 million unique individuals are admitted to a jail each year.[3]

Because of the size of the jail population and the scope of services jails provide, a vast amount of critical data are generated by jail systems. Although some jails significantly leverage these data to inform policies and improve both operations and inmate outcomes, many more do not, for a variety of reasons. It follows that there is much untapped potential for jails to operate in a more data-informed manner.

Why should a jail operate in a data-informed manner?

Jails are under constant scrutiny, and taxpayers demand jail services that are both effective and efficient. Jails are complex organizations, and the desired outcomes are much more attainable when data are collected, analyzed, and used to drive decision-making. A data-informed approach can yield benefits in key functional areas such as:

  • Development of key performance indicators and systems for measuring progress.
  • Objective evaluation of the impact of policy changes, programs, and innovations.
  • Identification of trends, projection development, and planning for future needs.
  • Facilitation of information sharing with justice agencies as well as public health organizations.

To examine this issue, the RAND Corporation in partnership with the University of Denver (DU) analyzed insights from a workgroup of jail administrators, correctional experts, and representatives of national associations. This work, sponsored by the National Institute of Justice (NIJ), identified a total of 43 needs related to jails’ use of data, with 13 high-priority needs associated with the four following themes: (See Figure 1: Top-Tier Needs.)

  • Leadership and Organizational Issues
  • Information Sharing
  • Data Collection and Analysis
  • Applying the Data

A fifth theme, Procuring and Implementing a Jail Management System (JMS), related to four of the lower-priority needs, but none of the 13 top-tier needs, the research team determined.

The workgroup’s methods and findings, discussed below, suggest that organizational leadership is essential to a data-informed jail. Leaders must understand and commit to a data-driven decision-making approach, develop and nurture an organizational culture that values data, make the requisite investments in staff and technology, and foster information sharing with relevant justice, public health, and social services organizations.

Figure 1: Top-Tier Needs by Category

CATEGORY

ISSUE/PROBLEM

ASSOCIATED NEED

Leadership & Organizational Issues

Jails with cultures that do not value data are missing out on the potential benefits of proactive data collection and analysis.

Develop educational resources and toolkits that help administrators understand the benefits they could achieve from proactive data collection and analysis.

Leadership & Organizational Issues

Jails need to create an organizational culture that values data.

Incorporate education on the practices and benefits of data management into training for academies, leadership positions, and informal leadership among staff.

Leadership & Organizational Issues

Leadership would benefit from improved statistical literacy to better understand the purpose and implications of analysis.

Create online, self-paced curricula that can be understood by leaders of varying professional backgrounds.

Leadership & Organizational Issues

It can be challenging for administrators to determine what metrics are most appropriate, how to define them, and how to track them.

Research and promote effective strategies for identifying and monitoring key indicators.

 

Leadership & Organizational Issues

Data may not be captured and managed well when staff don’t understand the importance and purpose of it.

Develop effective strategies (e.g., use cases, documentation of return on investment) to champion data-management objectives and to educate line staff on how data collection contributes to the mission of the jail and affects their day-to-day work.

Leadership & Organizational Issues

Developing data management staff (e.g., data analysts) has not been an investment priority for administrators.

Develop effective strategies (e.g., use cases, documentation of return on investment) to educate administrators on the urgency of maintaining data management staff.

 

Information Sharing

Many jail management systems cannot easily interface or integrate with other partners’ information systems (e.g., those of the criminal justice system).

Develop effective strategies to assist jails — and the jurisdictions they support — in planning for the procurement and implementation of information systems chat can be part of an integrated, jurisdiction-wide solution.

 

Information Sharing

Misperceptions about legal issues associated with data sharing (e.g., the Health Insurance Portability and Accountability Act) prevent agencies from even attempting to share information.

Research and publish guidance documents targeted to jail administrators that identify and counter common misperceptions.

Develop effective strategies to help diverse stakeholders build trust by identifying and reinforcing shared values and work through any misperceptions or perceived conflicts.

 

Information Sharing

There is often organizational resistance to sharing one’s own data, even where legal, technical, and security issues with data sharing have been or can be resolved.

Develop guidance documents on risk management strategies for information sharing (e.g., always have a memorandum of understanding, limit scope).

 

Data Collection & Analysis

The general quality (e.g., accuracy and reliability) of data needs improvement.

Publish guidance on effective strategies to improve the quality of manually entered data (e.g., better training, use of predefined fields in drop-down menus).

Applying the Data

Many jails lack the kind of data-informed accountability management models used in law enforcement (e.g., CompStat).

