Report on the Utility of a North Carolina Jail Database

Duke University Bass Connections Team, 2020-2021

The Duke University Bass Connections Team includes

Brandon Garrett, JD (faculty advisor)
Quantitative Research Team: Qualitative Research Team:
Ruth Wygle, MA (doctoral team lead) Catherine Grodensky, MPH (doctoral team lead)
Kalley Huang Lucy Callard
Effie Mehbod Nicole Calhoun, MSN, FNP-BC, PhD
Albert Sun Rhea Jain
Brianna Joyce
Ben Keschner
Spencer LaFata
Katherine LoBue
Birk McCaffrey
For any questions about this report, please contact Catherine Grodensky at catherine.grodensky@duke.edu.

1
 Table of Contents

Executive Summary 3
Introduction 6
Methods 8
Model State Databases: Description and Sample Analyses 11
Results from Stakeholder Interviews on the Utility of a North Carolina Jail Database 14
Policy Recommendations for North Carolina 31
Practical Recommendations for Local Governments and Jails 35
Conclusion 36
References 37

2
 Executive Summary

In North Carolina, at least 128,000 individuals spend time incarcerated in a local jail each year.
These individuals have unique legal, social, and medical needs that are often poorly understood by policy-
makers and service providers due to a lack of data about jail populations. In the majority of US states,
local jails are funded by city or county governments and run by local officials with little state or federal
oversight. Local governments may choose to fund and support data collection in their jails or they may
not, depending on the availability of resources, their administrative needs and priorities, and the presence
of political will or popular pressure for better jail administration. Since data are collected from some jails
and not others, and because counties develop their own unique data collection methods and systems, it is
extremely difficult to understand characteristics of the entire jail population in a state. A few states have
developed sophisticated systems to collect and report data from their local jails, but the vast majority have
no such effort and therefore know little about who is in their local jails, why, and for how long.
This report details findings from a research project aimed at exploring the need for a
comprehensive system to collect and report data on local jails in North Carolina. A team of PhD and
undergraduate students from Duke University spent one year exploring quantitative and qualitative
research questions related to the need for jail data in North Carolina. Quantitative research focused on
demonstrating the types of analyses that can be conducted using comprehensive jail data from other
states, and qualitative research focused on exploring the perspectives of stakeholders in North Carolina
and other states on the utility and desired features of a State-run jail database.
Model state jail database analyses
Quantitative analyses were conducted using data from statewide jail databases administered by
California, Texas, and Colorado.

• Colorado analyses were designed to understand predictors of jail population reduction in response
to COVID. The analysis found that the biggest predictor for whether a jail decreased their jail
population was the urbanicity of its county: urban jails decreased their population by 12 percent
more than rural jails on average, all else held constant.
• California’s Jail Profile Survey data were used to evaluate changes in jail populations after
Proposition 47 (“Realignment”) mandated the release of thousands of individuals from California
prisons. Findings showed that for each analyzed year, a majority of California’s 57 counties
experienced increasing jail populations after realignment, the opposite effect from the state
prisons.
• Data from Texas jails were analyzed to understand which counties are noncompliant with state
data reporting protocols and the proportion of Texas’ jail inmates residing in noncompliant jails.
Findings show that 1 in 6 people (roughly 17%) detained in jails in any given month in Texas
reside in a non-compliant jail, even though there are only 22 non-compliant jails in Texas
compared to 246 compliant ones (roughly 9%). Low-income county jails were also more likely to
be deemed non-compliant than high-income county jails.

3
Qualitative interviews
Analysis of semi-structured qualitative interviews with 15 criminal justice stakeholders (13 from
North Carolina and 2 from other states with comprehensive jail databases) found the following themes:

• Data on jail populations would be highly useful for informing policy, evaluating programs,
directing resources, and managing public health crises.
• Participants described many uses for data on North Carolina jails, focused on illuminating the
status quo inside jail facilities, convincing politicians or the public to support needed policies or
resources, and evaluating the effect of policies.
• The majority of local jails and counties in North Carolina have inadequate data on health and
healthcare in their jail populations, and better jail data could improve counties’ ability to budget
and support healthcare for their jails.
• Data are specifically needed for urgent health crises such as mental illness, which is
overrepresented in jail populations; opioid use, for which resources and treatment are available
but challenging to disseminate; and pregnancy, which is often undetected and high-risk in
incarcerated settings.
• COVID-19 is a recent example of a crisis for which data could have played a much larger role in
controlling disease transmission in jails.
• Resources (particularly for technology and staffing) should be allocated to support valid and
consistent data collection in jails, especially under-resourced jails in rural areas.
• Oversight mechanisms are critical in ensuring that data are reported by all jails consistently, and
that data collection needs to be standardized across all facilities in terms of the information
gathered, the questions asked of jail residents, and the frequency of reporting.
Recommendations for North Carolina
Based on the findings from this project, the Bass Connections team has developed the following
specific recommendations for the state of North Carolina:
1. North Carolina should require jails to report daily counts of sentenced and unsentenced
populations immediately, as this information is some of the most basic and essential data for
evaluating policies to reduce jail populations and for understanding jails’ responses to crises such
as COVID-19. Daily total counts (not sentenced and unsentenced) are already routinely reported
to the North Carolina Department of Health and Human Services but are not made publicly
available to other stakeholders in the state who need it for their work.
2. Beyond daily count data, it is important that North Carolina jails are required to report the most
important demographic and health variables: race/ethnicity to understand systematic
disparities in treatment behind bars; age and gender, for the purpose of understanding group-
specific needs; and health variables including mental illness, opioid use and treatment, and
pregnancy.
3. North Carolina should provide funding to counties to support collection of valid data,
particularly in under-resourced counties, and establish oversight mechanisms to increase
compliance with data collection standards. Finally, the collected data should be made publicly
available through a user-friendly, easy-to-understand online interface.

4
Recommendations for jails and counties
In the absence of state-wide policy reform requiring data reporting from jails, counties and jails
can take steps to improve their use of data to inform their own practices and to share with external
stakeholders serving their populations.
1. Jails that do not already do so should establish standard procedures to record daily jail
population count data electronically, ideally including sub-counts of those who are sentenced
and unsentenced.
2. In the absence of a standardized state screener for mental health conditions, jail officials should
talk to their local or state health department to find a valid mental health screener a for their
correctional setting.
3. All jails should administer lab pregnancy tests to all reproductive aged female inmates and
report these results to county and state health departments.
In conclusion, North Carolina would benefit from State-supported jail data collection and
reporting. Understanding jail populations is the logical and urgent next step in unraveling the complex
web of mass incarceration in the US, and in mitigating the negative impacts of incarceration on
individuals and communities.

5
 Introduction

Background on the Need for Data on Local Jails

Approximately 2.3 million individuals were held in United States correctional facilities on any
given day in 2019, although the actual number experiencing incarceration was far higher due to millions
of people cycling through facilities for relatively short periods of time (Henrichson, 2016; Sawyer &
Wagner, 2019). Individuals experiencing incarceration have unique legal, social, and medical needs that
are often poorly understood by policy-makers and service providers. Understanding the needs and
experiences of incarcerated populations depends on the availability of information from detention
facilities in the US, which is highly variable. One type of detention facility, the state prison, benefits from
infrastructure that supports the collection and use of data. State prisons are uniformly administered by
state governments through central departments of public safety or corrections, which create policies and
resources for collecting standard information and recording it in centralized data systems. These data can
then be used, for example, by prison medical providers to identify inmates who are diagnosed with
diabetes and to monitor how well their blood sugar is being controlled; by policymakers to examine the
number of individuals who are in prison for nonviolent crimes and calculate the costs of incarceration vs.
diversion; and by prison administrators to understand trends in violence and suicide in their facilities and
design effective prevention strategies.
Unfortunately, similar data are rarely available for local jails. Jails house fewer individuals than
prisons on any given day, but due to the large number cycling through with shorter stays, nearly 18 times
more people are admitted to jails than prisons annually (Copp & Bales, 2018). In the majority of US
states, local jails are funded by city or county governments and run by local officials with little state or
federal oversight. Local governments may choose to fund and support data collection in their jails, or they
may not, depending on the availability of resources, their administrative needs and priorities, and the
presence of political will or popular pressure for better jail administration. When local governments and
jails do engage in data collection, they typically use tools and systems they design themselves, producing
data that are ultimately unable to be compared or aggregated with other jails. For these reasons, it is
extremely difficult to understand characteristics of the entire jail population in a state.
In recent months, the COVID-19 pandemic has provided an example illustrating the importance
of data in monitoring and controlling the disease in jail populations, which are highly vulnerable to
infection because their close quarters hinder social distancing efforts. Currently the best data available on
jails in the US are collected by the federal Bureau of Justice Statistics (BJS), which surveys a
representative sample of jails annually (Annual Survey of Jails) and a census of all jails every five years
(Census of Jails) (Statistics, 2019). Unfortunately, these data include only the most basic characteristics of
those in jail and are often not publicly available until years after they are collected. Ideally, a strategy for
managing COVID-19 in jails would require more immediate and frequent measures of the number of
people being detained; information on pretrial status and criminal charges for those being held, for
purposes of expediting release for those posing little threat to society; and information about underlying
medical conditions in facilities that may make inmates more vulnerable to the disease (National
Academies of Sciences & Medicine, 2021). To enable better tracking of COVID-19 in jails the BJS added
a Special Addendum on Coronavirus to its 2020 and 2021 Annual Survey of Jails, which collects monthly
jail population data and COVID-19-specific measures on testing and cases (Minton, Zeng, & Maruschak,
2021). However, the Annual Survey of Jails still relies on a nationally representative sample of jails
(rather than a census), which limits its usefulness to states and localities who need data for their own

6
COVID-19 efforts. Furthermore, the Addendum is likely a temporary effort that may not be able to
provide data for future crises.
The vast majority of states have either not attempted to collect data from jails or have conducted
studies limited to particular years or counties and have not made the data publicly available. However,
some states have invested time and resources into creating comprehensive jail data collection
infrastructures a few have even developed user-friendly online interfaces to increase the public
accessibility of their data. In many instances, these states’ data collection efforts were legislatively
mandated and are supported with funding from state budgets, and hold tremendous potential for
addressing the gap in knowledge about jail populations in the US.
The Current Project: Informing Jail Data Collection in North Carolina
A comprehensive statewide jail data collection system in this state may enhance state and local
governments’ ability to engage in planning and resource allocation, program evaluation, and especially
criminal justice reform, which has been a focus of bipartisan legislation for over a decade. In 2018, a
legislative mandate for comprehensive jail data collection was passed in the North Carolina House of
Representatives but failed to pass the Senate. In Spring 2019, a much smaller-scale statute was ratified as
part of an unrelated piece of criminal justice legislation (the First Step Act, H511). This statute required
North Carolina’s Government Data Analytics Center, under the Department of Information Technology,
to conduct an exploratory study to understand the data currently available from local jails and to explore
the feasibility and utility of a comprehensive jail database in the state ("House Bill 511," 2019).
However, the statute did not specify the research methods to be used in the exploratory study, and did not
require that the study consult potential users of the database or administrators of model databases in other
states.
The current project, a student- and faculty-run research study sponsored by the Bass Connections
program at Duke University, uses multiple research methods to explore the current need for data on North
Carolina’s jail population, the potential uses of a state jail database, and the specific features required for
a database to be most useful (Duke University, 2020). This report presents analyses of model state jail
databases in Texas, California, and Colorado to illustrate the potential of these systems, and then presents
findings from 15 qualitative interviews conducted with officials in those states and stakeholders in North
Carolina about how such a system may benefit the state. This research study was designed to produce
findings that would be useful to North Carolina state administrators and policymakers as they design
comprehensive jail data collection efforts in the state. Therefore, after the methods and findings are
described, the report ends with a set of policy recommendations for the state to consider as it moves
forward with this effort, as well as recommendations for individual localities and jails that are interested
in improving their data collection. Overall, the report’s findings are supportive of the creation of a
statewide jail database in North Carolina, but several conditions are necessary to ensure that the database
provides enough benefit to justify its cost.

