Step 3: Harness the power of data

Step 3: Harness the power of data

October 2021

To build a common understanding of a problem, its causes, and the potential impacts of various options for change, data is essential. Most local criminal legal system data at the county level is held by public institutions that have a duty to make that data transparent and accessible to local stakeholders (with appropriate protections for confidentiality).

Becoming familiar with available data and options for using it to answer different questions about local justice systems is not just the domain of professional researchers, data analysts, or people working inside government. Although some of the detailed steps in data analysis might be beyond the skillset of many people, everyone can understand what the data shows about local trends. Information alone does not usually change policies, laws, or practices, but it is a crucial component to bolster arguments about the current problem and proposed solutions.

Typically, administrative (official) data from local criminal legal system agencies is central to an analysis of jail populations. The primary focus is on jail data, but data from other government and non-government agencies can also be relevant. The goal of this section is to walk through how to think about the questions you want to answer, what data might help you answer those questions, and the possibilities and limitations of different types of data. Vera’s Technical Guide to Jail Population Data is an in-depth resource with suggestions for ways to access jail data and an explanation of the categories and types of data in a typical jail data system and what they mean in plain language. The Technical Guide also provides information on how to do some basic organizing and analyzing of standard jail data and how to use the findings. Some findings are helpful to understand the current jail population and the top contributing factors to it, while others are helpful to think through the likely effects of specific policy alternatives that may be proposed as solutions.

  • Clarify the questions you want to answer. It is important to clearly list the questions you are hoping the data will answer and to refine them as you figure out what data is actually available. Usually, the overarching questions of interest are why are people ending up in jail, and why do they stay? Getting even more specific with your questions is helpful to build your analysis.
    • Examining the status quo.
      One set of questions has to do with describing the current situation: How many people are in jail? How do jail populations vary over time? How long do people stay in jail? What are the legal characteristics of people in jail (pretrial, sentenced, probation violations)? What are the most common charges that people are booked into jail for? Which law enforcement agencies book people into jail? How many people are in jail because of inability to pay bail? How many people are in jail because of technical violations of their bail or probation?
    • Measuring disparities.
      Another set of questions has to do with looking at differences across groups of people: Are some races, genders, ages, or other groups booked into the jail at higher rates than others? Do some groups stay in jail longer? Are some groups more likely than others to face bail amounts they cannot pay? Are some groups more likely to be diverted from jail?
    • Modeling policy impacts.
      A further set of questions has to do with estimating what might happen in the future if specific changes to policy were put into place: If the county implements an automatic pretrial release policy for certain charges, how might this affect the jail population? How would it affect specific subgroups within the jail population? If the county stopped charging people for certain low-level offenses, how much might this reduce the jail population? How would it affect each subgroup of people in jail?
    • Monitoring progress.
      Finally, some questions have to do with assessing the effects of policy changes like the ones listed above: How did a change in policing practice—such as no longer arresting people for ordinance charges like sleeping in public spaces—affect jail population trends?
  • Obtain, clean, and analyze jail administrative data. Jail administrative data refers to the formal records that a jail keeps. Most jails have an electronic database that staff use to record information when a person is admitted into the jail, at various points during their stay, and on release. These are public records and, with appropriate protections for privacy of the personal information of incarcerated people, should be accessible to local citizens. There are various approaches to obtaining and distributing jail administrative data that you can explore in your jurisdiction. See the Technical Guide for more details.

    Jail data on its own tells a very narrow story. In order to put the data from the jail into local context, it is helpful to collect as much other local data as possible to gain a more complete picture of the factors that increase jail population in your community. You can start by looking for county and/or state data in common, publicly available data sources. Possible sources include census data, local budget data, police/crime data, and public health data. For data on arrest patterns, one place to start is Vera’s Arrest Trends dataset.

    Review the Technical Guide to Jail Population Data for more ways to obtain, clean, and analyze jail administrative data.
  • Collect and analyze qualitative data. So far, this toolkit has focused on how to use official data from local government sources. This data is almost always quantitative (numerical). It is crucial to understand trends, but that does not tell the full story of what is happening or how a situation came to be. Also, official data is not fixed or perfect: It is generated by humans working within complex institutions, which means mistakes or bias are not only possible but likely. For example, there is little consistency or clarity in how criminal legal system agencies track experiences of Latinx people, as some systems have few categories for race/ethnicity.[]Sarah Eppler-Epstein, Annie Gurvis, and Ryan King, “The Alarming Lack of Data on Latinos in the Criminal Justice System” (Urban Institute, December 2016), http://urbn.is/cjdata. Understanding the processes and assumptions behind the official data helps put it in context.

    Qualitative data focuses on the experiences of people and groups, and it can provide a deeper understanding of how to address overuse of jails and local criminal legal systems. A “mixed methods” approach—using quantitative administrative data and qualitative research data—will strengthen the overall analysis by combining the power of numerical trend analysis with the nuance and detail of what people say about their own lived experiences.

    In some cases, partnering with a research organization or local university to do qualitative data collection is advisable. This organization can help clarify research questions, determine what kinds of data collection methods to use and whom to involve, and analyze the findings in a systematic way. This organization will also have experience in how to protect people’s privacy and confidentiality when collecting and analyzing data, as talking about criminal legal system involvement can be sensitive and involve certain risks.

Qualitative questions and methods

Group Created with Sketch. Learn more
  • Explore data transparency options. Once your team has determined the central questions and findings, it is helpful to create a way to share the findings with your community. Some common approaches for data visualization include the following:
    • Public dashboards.
      These are websites that present graphic representations of key data points, with varying levels of detail. Some offer the ability to update numbers as new data comes in. Sometimes they are hosted on the website of a county agency—such as the sheriff’s office or the county criminal legal system committee—while other times they are on the website of a nongovernmental or research organization working with the data. One common software is called Tableau, but there are many other options. For example, Hays County, Texas, has a dashboard that shows different aspects of jail population characteristics and numbers.
    • Infographic publications or factsheets.
      These are digital and print publications, typically under four pages, that include visualizations of key findings with some explanatory text and citations. Although these publications do not usually allow for ongoing updating of data, they can be very helpful to present a set of findings at a particular juncture and share them widely with many stakeholders. For example, Vera has produced factsheets for state-level and county-level data, in different formats.
    • Data transparency.
      Public data visualization helps to build transparency and community awareness and understanding of findings. A further step in data transparency is to provide the dataset (with any personal identifying information stripped out) so that people can download it and do their own analyses. Some counties have developed principles or guidelines for data transparency—for all local government data, not just jail data—to ensure access and clarity.