Conduct research to identify jails that are effectively applying these models and disseminate their successful strategies.

Applying the Data

It is very difficult to compare data and metrics between different jails.

Encourage data definitions (e.g., at the national or state levels) for jails to enable better comparison.

Nature of the problem

Despite the volume of accessible data, jails traditionally have not been viewed as information processing organizations. Although there are exceptions, most jails do not consistently operate in a data-informed manner, and therefore cannot fully leverage the power of data to support decision-making, inform policies, and improve both operations and outcomes. For example, searches for state and local jail data on a data portal in the website of Measures for Justice, an organization that collects, standardizes, and publicizes county-level criminal justice data from across the United States, establish significant gaps in jail performance data available from jurisdictions.[4]

Some jails still operate without the benefit of an automated information management system.[5] Many jails use obsolete systems built by companies that have gone out of business or coded in defunct programming languages that do not conform to modern information standards.[6] It is unsurprising, therefore, that county jail data has been described as “often fragmented, incomplete, and unreliable.”[7] Furthermore, many jails tend to use data in an ad hoc fashion, focusing on particular issues of importance as they arise rather than leveraging these data to operate pursuant to a comprehensive strategic plan. Adding to the challenge is the fact that jail staff are often overworked, and the jails themselves are under-resourced. As a result, there is much untapped potential for jails to become more data-informed. Ultimately, leadership is key to achieving a data-informed jail — one that maximizes the power of information to improve not only operational effectiveness and efficiency, but also inmate outcomes.

Research purpose

The joint RAND-DU collaboration, “Data-Informed Jails: Challenges and Opportunities,” is part of a multiyear research effort, the Priority Criminal Justice Needs Initiative, to identify innovations in technology, policy, and practice that could benefit the criminal justice sector.[8] The initiative, developed and supported by NIJ, aims to generate findings that will inform NIJ’s research agendas moving forward. In response to the significant challenges and opportunities discussed above, the data-informed jails research aimed to produce a better understanding of the obstacles jails face in fully leveraging internal and system data to improve outcomes and what is needed to overcome these obstacles.

Methodology

The RAND-DU team assembled a group of 15 subject matter experts to participate in a two-day workshop. Participants included correctional administrators, researchers, and representatives from relevant national organizations.

Before the workshop, participants received literature on jail information systems and the benefits of data-informed decision-making, as a resource and discussion guide. During the workshop, RAND-DU staff conducted highly structured exercises with the group to help identify and elicit information about the most pressing problems in jail settings. The group then assessed how these problems could be addressed. Discussions focused on several major areas relevant to leadership and organizational issues, procuring and implementing a jail management system, data collection and analysis, data application, and information sharing.

From these discussions, the research team identified a set of discrete “needs” — a term used to describe a specific area to be addressed, tied to either solving a problem or taking advantage of an opportunity for better system performance. This process yielded a total of 43 needs (see Figure 2).

Breakdown of Needs by Area
Figure 2: Breakdown of Needs by Area (View larger image.)

Needs and themes

To provide structure to the large set of identified needs, participants ranked each need in terms of expected benefit (relative importance of meeting that need) and probability of fully meeting that need. These ratings were multiplied by each other to produce an expected value score, and that score was used to group the needs into top-, medium-, and low-tier priority needs. In the final analysis, 13 of the 43 identified needs were ranked in the top tier and are listed in Figure 1. The 43 needs fell under five themes: leadership and organizational issues, information sharing, applying the data, data collection and analysis, and procuring and implementing a jail management system. The 13 top-tier needs fell under four of those themes — all but procuring and implementing a jail management system.

Organizational and leadership issues

Overall, more than 30% of the needs (and nearly half of the top-tier needs) fell under the theme of leadership and organizational issues. Although many jail administrators fully understand and value the power of data-driven approaches to improving outcomes, participants acknowledged that this is not universal. One participant noted that many jail leaders suffer from a “failure of imagination” with respect to how data can support and even transform the way jails perform. Emphasizing the value of data to leadership, one participant quoted his sheriff saying that “vision without data is a hallucination.” To bridge this gap, the participants recommended the development of education and training programs, case studies, and toolkits geared toward administrators that demonstrate the benefits and return on investment from effective use of data. Further, the participants recommended the creation of online, self-paced curricula designed to provide jail leaders with a basic level of data literacy.