7
 Methods

Beginning in August of 2020, a team of Duke University undergraduates, PhD students, and
faculty began a year-long project involving in-depth didactic learning about the US criminal justice
system and empirical research into jail data collection systems in North Carolina and other states. The
team began by reviewing online information about jail data collection efforts in all 50 US states and
identifying states with the most well-developed systems to use as models for the next research phases.
Students then divided into two sub-teams to conduct quantitative and qualitative research, respectively,
based on findings from the 50-state review.
Identifying Examples from Model State Systems
To demonstrate the utility of standardized jail data collection efforts at the state level, a sub-team
of students generated example analyses using data from the current efforts in California, Texas and
Colorado. The goal of this exercise was to show stakeholders the potential uses for such data should it be
collected in North Carolina. To begin, students accessed the data for each state. In the case of Colorado,
data was downloaded directly from the State’s data dashboard (Colorado Division of Criminal Justice,
2021). While data from California and Texas are publicly available, they are published in PDF format to
each State’s Department of Corrections website. However, these data have been compiled into a more
user-friendly format by a team of researchers at the University of Pennsylvania and are available for
public download through the University of Michigan’s ICPSR web portal. As a note to stakeholders in
States that are interested in developing standardized jail data collection systems, we highly encourage that
these systems be developed such that the public can easily and directly analysis-ready files from the
website of the State authority overseeing the data collection. As all data used were secondary and publicly
available, approval from the Duke University Campus Institutional Review Board was not necessary.
Students subsequently reviewed and cleaned the data. All data cleaning and statistical analyses
happened in R. The data cleaning process especially involved assessing the data for missingness and
handling accordingly. We raise this point here to emphasize another important feature of good
standardized data collection: ensuring quality checks during the collection process. For example, certain
facilities in all three states consistently failed to report any data. In other instances, certain variables had
excessively high rates of missingness (often deeming the variable unusable for most statistical analyses),
which suggests that the question wording may have been too difficult or that there was not stakeholder
consensus during the instrument-development process (i.e., jail administrators did not feel comfortable
sharing this information for these purposes).
Prior to beginning analyses, students developed research questions and analysis plans. These
questions and plans went through multiple rounds of peer review with a focus on feasibility and policy
relevance. Students also conducted literature reviews to understand the current research landscape around
jails. In some instances, the developed research questions required that the jail data be analyzed with data
from other sources. For example, one student used US Census data to develop variables for median
county income and a binary indicator of whether or not the county was considered rural or urban. This
effort raises a third additional point of importance for those considering implementing jail data-collection
efforts in their states: data should be created in a format that easily allows for file merging with other
sources, likely with county as the variable to be used for record matching.
Lastly, students drafted extended reports and one-page summaries with the findings from their
analyses, excerpts of which are presented in the next section of this report. The goal for including these is

8
to demonstrate to stakeholders in North Carolina, and other states currently lacking a standardized jail
data collection infrastructure, both the kinds of research questions that standardized jail data can help
answer and examples of research products targeting a variety of audiences.
Qualitative Inquiry Focused on North Carolina
After completing the 50-state review, a second sub-team of students designed a qualitative
research study to explore the utility of a jail database in North Carolina. All research methods were
reviewed and approved by the Duke University Campus Institutional Review Board. Students spoke with
experts in jail administration and research to develop draft qualitative interview guides and a list of initial
stakeholders to be interviewed. The list of stakeholders included scientific researchers, policymakers,
legal experts, service providers, and advocates working with jail populations in North Carolina, as well as
sheriffs and jail administrators responsible for jail data in North Carolina and the three model states.
Students added individuals to the list during the study through “snowball sampling,” wherein
interviewees were asked to recommend other individuals who may also be knowledgeable about the topic.
Qualitative methods allow for a more nuanced understanding of a research topic than is often
available using numeric or categorical data. Rather than asking closed-ended questions with pre-set
answer choices, qualitative interviews utilize open-ended questions and probes to allow respondents to
share their opinions and lived experiences in their own words. Interview guides for this study were
designed to be “semi-structured,” including many pre-determined open-ended questions designed to focus
the discussion on the state jail database but also allowing for flexibility in probing for information on
additional relevant topics that may emerge during the interview. Undergraduate interviewers were trained
on qualitative interviewing skills by a PhD team leader who was an experienced qualitative researcher
and practiced using the guide in pairs before using it with study participants.
Slightly different versions of the interview guide were developed based on the nature of the
interviewee’s occupation (i.e., health, legal, legislator, sheriff or jail administrator, and “general”) and
whether they resided in North Carolina or one of the model states. The appropriate interview guide was
chosen for each interviewee based on a team conversation prior to the interview itself. All interview
guides consisted of the following categories:

• Interviewee’s background
• Types of decisions the interviewee makes regarding jail populations
• Sources of information used by the interviewee to make decisions, and any unmet need for
information
• Interviewee’s opinion about utility and desired content of a North Carolina jail database
Potential interviewees were emailed an invitation to speak with student interviewers for 30-60
minutes via phone or Zoom (due to restrictions on in-person research imposed during COVID-19). Emails
made clear that participation was voluntary, and interviewees reviewed an informed consent form and
gave verbal approval prior to the interview. Proper Zoom safety mechanisms were utilized to protect
anonymity of the interviewee. All interviews were audio recorded with the permission of the participant,
and afterwards the interviews were transcribed by a professional transcriptionist. To bolster recruitment,
procedures were modified two months into the study to allow interviewees to be compensated with a $25
Amazon gift card for their participation.
Fifteen interviews were conducted by seven research team members. Interviewees had
overlapping areas of expertise, including 12 individuals from North Carolina, 2 individuals from states
which have or are developing a statewide jail database (CA/TX/FL), 3 sheriffs or jail administrators, 2

9
individuals with leadership roles in state government or corrections, 2 advocates for Justice-Involved
Individuals (JII), 5 researchers studying JII, 2 healthcare providers working with JII, 2 employees of local
court systems, and 1 state legislator.
Students conducted thematic analysis of all interview transcripts using Dedoose Qualitative
Software (Salmona, Lieber, & Kaczynski, 2019). First, team members created a codebook containing
topical and thematic codes, each with definitions and examples for when these codes should be applied to
text in the transcripts. Afterwards, pairs of qualitative team members used Dedoose to apply the codes to
relevant text excerpts in a subset of transcripts, comparing their applied codes with one another and
discussing them to reach consistency. After an initial round of coding, the codebook was revised to clarify
codes that were initially unclear to team members, and to add and remove codes as needed. Afterwards,
the finalized codebook was applied to all interviews by 2-3 students. After all interviews were coded,
qualitative team members created analysis reports of themes that emerged as most important from the
interviews. They did this by using Dedoose to export all of the text excerpts assigned with the relevant
codes and reading the text excerpts to identify and describe themes and sub-themes. These thematic
analysis summaries provide the results presented in this report.

10
 Model State Databases: Descriptions and Example Analyses

In this section, we present the results of analyses using data from the state jail databases in
California, Texas and Colorado. The goal of these analyses was two-fold. First, by engaging with these
existing databases, we were able to see firsthand the features of standardized data collection that are most
useful for research purposes in order to make suggestions to those engaged with bring data collection
efforts to more states. Through our efforts, we developed four suggestions:
1. Data should be made available for download into csv files or other file types that are easily
integrated into statistical software environments (i.e., not PDFs).
2. Questionnaires and data reporting tools should be reviewed with jail officials prior to
implementation to ensure an understanding of what information is being requested and for what
purposes.
3. Data should be formatted in such a way that merging the jail data with data from other sources is
a straightforward action.
4. When possible, bivariate data should be reported on with respect to race/ethnicity, and gender to
provide a better understanding of disparities among jail populations.
The second goal of these analyses was to demonstrate the types of questions (and show potential
formats of answers) that standardized jail data can help answer. We show in the example analyses below
that this type of data is of great potential utility to stakeholders because it can be used to:

• track jail populations in real time;

• test the effectiveness of certain policies; and
• help assess facility compliance with state codes and standards.
Tracking jail populations in real time: the case of Colorado jails during the COVID-19 pandemic
The COVID-19 pandemic has placed immense strain on jail authorities as they balance concerns
over public health and public safety. With overcrowding and characteristically unsanitary conditions, jails
are uniquely positioned to spread infectious diseases such as COVID-19 in its quarters and to local
communities. At the outbreak of the pandemic, decision-makers at jails across the country made an effort
to decrease the size of their jail population to prevent the spread of COVID-19. These efforts often
involved slowing the intake of new persons, as well as releasing individuals who met specified criteria
(e.g., older than a specific age; being held for a certain category of offense). Because of the concerted data
collection efforts in the state, officials and other stakeholders in Colorado were able to closely track jail
conditions during the pandemic and will have access to this data as they work to develop tentative plans
for future disasters.
For this example, we sought to understand how jail population decreases differed for facilities in
urban versus rural areas. Using regression analyses, we found that, on average, jails in urban or
metropolitan areas saw population decreases that were ~12% greater than jails in rural areas. The chart
below shows the change in the total jail population during the period of the COVID-19 pandemic,
stratified by locality of the jail. These findings suggest that rural incarceration was somewhat resilient to
COVID-19, which fits with larger national trends. According to the Vera Institute of Justice, there’s been
a "jail boom" in rural incarceration rates across the country where rural county budgets are allocating
more money toward building and expanding new jails instead of housing, transportation, and public
health. Further research should be done to see whether the same fiscal incentives that are driving the jail
boom are primary causes to our findings about why rural jail populations were so resilient to COVID-19.