According to the participants, better education is needed to help leaders create and nurture a culture that values data and supports data-informed practices. Staff at all levels should receive training on the importance of collecting and utilizing quality-assured information. These values should be reinforced consistently through in-service training and regular interactions between staff and supervisors. Further, the participants called for best practices and effective strategies to demonstrate the relevance of data to staff in specific contexts (e.g., communicating how analysis of data about assaults on staff can lead to better interventions and improved safety). Better staff understanding of how data efforts relate to the jail’s mission and the staff’s day-to-day work experience may improve the quality of data collection.

During workshop discussion the participants noted that many jails lack an overarching strategic plan with respect to their data objectives (i.e., what data should be collected and why). Some have not established key performance indicators, so measuring progress toward objectives is virtually impossible. Resources are needed to highlight best practices to help jails create a strategic plan around their data needs. Furthermore, jails need training and technical assistance to help them create an initial plan and evaluate and modify that plan based on emerging needs.

Finally, some jail administrators fail to prioritize the need for information technology and analytic staff, which can significantly limit efforts to become more data-informed. That said, participants readily acknowledged that dedicating adequate resources to supporting this function can be challenging, particularly in jails that suffer from chronic understaffing in security positions. Further, many jails operate under the authority of the county sheriff, and there is generally greater public support for the more visible law enforcement component of county sheriffs’ offices rather than for their jail operations.[9] It can be difficult, therefore, for jail administrators to successfully lobby their sheriffs for the resources required to achieve their data-management objectives. The participants argued that education is needed to help administrators understand the value of data support staff. Case studies and cost-benefit analyses can help demonstrate the benefits in qualitative and quantitative terms.

Information sharing

Participants agreed that timely information sharing — both within the jail and between the jail and other justice, social service, and public health agencies — can improve outcomes and save resources. However, information sharing efforts have historically been fraught with pitfalls, both real and perceived. For example, patient privacy regulations (e.g., under the Health Insurance Portability and Accountability Act) often are misinterpreted, according to the group. As jail administrators tend to take a conservative stance with respect to protected information, they can be resistant to sharing mental and behavioral health care records even with others who have a genuine need to know. To address this challenge, participants called for research to identify and dispel common misconceptions about patient privacy regulations that might unnecessarily deter jails from sharing information. Further, effective strategies are needed to help diverse stakeholders build trust by identifying and reinforcing shared values and by working through any misperceptions or perceived conflicts.

Technical challenges (e.g., informational silos and firewalls) can be a major obstacle. The participants suggested that jurisdiction-based — rather than agency- or department-based — information systems that are built on a common platform would help resolve many of these challenges. Research is needed to explore the feasibility of building a jurisdiction-based information system and to identify effective planning and implementation strategies.

Ultimately, for some jails, organizational culture prevents effective information sharing, even when other hurdles have been cleared. The participants noted that, in many cases, information sharing is dependent upon the leadership qualities of jail administrators and their level of comfort with transparency. Administrators tend to be risk-averse, and many have been trained in a culture that supports keeping things in-house. To overcome these obstacles, the participants recommended the development of guidance on risk-management strategies, including examples of effective memorandums of understanding between agencies and lessons learned from other jails.

Data collection and analysis

The participants noted that poor or inconsistent data quality can undermine a jail’s ability to operate in a data-informed manner. Guidance documents are needed that highlight effective strategies (e.g., better training and improved software structure, including predefined fields in drop-down menus) to improve data quality.

Applying the data

Participants identified two high-priority needs in this area. To maximize the value of data, a performance management system of some sort is typically required. Participants discussed CompStat[10] as an example of a data-driven model for problem identification, development of targeted interventions, and assessment of the impact of those interventions. Research is needed to identify jail systems that are successfully applying this model and document promising practices and strategies that are producing desired outcomes.

The lack of common terminology and definitions was identified as a major impediment to fully leveraging jail data on a macro level. Participants recommended the development of national standards that govern how jails collect key data and stipulate common definitions of terms (e.g., recidivism, use of force). Standardization would allow for improved comparisons between jails with similar characteristics (e.g., size of population, demographics), the tracking of performance metrics over time, and the identification of outliers (e.g., jails with extraordinarily high or low violence rates) to be studied.

Conclusion

Jails have access to vast amounts of data that, if leveraged properly, can improve decision-making, inform resource allocation, and lead to better outcomes for the organization as well as for the inmate population. It can be highly challenging for jails to make the shift to a data-driven management approach, but the high-priority needs identified by the expert panel provide a roadmap of what must be accomplished to move toward that goal.

Date Created: January 28, 2021