11
 Also noteworthy is that the Colorado Jail dataset was signed into existence by the bipartisan
Colorado State House Bill 19-1297 in mid-2019, and the bill expanded previous and outdated efforts to
catalogue Colorado jail data. The dataset is currently the most comprehensive, open-source published, and
catalogued state-wide jail database in the United States. Colorado’s data is published in Comma-separated
values (.csv) format, allowing data researchers and the public to directly understand and run models on its
data without needing to scrape it from a website. As such, it should be a motivating example for other
States to adopt a similar, modern system of data collection on their jails.

Testing policy effectiveness: the case of California jail populations before and after Realignment
Routinely collected jail data can be a vital tool for evaluating the effectiveness of policy. This is
true for stakeholders in California who are eager to assess the impact of the Public Safety Realignment
initiative, which was signed into law in 2011. This legislative push was in response to a 2011 ruling by
the US Supreme Court that California’s overcrowded prisons were violating the constitutional rights of
the individuals being held in these facilities. One major component of Realignment was an increased
reliance on local jails to hold individuals on behalf of the State. Using data from the California Jail Profile
Survey, researchers are able to easily track the jail population in the state and gain a better understanding
of how Realignment has affected conditions at these local facilities (California Board of State and
Community Corrections, 2021).
One way to easily track such things is by creating interactive visualizations, such as the one
available here, https://kalleyhuang.shinyapps.io/california_jails_app/, created by a student member of our
research team. Users of this tool can easily see, at the state and county-level, how the size of the jail
population changed over time, as well as how average length of stay and the ratio of sentenced to
unsentenced persons experiencing confinement changed. The tables below show the percent changes for
these jail traits one, three and five years after the passage of the Realignment legislation. The median
percent change is shown, along with the percent change at the 25th and 75ths percentile in order to
demonstrate the distribution. For each analyzed year, a majority of California’s 57 counties experienced
increasing populations, decreasing ratios of unsentenced to sentenced, and increasing lengths of stay in
comparison to 2011.

12
 Percent change in average daily jail population
Year Median percent change 25th percentile 75th percentile
2012 2.53 -4.37 8.82
2014 14.2 4.30 24.6
2016 8.45 -6.46 19.3

Percent change in ratio of unsentenced to sentenced daily jail populations*

Year Median percent change 25th percentile 75th percentile
2012 -0.236 -1.05 0.0988
2014 -0.531 -1.15 0.142
2016 -0.259 -1.00 0.405
*a negative percent change would indicate that more sentenced than unsentenced individuals were part of the
average daily population than prior to Realignment.

Percent change average length of stay, pretrial and sentenced populations

Pretrial population Sentenced population
Year Median 25th 75th Median 25th 75th
percent change percentile percentile percent change percentile percentile
2012 0 -1.25 0.0988 5.98 -0.5 10.8
2014 2.35 -1.00 0.142 10.8 2.16 23.7
2016 2.06 -0.25 0.405 10.5 2.7 26.0

Assessing facility compliance: the case of Texas jails

Another potential use of data collected from a standardized, statewide effort is to assist with
assessing the compliance of facilities with various codes and standards. Examples of jail features that
might be assessed routinely for code compliance include sanitary measures, nutritional provisions, and
staff training. Stakeholders are also commonly interested in knowing such things as the ratio of staff to
confined persons, average length of stay, and availability and utilization of programming. Standardized
jail data collection can provide this important information in a routine fashion. In the case of Texas, the
Texas Commission of Jail Standards routinely publishes the list of jails that have failed to meet standards
and are thus determined to be “non-compliant” (Texas Commission on Jail Standards, 2021). While
compliance alone is an important indicator of jail function, this information used in combination with
other collected data can tell an important story about where and why individuals are experiencing
confinement in unacceptable conditions.
Looking at the location and population of facilities deemed non-compliant by the Commission,
we found that at the end of 2017:

• Approximately nine percent (8.9%) of jails in Texas had a status of “non-compliant.”

• One in 6 persons confined in a Texas jail at the end of 2017 were being held in a jail with a status
of “non-compliant.”
• Jails in counties where the median income is lower than the state average are more likely to have
a status of “non-compliant” than jails in counties where the median income is higher than the
state average.

13
 Results from Stakeholder Interviews on the Utility of a North Carolina Jail Database

To explore the potential utility of a jail database in North Carolina, we asked participants to
describe the informational needs they encounter when making decisions about jail populations and to
speculate about how they and others may use data from a State-run jail database. We also asked about
potential barriers that may impede State efforts to collect data from jails. In this section we present the
uses of a database that were stressed as most important by stakeholders, starting with more general uses to
inform state and county policy and leading into more specific uses in healthcare. Specific examples of
using data to inform mental healthcare, pregnancy, opioids, and COVID-19 are discussed in detail. We
also present the barriers to jail data collection discussed during the interviews as implementation factors
to consider in state jail database development.
1. Data from Jails Can Inform State and County Policies
Most participants stated that comprehensive jail data could be used to inform state and county
policy in North Carolina. The following themes were identified in participants’ discussions of data usage
for policy: data on jails can 1) illuminate the status quo inside jail facilities, 2) convince politicians or the
public to support needed policies, and 3) evaluate the effect of policies impacting jails and jailed
populations.
Jail data is helpful in illuminating the status quo inside facilities, which is important in helping
policymakers, service providers, and legal practitioners understand the needs of jail populations. A North
Carolina-based criminal justice advocate explained how basic information about a county’s jail residents
is needed to inform decisions about allocating local resources, but is often not available:
“It’s important to know who’s in there and for how long, because how long someone’s
there can determine how many reentry services they may need. If somebody’s been in jail
three to six months, we probably know they might need some help with housing when they
come out versus somebody who’s been there for two or three days. That’s why the
demographic information—who’s there for why and for how long—all of that is
important.
When we talk about over-crowdin’, when we talk about, how do we provide better
outcomes, we can’t have any evidence-based solutions because we don’t have the
evidence. We’re not able to gather and collect evidence. That’s why it’s very important
that we have the data so that we can know what’s happening.”
A specific example of how information on current jail practices could be useful in guiding policy
was provided by a North Carolina policymaker, who stated that it would be helpful to have data on the
medications that are currently being prescribed in jails to help inform policy response to the opioid crisis:
“There have been debates around the type of medications that inmates have access to or
deal with. Some folks still believe that there’s this tug-of-war around the level of type of
prescriptions, medications, that folks have just so you have opioid issues out in the
community that they still, sometimes, end up continuing even when folks are incarcerated
because of the level there. If you don’t have the data, you can’t say whether or not that’s
really happening or not”.

14
 Data illuminating the status quo are helpful not just for policymakers, but also for legal
practitioners. One local county employee felt that jail data would be helpful for public attorneys in the
district:
“I think some of the stuff that would be helpful would also help inform our district
attorney’s office, our public defender’s office. Which clients are staying in longer? What
types of charges are they staying in longer for? Is there a difference, a disparity between
racial groups? Is there a disparity between gender? Who’s getting out to treatment?
Who’s not?”
Data can also be helpful in informing politicians and/or the public about needed policies.
Interviewees discussed garnering support specifically for healthcare-related policies and for increased
funding for jails. An individual with involvement in the North Carolina Sentencing Commission pointed
out how useful the Commission’s data have been in informing and garnering support for state sentencing
policy: “[Data are] a good way to convince representatives and senators that we need the money for the
state, and without that research base you would never get it.”. Jail data would also be helpful in gaining
support to fund additional services for jail populations. A local criminal justice employee in North
Carolina said: “I think if they knew how many people came to the jail with relatively minor offenses, with
underlying issues, behavioral health issues, if they knew how many people came, I think that would be
very helpful to supporting more community-based mental health and substance use treatment facilities.”
With regard to the importance of data collection, two individuals described how data on
healthcare and outcomes in jails could facilitate legislative change in the state. One correctional
healthcare provider and researcher discussed a recent bill defining certain health protections for pregnant
women who are incarcerated in North Carolina. The bill, known as the Dignity for Women Who are
Incarcerated Act, prevents the use of restraints (or “shackles”) of any kind on people in detention that are
pregnant from the second trimester to six weeks postpartum. Because jails lack systematic data collection
on pregnant women, actors lobbying for the bill relied mainly on anecdotal data, necessitating significant
lobbying efforts for legislative change:
“Fast forward at the end of last year, fall or winter of last year, the Dignity Legislation
came back up again. There’s a couple of different interests pushing this along, and the
North Carolina OB-GYN Society, at my urging, joined in with this. It got picked up by a
member of each party in the House, who really championed it. There was a lot of behind-
the-scenes work trying to lobby for law enforcement support of this bill”.
A correctional healthcare researcher described how the availability of data on pregnancy
outcomes from jails directly contributed to a countywide policy shift to avoid the arrest of pregnant
women in their third trimester:
“Cook County [Illinois]… showed [data] to their sheriff… the number of incarcerated
people who were in their third trimester, so that people wouldn't have to give birth while
in custody. The sheriff changed the policy and, in short, said other than the women who
are arrested on certain high-level charges, if a pregnant woman gets arrested, we’re
gonna release her off her charges if she’s in her third trimester. That was a good story…
about the value of data being able to advocate for change”.
Interviewees also discussed how data could be used to evaluate the effect of policies. A California
research expert explained how jail data have been used in that state to evaluate the effect of state Justice
Realignment policies:

15
 “Proposition 47 has happened in California, which changed some felony offenses to
misdemeanors. We can track on a graph Prop 47 happened, and now we have this
decline. That’s been useful…Showing shifts that are occurring in sentenced and non-
sentenced is helpful, especially when people start talking about pre-trial programs, and
people start talking about bail and bail reform…. We can be able to say in 2007 was our
highest incarcerated population. We’ve reduced that by over 30 percent. Being able to
show the impacts of legislation and system changes has been helpful”.
Even the most basic data on daily jail populations can allow policymakers to understand the impact of
new bail policies on pretrial detention, the impact of local efforts to reduce jail populations in response to
COVID-19, and changes in incarceration practices when new judges or District Attorneys are elected to
office.
A specific variable that some respondents stated would be helpful for informing policy is the race
and ethnicity of jail inmates. Information on race and ethnicity is important for understanding the racial
disparities that exist in the criminal justice system and for developing and evaluating strategies to improve
racial equity. One correctional health provider stated,
“We need more data to look at actual outcomes and morbidities and inequities and
disaggregation by race and ethnicity. We definitely need more data to try to help move
the needle, but even just showing that the problem exists can be a powerful tool for
making policy changes”.
Currently the North Carolina Department of Health and Human Services requires jails to report the
demographics of their inmate populations monthly, including gender and race/ethnicity. However, it is not
clear how jails or the prison system assign racial categories. Another correctional health provider
explained the difficulties in interpreting prison system data:
“If you look at the DPS [Department of Public Safety] statistics website, you will be
horrified at the categories that are offered to you as options for ethnicity and for race.
It’s totally unclear how those are assigned to people, whether it’s eyeballed by someone
or self-report or some combination thereof, and doesn’t account for individuals who
identify as being of mixed race. There are so many problems…. In the context of the
carceral system, there are so many questions about racial and ethnic disparities. If we’re
measuring race in a way that doesn’t make any sense, none of our conclusions are gonna
be all that helpful”.

2. Jail Data Can Inform and Improve Healthcare for Jailed Populations
Out of many potential uses of a State-run database discussed during these interviews, participants
spoke the most frequently and descriptively about potential uses in healthcare. Healthcare in North
Carolina local jails is highly heterogeneous, with some jails contracting with private companies to provide
all their healthcare, some providing the services directly, and others partnering with their local health
department. Although all local jails are technically overseen by their local health department, this
oversight is typically minimal, consisting of a written agreement on an annual health jail plan. Little direct
supervision or monitoring of jail healthcare is conducted by the health department. Additionally, research
has demonstrated that jail populations have higher prevalence of many chronic and acute health
conditions than the general population, but that jails are often ill-equipped to detect and address health

16
conditions due to limited resources and the short-term nature of most jail stays (Pollitt & Woollard,
2019). Respondents in the current study spoke about significant limitations in jail healthcare and how data
would be helpful in addressing them.
Local Jails and Counties Currently Have Inadequate Data on Jail Populations
Although jails already document a fair amount of health information from inmates in medical
records for the purpose of providing healthcare, the information is not collected with the purpose of
broader population-level analysis that may support evaluation, policies, and programs. Health information
documented during jail intake health screenings and later physical exams often includes basic vital
statistics, lab tests for pregnancy and infectious diseases, self-reported inmate medical history and current
medications, and inmates’ answers to jail-administered screening questionnaires on topics such as mental
health, substance abuse, and suicidality.
Entering and aggregating such data points across a single jail or multiple jails would be one way
of trying to understand broader population issues. Unfortunately, there are significant challenges in
creating a usable dataset from the information already collected. First, different jails use different
screening instruments for the same conditions, meaning that the results are not comparable. Often, the
screening instruments themselves are inadequate. As one respondent with experience working in in
multiple local jails in North Carolina described:
“The screening forms are crap. They’re that standard form that somebody pushed out
about 40 years ago that, to this day, jail administrators think is the standard form that
you have to use, and it is not. It needs to be tailored for your jail, for your demographic
in your county, in your region. They’re crap. The policies that they write to go along with
the mental health are crap because, again, it was somebody’s best guess 40 years ago”.
Additionally, respondents said that information that is self-reported by jail inmates may be
inaccurate. During the main health screening at jail intake, many inmates experience stress and confusion
that may lead them to omit or misreport some of their health information. For example, an inmate may
forget to list one of the medications they are currently taking or may decide not to report current drug use
due to fear of incurring additional criminal charges. One jail health provider described that they are
usually able to collect more accurate data when they ask questions later in the jail stay, at the “history and
physical” exam:
“It’s just almost the same questions as we did at the receiving screening, the intake
process, but they’re a bit more settled. They know they’re gonna be there. They tend to
know what’s happening with them legally-wise, so they’re a bit more stable, I feel like, by
day 14.”
Unfortunately, these potentially more accurate data are collected after the vast majority of jail inmates
have been released, and so are only available on the unique few who have longer jail stays and do not
represent the broader population. Due to these and other problems with data in jail health records, a
concerted, coordinated effort is needed to collect quality healthcare data across jails using standard
measures and taking place separately from the stressful intake process.
We asked respondents to list the health variables they thought should be collected for a North
Carolina jail database, and we also read a list of potential variables to assess the importance of each. The
health variables respondents felt would be important included: inmates’ history of substance abuse
(particularly opioids) and mental health conditions and treatment; chronic diseases such as heart

17
conditions and diabetes; acute conditions such as injuries and suicidality; women’s reproductive issues
such as pregnancy and menstruation; and infectious diseases such as HIV and Hepatitis. They also felt
that characteristics of jail healthcare facilities should be collected to understand variations in jail
healthcare services.
Better Jail Data Could Inform Budgeting and Treatment of Health Conditions in Jail
Respondents listed several ways in which accurate information on health conditions and
healthcare would be useful for jails themselves and for others working to improve health in jail
populations. First, individual jails and counties would benefit from accurate information on inmates’
health conditions and treatments in planning their healthcare budgets. One former jail administrator
described the difficulties of budgeting without good data:
“I did my budget myself. Budgets in jails are squishy. It’s kind of a best guess, ‘Let me
take an average of the last three years, but I’m really not sure’. If you ask jail
administrators what are the biggest health issues for their inmates, a lot of ’em are gonna
guess, ‘Well, I don’t really know’. Well, if we knew how many of ’em had syphilis, or if
we knew how many of ’em were HIV-positive, not who but how many—that’s a really big
part of the jail budget is healthcare. If I know that over the last three years I’ve had this
many inmates with heart conditions, and I know that I can quantify how much they cost
me, then I know how much to budget for, or I know how many nursing staff that I need, or
I know a whole slew of other things.”
Jails would also benefit from healthcare data they could access over time and across jails to
support care provision for high-need populations. Individuals entering jails, particularly those with
complex medical conditions, generally cannot provide jail health providers with all the information
needed for their care. They face a similar challenge at release from jail, lacking information about their
treatment in jail that could inform their care in prison or the community after release. The North Carolina
prison system manages an integrated electronic record system across all facilities where medical, safety,
and administrative information is entered and accessible to anyone working with a given inmate. Jails, on
the other hand, are usually completely unaware of a new inmate’s needs at intake, even if they were
thoroughly documented by another jail nearby or during a previous prison incarceration spell. One former
jail administrator described a situation in which a new jail inmate severely harmed himself during intake,
at severe cost to himself and the jail, that may have been avoided with a better information system:
“We save him, and we call our mental health provider to do some telepsych with him.
The minute we tell the doc who this guy is, and he said, ‘Oh, my God. You didn’t let him
in your jail, did ya?’ I said, ‘Yeah, we did.’ He said, ‘Any other jail won’t let him outta
the sallyport. They process him while he’s in the car because they’re gonna send him
right to central prison for safe keeping because he is a cutter and an inserter, and
everybody knows it.’ I said, ‘Well, how does everybody know it because he just came
from a neighboring jail,’ and the doctor said, ‘Yeah, he tried to do it over there last
month.’ I didn’t know…. Well, he’s a paranoid schizophrenic. He presents well when he
needs to, and then the minute you turn your back, he’s gonna open a vein and try to bleed
out. When he goes to prison for prison sentences, they’re very familiar with him. My
doctor was very familiar with him, but I didn’t know ’cause there’s no database to tell
me.”
Because they lack data, a small number of well-resourced jails spend time and money to develop
their own unique systems for collecting and analyzing the information they need. One North Carolina

18
court employee described having to create a detailed, “homegrown” system for tracking the information
they needed on criminal defendants released on pretrial, as part of an effort to reduce the number of
individuals with behavioral health needs being held in the jail. The employee illustrated how the
information collected through the homegrown data system is superior to that collected through the daily
jail administrative dashboard:
“We also created, again, a homegrown booking sheet that provides information about
individuals at the time of booking that relates to behavioral health, substance use, mental
health. That information is also then documented for us so that we can access it and look
over time. The daily [jail administrative] dashboard, with all the demographics of the jail
population, is just a daily. It doesn’t keep data over time. We would really like to have
that. Our behavioral health document does. We can, say, look, for the last three months,
how many people came in identifying with a substance use issue or an opioid use
disorder. We can get those numbers.
…Our pretrial case manager [does] an intake with everybody at the jail for booking….
He does an in-person interview with everybody. We can compare what we’re receiving in
terms of the [standard jail] behavioral health reporting with what he’s seeing in terms of
his in-person interview, which usually occurs the next day, and tends to be—people are
more willing to disclose at that point than they might be right at the booking stage with
the detention officer. That’s our best way of comparing some of the data.”
When asked about what it would be like to make decisions about the jail and pretrial population
without the data from the homegrown system, the employee said:
“Well, I think the consequences are that we aren’t able to measure the effectiveness of
our programs if we don’t have reliable jail data. Our office department exists to try and
safely reduce pretrial incarceration rates. Also, to make sure that we are reducing the
number of people with behavioral health issues in the jail and that we are shortening
lengths of stay by our interventions, and that we are keeping people healthier and safer
by our interventions, and that our pretrial reforms are working. If we don’t have reliable
data, we have a hard time measuring and improving the efficacy of our programming….
Well, and I think the other thing is that if we don’t have accurate information about
behavioral health issues, then we risk not being able to provide interventions to
individuals that are critically in need of them.
….We’ve created these homegrown ways to get some of the data that we need to evaluate
our programs and also to make sure that the county’s investment in this department and
the work we’re doing on justice reform issues is ultimately cost effective and doing what
it’s intended, which is to safely keep people in the community and to reduce disparities.”
All respondents stated that the lack of data on health conditions and healthcare in North Carolina
jails is a barrier to understanding and treating this population at a public health level and to advocating for
State resources and policy changes to help improve health conditions for jail populations. A corrections
official in California described how having routinely collected jail health data enables local sheriffs in
that state to demonstrate the need for more State funding to address health issues:
“Jail Profile Survey development has been largely driven by the sheriffs’ needs and their
needs to show capacity issues or programming issues…. We collect a variety of things
about mental health so number of open mental health cases each month, number of new

19
 mental health cases each month, all the way down to number of medical encounters like
total number of dental encounters”

3. Data are Urgently Needed to Address Specific Health Needs in Jail Populations
Four health conditions (mental illness, pregnancy, opioid use, and COVID-19) were discussed
most frequently during interviews and are described below to illustrate how better health data may help
justice-involved North Carolinians.
Untreated Mental Illness is a Major Challenge for Jails
Respondents spoke about mental health as one of the most concerning health issues facing jail
populations in North Carolina and the country. Mental illness is highly prevalent among incarcerated
populations, in large part because a severe lack of mental health services in the community can contribute
to criminal behavior, and ultimately arrest and incarceration, among the mentally ill (Esquibel, 2021).
Several respondents also pointed out that spending time in jails and prisons likely exacerbates rather than
ameliorates mental health conditions, potentially leaving inmates worse off and at higher risk of
recidivism after release. As one former North Carolina re-entry case manager shared, “I think if we had a
better mental health system a lotta people that are in jail wouldn’t be in jail. Our mental health system is
probably our weakest system in North Carolina.” One North Carolina non-profit employee explained that
understanding the number of mentally ill jail inmates is an important first step in addressing the problem:
“I think national average 20 to 25 percent of the people in the jail have a serious and
persistent mental health issue. It’s not gonna be addressed in the jail setting. It’s only
gonna make things worse, right? You start doing the math on that $80 a day. How long
are they staying there, what could we have done differently to pull them out of that system
or avoid that system entirely and get them community-based treatment? You have to
know those numbers—20 percent here, or maybe there’s 40 percent between mental
health and substance use.”
The Texas Commission on Jail Standards can serve as an example of how data on mental illness
in jails can inform programming and intervention. A Texas sheriff described how the state uses mental
health data to identify and understand “frequent flyers”, who are individuals who repeatedly require
jailing or emergency services at high cost to the community:
“We currently are doing a lot of work in Texas on the mental health, the forensic
population in our county jails, which is a problem all across the nation. Probably one of
the most costly parts of running a county jail is providing housing and the medication
and then treatment for those that come into our custody that suffer from mental health
issues. Currently in Texas, we have a waiting list of almost 1,400 inmates that have been
court-ordered by a district judge to receive a mental health evaluation. The wait list now
is running about a year. That population, the forensic population particularly in the
criminal justice system, is huge.
Now, with that being said, it’s no longer about just the need for beds and a behavioral
healthcare system. There are other outlying resources that need to be taken into
consideration that are important for when you deal with people that have mental health
issues…. More importantly, to improve the behavioral healthcare issues that we face in
running our county jails because that is a population that is very trying at times because

20
 of the manner of their behavior, and a danger to themselves as well as to my staff….
Well, one of the things that we have over the past couple of years implemented, and it’s
slowly beginning to come together, is to find out what we call, the ‘frequent flyers’, and
track the data of where these people are going in and out of the behavioral healthcare
system, and what their past histories, criminal histories as well as their medical histories
have been. The data collection that’s being collected and now being analyzed can be
used by our health and human services agency, a state agency, to try to determine how
can we better serve and divert those people from the criminal justice system.”
An example of how jail data have already been used to address the mental health crisis in North
Carolina jails is the Stepping Up Initiative, a partnership between the Council of State Governments
Justice Center, the National Association of Counties, and the American Psychiatric Association
Foundation to support counties’ capacity to collect and use mental health data to reduce the number of
people with mental illness in jails (Wrenn, McGregor, & Munetz, 2018). As a North Carolina non-profit
employee described,
“The Stepping Up Initiative is actually a national initiative. It started at least 15 years
ago it feels like, 10, 15 years ago. What that was about was reducing mental health issues
in the jail setting. They had to go through this data collection process of really
understanding the prevalence of serious and persistent mental health in their jail, and
then developing strategies to reduce that population in the jail.”
Nearly half of the counties in North Carolina have participated in this program to date, providing an
important resource that could be leveraged to provide similar data on a state level (Gangi, 2021). This
program is discussed further in Policy Recommendations for North Carolina.
Opioid Screening Data are Collected but Could be More Useful
Some respondents specified opioid use and opioid use disorder (OU/OUD) as particular health
information that jails should collect from their residents. Several participants noted that information about
opioid use, including treatment utilization, were collected from jails they or their colleagues work in.
However, the accuracy and completeness of OU/OUD data was often limited, as jails used inadequate
tools to screen inmates at intake, and OUD definitions used by jails were inconsistent with current
medical guidelines. Important consequences of poor screening measures were that jails and treatment
providers were unable to identify eligible candidates for treatment programs or to determine who requires
linkage to care upon release from jail.
A non-profit employee focused on substance abuse treatment in North Carolina felt that better
screening tools could provide data on the number of eligible candidates for treatment programs.
According to this individual, this data could then help improve the program by highlighting the factors
contributing to low turn-out.
“It is not always easy to work with jails. They have a lot of employees and they all have
their different skillset, and they all have their own capacity. You don’t want to harass
them, but you really would like to know how many people did you guys screen and how
many screened as having an opioid dependence. Then we could say—if we knew that they
had screened and there was 100 people, but we only had ten come to our class the next
week, day, month, well, then we’re not doing our job, right? How are we missing those
90 people in between? What can we do? Do we have to advertise better? Do we have to
have different hours? Do we have to talk to the jail about having two classes? There’s a

21
 lot that’s missing in the middle. A lot of it comes to just hoping that your partnership is
strong and that they’re dedicated to what you’re delivering that they’ll get people to
come to the class.”
Barriers to OU/OUD data collection among incarcerated populations in general included
insufficient technology and computer system capabilities to understand inmate drug history and to
connect inmates to healthcare providers for medication treatment. The same non-profit employee
explained,
“Jails also don’t have access, at least not all of them, or probably even the majority of
them, they don’t have access to treatment data. Let’s say, I came into the jail, I was
booked in, and they asked me the questions and they would say, ‘[Interviewee Name], are
you under any medication for opioid maintenance?” and I could say, “No.” They could
say, “Do you have an opioid dependence?” and I could say, “No.” I could say whatever
I wanted to say. They have no idea whether or not this is true, and that is a challenge,
because they can’t help me, right? If they had access to that information, they could say
we see that [Interviewee Name]’s been on Methadone for a while and we don’t want her
to stop taking that Methadone. How do we connect with her healthcare provider to make
sure that she can still get that medication while she’s incarcerated? They don’t have that
at a lot of jails just because those systems just haven’t really been created. You’re
talking about computers talking to each other and really confidential sensitive
information.”
Incarceration is Risky During Pregnancy, but Many Pregnancies are Undetected
Another health topic interviewees felt should be more extensively investigated in jails was
pregnancy. Some jails test women for pregnancy at intake but others rely on women to self-report their
pregnancy status, which may be unknown to them at the time. One jail health provider stated that the jails
where she works test every woman for pregnancy at intake, and she can regularly run reports on the
number of pregnant inmates for health administration purposes. But an advocate for justice-involved
individuals across the state of North Carolina said she lacked the information on pregnancy that she
needed for her job. She explained,
“I focus on the incarceration of women and girls a lot. Who’s pregnant is huge, and we
never know that. Jails don’t do pregnancy tests. Just arbitrarily, they don’t do them. They
rely on folks to self-report. We also don’t know, if someone’s pregnant, how long they’re
stayin’, what they’re comin’ for. Are they receivin’ treatment? All of that. Unless a sheriff
is very cooperative and says, ‘Hey, there’s a pregnant woman in my jail,’ we don’t know
that she’s there. That’s important for another part of my work because we have the North
Carolina Pregnancy Bail Fund where, if someone is pregnant in pretrial, we will bail
them out, but we never know that they’re there because there’s no data to show.”
Several respondents thought that pregnancy would be an important variable to include in a state jail
database in North Carolina, including a respondent from Texas who said that the Texas database collects
that information.
One correctional healthcare provider spoke in detail about her unsuccessful efforts to learn the
number of pregnant women in North Carolina jails. Many North Carolina jails send their pregnant
inmates to be held in prison due to the possibility that they may need medical care the jail is unable to
provide. This practice is known as “safe keeping” and is used widely not only for pregnant inmates but

22
also for those with other medical or custody situations that may put them in danger in jail. Despite these
women being held in prisons, which generally have better record-keeping and data collection than jails, it
is still a challenge to understand the scope of pregnancy in jails. She describes:
“I don’t know that there are even any systematic records about how many people are
sent from jails, thinking about safekeeping in particular…. Pregnancy is not recorded as
an administrative or demographic characteristic for most folks. The data that I have
about the populations that we take care of is largely data that I have gathered or paid
research assistants to painstakingly abstract data from the medical record. It’s not data
that’s collected in any sort of systematic fashion…. I don’t think that anything, like I said,
is being aggregated in a systematic way. Again, that NC DPS—excuse me—DPS
statistics website, I’ve definitely looked to see last year, what counties sent reproductively
aged women as safe keepers to give me some sense of the distribution and where folks are
coming from. That is an imprecise measure of pregnancy and certainly, doesn’t include—
there are certainly counties that don’t send folks as safekeepers. I think Mecklenburg is a
notable one where they don’t send folks—it’s a big county. Wake County doesn’t. I think
a lot of the bigger jail systems can provide that care. Then there’s just zero data. Right? I
can tell you I suspect that the safekeeping tells us how many smaller jails have pregnant
people in them and how many pregnant people go through those small jails. It is really
unclear how many people go through those other—pregnant people go through those
other jails. It’s possible that someone is recording that there, but I certainly don’t have
access to it.”
Understanding the number of pregnant women in North Carolina jails and the number sent to
prisons for safekeeping is important for ensuring that these facilities have adequate resources for their
care. This information is also important for designing attempts to divert pregnant women out of jail
during the high-risk prenatal period. The experiences and living conditions associated with incarceration
are often highly stressful and can lead to poor pregnancy outcomes. A correctional healthcare provider
explained how more data about pregnancy in jails may help:
“I’m also really interested in figuring out how we can not incarcerate people, and
knowing what counties are sending folks and where are pregnant people getting
incarcerated would be really helpful information for how to intervene on that. We have
great case managers and social workers connected with various programs at [name of
respondent’s clinic] who could provide referrals for residential treatment programs or
any number of other resources that would potentially divert people out of jails and keep
them from being sent to the prison. We really just don’t know where they’re coming from
or how frequent that is. It makes it much harder to target intervention and
programming”.
Better Data were Needed to Manage COVID-19 in Jails
Because the study period for this project coincided with the COVID-19 pandemic, we asked
participants about their experiences addressing or responding to COVID-19 in jail populations, and the
role of jail data in the response. Interviewees broadly recognized that more data would have informed the
management of COVID in jails, specifically data on COVID case numbers and positive and negative test
rates, in addition to incarcerated individuals’ ages and the underlying health conditions that are known
risk factors for COVID-19. When asked what type of COVID-related data would have been helpful in

23
their work during the pandemic, an advocate for justice-involved individuals in North Carolina
responded:
“Daily county-level jail populations. Obviously once they started testing, testing rates
and positive and negative rates for each jail. Information about the age of the people in
the jails, like a yes, no on whether they had underlying health conditions if that were able
to be public. Then I think anybody who—any information about bail that would be
relevant to people who could be bonded out quickly or should be reconsidered for
release. Yeah, I guess a breakdown of the people who are in jail and what their charges
are to understand who would be the most easy to get released.”
Not only were these data points unavailable from many jails, even basic COVID information like
the number of cases in a facility were often lacking due to some sheriffs not sharing data. Some sheriffs
were unwilling to share information about testing and COVID cases with those outside their jail, and
some were unable to devote the extra resources necessary to collect and report testing and COVID data. A
North Carolina advocate commented: “I think some jails have been more open about what their stats are
in terms of testing and rates and also just total numbers currently in their units. On the whole, it’s been
really patchy. It’s been up to the sheriffs to decide whether they were gonna be public about it.”
Another challenge in measuring COVID in jails is poor understanding and standardization of key
COVID indicators. A North Carolina jail healthcare worker discussed this challenge related to collection
of data on COVID-19 symptoms:
“I was doing COVID screening, and they wanted to know how many were symptomatic of
COVID. A lot of ’em said, ‘I’ve got a sore head.’ A lot of ’em said, ‘I’m feeling tired.’
You know? All these sort of symptoms. We didn’t count them at all. Even loss of taste,
loss of smell, things like that we didn’t count because they weren’t quantitative. We had
to measure the blood pressure. If they had increased blood pressure, increased pulse,
increased temperature, increased respiratory, it all had to be clinical information
because you just can’t go on anything that’s just their story, unfortunately.”
In contrast to North Carolina, California was able to modify its existing Jail Profile Survey and
develop additional surveys for special data collection and analysis of daily jail counts to understand the
impact of COVID in jails. A California jail expert described this process:
“We ask the average daily population so that would be a weekly daily population total
number of bookings, total number of releases, and then of those releases how many of
them do you estimate were releases to mitigate the impact of COVID-19? It wasn’t
necessarily they were released because they had COVID. They were released to make
space. They were released to better handle the situation in the jails, and then we screen
that data weekly. We also developed what we called a COVID-19 Cases and Local
Detention Facility Survey, also a weekly survey.”
The COVID Prison Project (CPP) provides an example of how COVID-19 in jails could have
been understood with better data. Public health researchers established the CPP in March 2020 to collect
and monitor COVID-19 statistics from correctional facilities across all 50 states. The project collected
COVID-19 case numbers, testing, deaths, and state vaccination distribution plans for their justice-
involved populations. The CPP also tracks COVID operational polices related to quarantining, movement
restrictions, isolation of inmates, and face mask requirements. One respondent involved in the CPP
described the motivation and purpose behind the project:

24
 “(COVID) was going to be so much worse for incarcerated populations for a couple of
reasons, right. Jails and prisons tend to be closed settings where there's a lot of people in
space, where it's very hard to engage in social distancing. People who are incarcerated
tend to be more sick than the general population, which puts them at risk of suffering
more severely from COVID-19. Really, those things made me very worried about what
was going to happen. We started looking around at what we knew. What we knew, what
we didn't know, what we could find out. All this information was coming out about those
cruise ships. We had very clear understanding of how many people were on those ships,
and where those ships were at. I was, well, can we find this information for incarcerated
populations?.... I knew that it's very hard to get this information out of jails and prisons,
right? Very hard to get information on health conditions, or even what the health policies
might be. I was dubious that it existed, so it was surprising that a fair number of states
even in March or April had launched public facing dashboards….
We've been collecting data on 53 prison systems and some jails across the country for, I
don't know, close to 18 months. We've been taking everything that they're giving us, so
we've been curating a data set behind the scenes, and then we launched a public facing
dashboard that has been pretty widely used by researchers, and the media and advocates
to keep prisons accountable during the pandemic.”
Because state prison systems have centralized data collection systems, the CPP was able to begin
collecting data from states from the very beginning of the COVID pandemic to increase public and
scientific understanding of how the disease was affecting prisoners. A jail data reporting system in North
Carolina could have allowed for the reporting of jail data as well, contributing to a more complete
understanding of COVID in correctional populations.
These findings related to COVID further illustrate the importance of jail data, and how the lack of
jail data collection in North Carolina may impede quick and accurate responses to public health crises.

4. Implementation Factors May Impact the Utility of State Jail Database in North Carolina
In addition to identifying the ways in which a North Carolina jail database would be useful to the
state and its localities, we also sought to understand issues related to implementation that may inform the
design of the database. We asked participants to describe the resources and funds they felt would be
needed to create a useful database and the desired outputs of a jail database. We also asked participants to
describe facilitators and barriers to jail data collection that may influence the success of a large-scale
effort. In this section we describe our findings related to the ideal implementation of a state jail database
in North Carolina: 1) jails (particularly rural and smaller jails) need resources and staffing to collect and
report data accurately and consistently; 2) oversight mechanisms by the State are needed to ensure
consistent reporting from all jails; 3) data must be collected in a standardized way across jails; and 4) data
must be presented in a way that is readily accessible and understandable by stakeholders and the public.
Additional Resources are Necessary for a North Carolina Jail Database
Several participants stated that adequate staffing was essential to jails’ ability to collect consistent
and accurate data. Unfortunately, a large number of jails are in need of more staffing. When asked about
the current feasibility of routine data collection in jails, a local jail employee in North Carolina claimed,

25
 “You would have to have someone that’s actually dedicated to collecting that [data] and
putting it into one place. In so many facilities, because you have such staffing issues, you
have people that are doin’ multiple jobs. Things will fall through the cracks, so to speak.
I think one of the greatest challenges would be, of course, the software to be able to do it,
but the staffing to actually collect the data and put that in place and say, ‘Okay, we need
to start doin’ these particular numbers.’”
Even in Texas, where jail data collection is required for the Texas Commission on Jail Standards, some
jails lack sufficient staff to collect and report the required data. When asked what additional resources
would be helpful in supporting jail data collection, a Texas sheriff responded:
“If we’re talking about long-term data collection, it would be nice for, particularly in our
rural areas, if we were allowed to have a designated person, employee, that could focus
on just that instead of all of the other duties that they are required to perform during
their 8 or 10 or 12-hour shift.”
Some interviewees mentioned that generally, smaller and more rural jails receive less funding and may be
less able to collect data than larger, urban jails. One North Carolina advocate explained, “We can just
dream about what’s happenin’ in the small lesser resourced counties. I think, for me, the dream is to get
as much information about these smaller places to see, how bad is it there? What’s happening? So that
we can begin to deploy some remedies.”
Other interviewees described a profound need for improved technologies and software in jails to
properly collect and analyze data. A former administrator for multiple North Carolina jails shared that
jails have different types of technology, some of which may make standardized data collection difficult:
“The challenges are gonna be the databases that they use. They’re all so different
depending on what vendor they have. The program that I had, I had to pay a couple a
thousand dollars and wait a couple of months for them to write the program for the
report that I needed as opposed to the previous jail that I had, where I could go in and
design my own report from that platform to spit out whatever information I wanted out of
it.”
Smaller jails may suffer more than larger jails from lack of adequate technology and support for data
collection and reporting. One correctional healthcare provider in North Carolina explained that private
correctional healthcare companies often have centralized resources to optimize data collection and
management, but these companies may be less willing to contract with smaller jails:
“The [private healthcare] contracts themselves are very much so that—I mean, [name of
company] wouldn’t go for a small site contract because they don’t agree with—really,
they don’t like anywhere that doesn’t do 24-7 care. Some of these poorer counties, for
example, just can’t afford that. They might want to, but they just can’t afford it. These
contracts very much define the level of care that’s being delivered, and therefore defines
how much information’s gonna be gathered. If you use a small contract to just, say, a
local provider and a nurse or something like that, then they’ll have paper format
probably, and the data just sits there. Then it’ll be gathered there. Maybe your rural
communities are not going to be able to gather information very well. Unfortunately,
even to watch some of the patients, [laughter] and they say, ‘Well, I was in such-and-
such a jail, and don’t send me back there!’”

26
 Technology even presents a barrier to jail collection in California, which has been collecting jail
data for its Jail Profile Survey since the 1980s. A California correctional expert explained,
“We’re always behind in technology and doing updates, maintenance. Revamp to make it
less staff intensive for us with technology can be challenging. I’d say we’re behind. Even
though I talk about having a database and doing screening it’s still a very manual
process for us, so anytime something’s added it changes the system. Any time we start
flagging or really focusing on the variable, it’s a very manual process.”
A Texas sheriff noted that technology is more of a barrier to data collection in smaller counties, which
may keep entirely paper records:
“The capacity to collect data varies by county. Some counties do not have a jail
management system, and they are counting these things by hand every month. They’re
literally going through paper and counting. Those tend to be the smaller counties, and as
counties update jail management systems, sometimes the ability to report certain data is
lost. That’s challenging, but overall I find that most counties are able to report.”
Oversight Mechanisms are Needed to Ensure Consistent Reporting
States can pass legislation to require consistent reporting of data from all jails, but there is a risk
that some sheriffs and administrators will fail to comply. Currently, North Carolina jails are required to
report basic monthly population counts and basic demographics such as sex and race to the North
Carolina Department of Health and Human Services, but not all jails comply with even this simple
request. As one former jail administrator said,
“Even though they mandate it, there are—I know of three jails that refuse to report. They
just don’t do it. When they get inspected, they get a ding for it, but there’s no teeth to it.
In my opinion, if one jail doesn’t report in North Carolina, then it makes all the data
almost worthless because you can’t say, ‘Here is what all the jails are showing’.”
State mechanisms for overseeing jail data reporting need to include incentives for compliance or
punishments for lack of compliance. A Texas sheriff explained that the Texas Commission on Jail
Standards has the authority to implement punitive measures when sheriffs do not report the required data
from their jails:
“Well, for us, honestly, we’re answering to the Texas Commission on Jail Standards, so if
for whatever reason, if that information continues to not flow to where it needs to be
flowing to, you’re gonna get a letter from our executive director… with an inquiry as to
why you’re not supplying that information. You’ll be given a certain number of days to
correct that. Then if you can’t correct it, then you’ll be summoned before the commission,
which meets quarterly here in Texas, to appear before the jail commission to explain
what your issues are. Then the jail commissioner will make a determination of—I hate to
use the word punishment but—the jail commissioner will make a recommendation of how
to correct the situation.”
Good relationships between the state and sheriffs and administrators can also encourage
compliance and makes the oversight process easier. The Texas sheriff noted that professionalism and
helpfulness of their oversight commission helped in the reporting of jail data:

27
 “Because we are, at the jail commission, very user-friendly. They do everything they can
do to assist us in operating our county jails, but they also have a statutory responsibility
that they have to make sure they carry out their duties…. They’re there to help us, not
hurt us. We understand that. Every now and then you have to reach out and tap someone
on the knuckles and say, ‘Hey, where’s your paperwork?’ or ‘Where’s your report?’ and
get their attention. Then they do it in a professional manner, and I respect that.”
A California corrections official also spoke about the importance of professional relationships in
encouraging compliance with jail data reporting, in the absence of enforcement mechanisms:
“We talk about the Jail Profile Survey being voluntary, and sheriffs are doing this
because they see value in it…. We rely on creating good relationships with the data
reporters and with the sheriffs to accomplish what we need to with the jail profiles. The
staff… in charge of the Jail Profile Survey is well aware that the relationships matter.”
Data Standardization and Presentation Affects the Usefulness of the Data
We also asked participants about the ideal presentation of a jail database. One of the foremost
emphasized points by participants was the need for an online, interactive platform that would provide
easy access to information on jail populations through a visually appealing, user-friendly interface. One
criminal justice advocate in North Carolina said “I like infographics. I need complex ideas and
information to be presented in simple-to-understand ways. If you got a picture, a graph, a few words, I’m
good.” This participant highlighted the importance of having a user-friendly approach to data presentation
since, in their words, “You’re gonna have people who understand all of the legalese and the jargon and
the system, and then you’re gonna have people who wanna know the bottom line.” In order to
accommodate the wide array of groups to whom jail data must be made accessible, it should be presented
in its entirety while highlighting the most important takeaways for less-experienced stakeholders.
In an interview with a jail research expert from California, the participant briefly described the
operability features and analysis capabilities of California’s jail data presentation platform. The raw
database created through the Jail Profile Survey is not intuitive or user-friendly for those wishing to
understand trends and make inferences. For that reason, the state took an extra step of creating an
interactive dashboard, allowing data to be filtered by variables, counties, and other demographic
information to fit the needs of legislators, jail administrators, academic researchers, and public citizens.
They also created a summary report to present information that is commonly requested by users.
The second theme related to data presentation speaks to the need for standardization in the data
reported across counties. For effective cross-county analysis, there must not only be standards for
variables and collection methods, but also for the way in which data is presented to stakeholders and the
public. In an interview with a North Carolina advocate, the participant noted that “being able to toggle
between statewide and local level information would be important... and then being able to compare
jurisdictions so that similarly sized cities or something can be compared.”
North Carolina has no such standardization across local jails at this time, which is a barrier to
compiling a meaningful database. A North Carolina policymaker expressed concern over this lack of
standardization:
“There does not seem to be continuity in systems and in processes across the board. It
seems like, although there may be some requests for certain information to be collected,
it doesn’t appear that we have a network across the state when it comes to the

28
 Department of Public Safety that you can just easily access, even the data that is
collected. Being able to make sure that it can be leveraged ’cause in my opinion, what
good is data that’s being collected if it’s not able to be reconciled and put into some way
that you can actually review it? Right now, it doesn’t appear that we have a system that
really lets us readily tap into data and information across the board...”
Standardized collection of jail data would likely need to be imposed by the state in order for a state jail
database to be useful and meaningful. The Texas county sheriff described Texas’s data collection and
reporting standard and highlighted how “the collection of the reports that go into the Texas Commission
on Jail Standards are uniform across the state, so every sheriff has to send the same report.”

29
 Policy Recommendations for the State of North Carolina

The findings presented in this report illustrate multiple potential uses of robust local jail data, and
also point to reasons why jails may have difficulty collecting and reporting robust data. In this section we
describe some of the policy changes North Carolina should consider enacting to begin routine data
collection from jails to improve conditions for jail populations.
Begin collecting and publicizing daily jail counts immediately
Respondents discussed a number of different variables that would be helpful to collect from jails,
but a few emerged as highest priority. The highest priority variable discussed by respondents was daily
counts of jail populations, with additional specification of the number sentenced and unsentenced. The
recent experience of managing COVID-19 in correctional facilities illustrates the usefulness of basic data
on daily population counts, which can indicate facilities that may be at risk of overcrowding and need
public health assistance to manage social distancing and testing efforts.
Daily counts of sentenced and unsentenced inmates can also facilitate the evaluation of efforts to
decrease pretrial jail populations, a common goal among many counties interested in reversing mass
incarceration. California was able to use the simple daily count (sentenced and unsentenced) data point to
help evaluate one of the largest decarceration efforts in US history, in part because of its Jail Profile
Survey. When jails are not required to document the numbers of sentenced and unsentenced individuals
explicitly, it can be very difficult to ascertain this information based on other factors. One criminal justice
researcher illustrated how this has been a barrier in researching jails:
“If we’re looking to determine pre-trial, if somebody is within the facility, pre-trial or
not, there might be a release date or an admit date and a release date but not always
accompanied by, what is the reason for that? Are they serving? Is this admission related
to serving of a sentence? Is this pre-trial and those type of things? Not being able to
differentiate. We can see the person’s there and what are the dates but not necessarily in
what way they were being admitted and released or the other important circumstances
surrounding that.”
Florida’s “County Detention Facilities Average Inmate Population” reports provide a good
example of how to collect and share basic jail information that would be helpful to many stakeholders
(Florida Department of Law Enforcement, 2021). Even before more comprehensive criminal justice data
reporting became required through a new “Criminal Justice Transparency Dashboard”, Florida was
collecting daily count data from jail facilities using standardized reporting instruments and publishing
monthly reports on its Department of Corrections website. These reports showed each jail’s average daily
population for the month, number of inmates held pretrial for felonies and misdemeanors, and total
percentage being held pretrial. They also show statistics for the entire state’s jail population stratified by
age (juvenile vs. adult), gender, and types of charges and sentences. Similar reports would be very useful
to multiple stakeholders in North Carolina and may not require significant additional time or monetary
investment to initiate, as jails are already required to report daily counts to the North Carolina Department
of Health and Human Services. These same basic data are being used in Florida by counties to forecast
inmate bookings and releases and estimate the resources required for special populations (Ranson et al.,
2018).

30
 Although Florida’s basic jail reporting system has been operating successfully for several years,
the state has faced challenges implementing the more ambitious “Criminal Justice Transparency
Dashboard” required by state law in 2018 (Florida Statutes, 2018; Pantazi, 2020, 2021). Although the
state spent two million dollars to build and staff the database in 2018 and 2019, the Florida Department of
Law Enforcement (FDLE) has encountered significant barriers in standardizing measures across criminal
justice agencies and in actually receiving data from agencies. One local Florida journalist published the
following description of the law’s implementation two years after it passed:
“FDLE missed its first deadline to publish a database by Jan. 1, 2019, and then
legislators passed amendments that gave them until March 1, 2019, to start publishing
data as it got it. The amendments called on FDLE to publish a complete database by Jan.
1, 2020. ‘They’re woefully behind on their timeline,’ said Sen. Jeff Brandes, one of the
law’s sponsors two years ago. ‘It’s an area of significant concern. We’re trying to address
this. We want better data. We want to make better decisions.’
As the public monitors COVID-19 outbreaks in prisons and jails, many of the state’s
counties are not publishing how many people are in their jails or what charges people
are being held on. As some prosecutors, like Jacksonville’s, say they’re seeking to release
more people awaiting trial than they used to due to the pandemic, there’s no way to
compare bond amounts to what they were months ago. As citizens go out into the streets
to protest and sometimes face charges like unlawful assembly or resisting arrest, some
prosecutors are dropping charges while others aren’t. There’s no one place to compare
state attorney policies or how effective those policies are at reducing jail populations”
(Pantazi, 2020).
Based on Florida’s experience, North Carolina may want to “start small” with basic jail reporting
before launching an effort that would require numerous governmental entities to begin reporting standard
data on multiple variables. Collecting basic jail counts would be an important step for North Carolina and
pave the way for collecting further data in the future.
Create infrastructure for collecting high-priority demographic and health variables
In addition to daily count data, North Carolina would improve its jail data collection efforts the
most by collecting these additional high-priority variables: basic inmate demographics (race, ethnicity,
gender, and age), mental health conditions, opioid use and treatment, and pregnancy. With a relatively
narrow focus on daily counts and a few high priority variables, the state can spend time and resources on
ensuring the accuracy and completeness of data across jails without overwhelming under-resourced jails
with a large list of requested variables.
Data on the race and ethnicity of inmates is similarly helpful in helping counties understand
where racial disparities are occurring in the criminal justice system and evaluate efforts to improve equity.
Currently jails are required to report this information to the Department of Health and Human Services,
but it is unclear how racial categories are being assigned to inmates and highly likely that methods and
categories are inconsistent across jails. Comprehensive racial and ethnic categories must be standardized
across jails, and they must be assigned according to inmate self-report rather than an “eyeball” assessment
by a correctional officer or health provider.
Collecting data on the prevalence of health conditions in jails is beneficial to informing public
health strategies to improve health outcomes. Data are particularly needed to understand the scope of
mental health conditions and pregnancy in jails to garner funding for public health efforts. Many jails

31
already administer pregnancy tests to female inmates at intake; now, all facilities should be required to
test women for pregnancy at intake and to report positivity rates to local health departments and to the
State to understand and support maternal health in corrections. Most, if not all, jails administer a mental
health assessment at intake, even if it is just a brief question about suicidality. However, the majority of
the currently existing data are collected for the purpose of healthcare, and suffer from lack standardization
and potential inaccuracies. A standard mental health screener that has been validated for use in clinical
(ideally correctional) settings should be selected for all jails to administer uniformly to inmates, and
results should be reported to the state to enhance understanding of mental illness in jails.
The Stepping Up Initiative should be an excellent resource for information on integrating mental
health screening and reporting into local jail procedures. Glenda Wrenn and colleagues write that
Stepping Up requires jails to track four key mental health outcome measures, which can be used to
characterize mental health in jails and evaluate the impact of policies over time: the number of people
with serious mental illness booked in jail, the length of time people with mental illnesses remain in jail,
connections to treatment, and recidivism (Wrenn et al., 2018). Because nearly half of North Carolina
counties currently participate in Stepping Up, requiring all jails to use the Stepping Up measures for the
entire state may be the easiest method of ensuring standardization and compliance in this area (Gangi,
2021).
Allocate funding to support accurate data collection in under-resourced counties
Many jails, particularly those that are small or in rural areas, lack personnel and technology
required to collect complete, accurate data. Any State requirements for jails to report data must be
accompanied with funding to support the effort. Some interviewees called for a “designated person” to be
in charge of data collection and reporting at each jail facility; this probably does not need to be a full-time
position dedicated to data at most facilities, but rather could be a person who has few competing demands
on their time and maintains a constant enough presence to ensure consistency in data collection
procedures. Certain types of information would be most accurate if they were collected at a time that is
separate from the regular intake procedures, in order to minimize inaccuracies caused by the stress and
confusion of intake. State funding could provide jails with extra staffing to meet that need. Additionally,
the State would benefit from investing in a standardized database software with licenses for all jails,
which would improve data collection in smaller jails and make sharing and comparing data across jails
more seamless.
Some North Carolina counties have developed their own data tracking systems and technology to
improve operations and direct resources but have done so at considerable expense. State funding to
support the development of a functional electronic data tracking system in each county, or the adaptation
of an existing data tracking system from a model county, would support the uniform collection of basic
data and may translate into large savings at the local level. Many jails currently have electronic
management systems, but they differ across jails and are often unable to produce the information required
for reporting. Identifying the best of these systems and paying for licenses and computer data entry
stations for each jail would go a long way toward standardizing data collection (not to mention improving
general jail operations) in the state.
Structure for accountability
In order to promote accountability in data reporting by jails, North Carolina will need oversight
mechanisms. There are no oversight mechanisms currently in place nationally or at the state level that
hold states accountable for reporting jail data. Even a legislative mandate will need “teeth” in order to
achieve compliance from reluctant sheriffs. Examples of how states can impose negative consequences

32
for non-compliant jails can be learned from the Texas Commission on Jail Standards. One recipient
discussed that Medicaid funding is currently denied to jail residents, but that if the policy was reformed to
allow Medicaid reimbursement to jails, this may be an incentive for jails to report data: “Hey, if you want
to get your Medicaid reimbursement, then you also have to report these data measures. There has to be
some carrot stick to mandate data collection and data reporting from jails.”
Make data publicly available and understandable
The way in which jail data is presented is critical to how the data can be used. Jail data should be
compiled in a way that is user-friendly and not just accessible to those who work in the specific county
jail. In order to create greater public awareness surrounding the issues that relate to jails, data need to be
presented in a manner that is digestible to the public. This means having a collection system that
incorporates visuals or infographics so that the data published is not just a set of numbers, nor is it
compiled in the mix of jargon. This also means that communications specialists and web designers should
be involved in the creation of the database. California, Texas, and Colorado have excellent examples of
jail databases that are user-friendly and interactive.

33
 Practical Recommendations for Local Governments and Jails

In conducting interviews for this project, we encountered several respondents with an interest in
improving data collection in their own local jails. Even without a coordinated State effort to collect and
aggregate data on jails, many local governments and jail officials would benefit from improving their own
data collection practices. Improving data collection in individual jails could have standalone benefits,
such as making it easier for jails to comply if the State does mandate reporting in the future and making it
possible to share data across localities and states for research purposes. These recommendations are
designed to improve the data collection capacity and rigor of local jails in North Carolina.
Create a system for recording daily population counts and other important variables
Jails that do not already do so should establish standard procedures to record daily jail population
count data electronically, ideally including sub-counts of those who are sentenced and unsentenced. In the
absence of an electronic jail management system, or even to supplement one, these data can be tracked in
a simple spreadsheet over time. Additionally, jails and local governments can decide on additional
measures that would be the highest priority to track over time and add procedures for collecting those.
Example variables that may be useful to collect are race and ethnicity, for the purpose of evaluating
equity in jail conditions; prevalent or expensive health conditions (e.g., diabetes, heart conditions, HIV)
for the purpose of budgeting and planning for health services; and variables that are of interest to local
community stakeholders or the court system such as bail amounts or social service needs.
Adopt valid screening tools for assessing mental health and opioid use disorders and treatment
In the absence of standardized state screeners for mental health conditions and opioid use
disorder, jail officials should talk to their local or state health department to learn options for valid
screeners for correctional settings. Adopting a valid screener will improve the jail’s ability to accurately
assess inmates’ mental health and opioid treatment needs and plan for treatment contracts and resources
accordingly. Adopting the same screener used by the local or state health department would further allow
the jail’s screening data to be compared with local or state data.
Test all reproductive aged women for pregnancy at jail intake
Finally, all jails should administer lab pregnancy tests to all reproductive aged female inmates.
Many female inmates are unaware that they are pregnant or do not share their pregnancy with jail staff,
leading to missed opportunities for important prenatal care. Many jails in North Carolina choose to send
their pregnant inmates to prisons for safekeeping, or to coordinate with case managers or local bail funds
to enable release. For women who are kept on-site, knowing about their pregnancy earlier can enable the
jail to coordinate ultrasounds, initiate prenatal supplements, and take other measures to reduce risk for the
mother and fetus. Data from the routinely administered pregnancy tests should be saved electronically and
shared with the local health department for the purpose of accessing resources and planning services for
this population.

34
 Conclusion

The findings of this report support the need for State-supported data collection and reporting in
North Carolina jails. Many important local and state-level policy-makers and service providers would
benefit from data about the number, demographics, pre-trial status, and health conditions of jailed
individuals. Currently, jail data in North Carolina are difficult to obtain because of the large number of
jails and the lack of standardized and consistent data collection across jails. However, California, Texas,
and Colorado provide examples of how State coordination and accountability can create statewide jail
databases that contain useful information for a variety of stakeholders, and can make them accessible to
the public on interactive web-based platforms.
A state-run jail database cannot be created overnight, and requires resources and significant
coordination and planning. This report contains recommendations that North Carolina begin by
developing systems to collect and report the most essential jail information (daily counts, pretrial status,
demographics, and priority health variables) before expanding to a wider set of measures. By creating an
infrastructure for collecting the most essential data accurately and consistently, North Carolina will be
poised to answer essential questions about its jail populations and potentially to expand to a cutting-edge
data system similar to those in model states.
Rigorous data are needed to help understand the millions of Americans who pass through jails
each year, and states play a key role in supporting the collection of these data. North Carolina has
demonstrated the political will to collect jail data through its recent study bill. The findings of this report
underscore the timeliness and urgency of that legislation and provide additional guidance for this
important new State effort.

35
 References

Bureau of Justice Statistics. (2019). Local jail inmates and jail facilities. Retrieved from
https://www.bjs.gov/index.cfm?ty=tp&tid=12
California Board of State and Community Corrections. (2021). Jail Profile Survey. Available from:
https://www.bscc.ca.gov/s_fsojailprofilesurvey/
Colorado Division of Criminal Justice. (2021). H.B. 19-1297 Jail Data Dashboard. Available from:
https://cdpsdocs.state.co.us/ors/Data/Data_Instruments/HB1297/Dashboard/HB19-1297v3.html.
Copp, J. E., & Bales, W. D. (2018). Jails and local justice system reform. The Future of Children, 28(1),
103-124.
Duke University. (2020). Bass Connections. Retrieved from https://bassconnections.duke.edu/
Esquibel, E. (2021). A Systematic Review of Mental Illness, Criminogenic Risk, and the US Prison
System.
Florida Department of Law Enforcement. (2021). Florida County Detention Facilities Average Inmate
Population. Retrieved from Bureau of Research and Data Analysis:
http://www.dc.state.fl.us/pub/jails/2021/jails-2021-01.pdf
Florida Statutes. Criminal justice data transparency, 943.6871 Stat. (2018).
Gangi, E. (2021). Mental Health & Incarceration – Stepping Up Initiative. Retrieved from
https://ncimpact.sog.unc.edu/2020/04/mental-health-incarceration-stepping-up-initiative/
Henrichson, C. (2016). Measuring Incarceration: Don't overlook local jails. The Marshall Project.
House Bill 511, General Assembly of North Carolina, Pub. L. No. 2020-47 (2019).
Minton, T. D., Zeng, Z., & Maruschak, L. M. (2021). Impact of COVID-19 on the Local Jail Population,
January-June 2020: US Department of Justice, Office of Justice Programs, Bureau of Justice ….
National Academies of Sciences, E., & Medicine. (2021). Decarcerating correctional facilities during
COVID-19: advancing health, equity, and safety.
Pantazi, A. (2020). A new effort was supposed to make Florida's criminal justice data radically
transparent. So far, it's failed. Florida Times-Union. Retrieved from
https://www.jacksonville.com/story/news/local/2020/07/02/new-effort-was-supposed-to-make-
floridarsquos-criminal-justice-data-radically-transparent-so-far-itrsquos-failed/41716651/
Pantazi, A. (2021). Florida was going to be a beacon of criminal justice transparency. Instead, it’s almost
3 years behind schedule. The Tributary. Retrieved from https://jaxtrib.org/2021/08/24/florida-
was-going-to-be-a-beacon-of-criminal-justice-transparency-instead-its-almost-3-years-behind-
schedule/
Pollitt, S., & Woollard, L. (2019). Barriers to access and inadequate levels of care in North Carolina jails.
North Carolina medical journal, 80(6), 345-346.
Ranson, J. W. A., Bales, W. D., Blomberg, T. G., Casey, W., Copp, J., & Pesta, G. B. (2018). Evidence-
Based Policy Planning for the Leon County Detention Center: Population Trends and Forecasts.
Retrieved from Florida State University:
https://criminology.fsu.edu/sites/g/files/upcbnu3076/files/2021-03/Leon-Detention-Center-FSU-
Final-Report-1.pdf
Salmona, M., Lieber, E., & Kaczynski, D. (2019). Qualitative and mixed methods data analysis using
Dedoose: A practical approach for research across the social sciences: Sage Publications.
Sawyer, W., & Wagner, P. (2019). Mass incarceration: The whole pie 2019. Prison policy initiative, 19.
Texas Commission on Jail Standards (2021). Non-compliant jails. Retrieved from:
https://www.tcjs.state.tx.us/non-compliant-jails/
Wrenn, G., McGregor, B., & Munetz, M. (2018). The fierce urgency of now: Improving outcomes for
justice-involved people with serious mental illness and substance misuse. In: Am Psychiatric
Assoc.

36