Skip navigation
The Habeas Citebook Ineffective Counsel - Header

The UCLA Law Covid Behind Bars Data Project-Doing Social Injustice Work From Inside a Law School, 2023

Download original document:
Brief thumbnail
This text is machine-read, and may contain errors. Check the original document to verify accuracy.
Doing Social Justice Work from
Inside a Law School
Sharon Dolovich1

Table of Contents

The UCLA Law COVID Behind Bars Data Project:
A Brief History................................................................................ 216
II. Why a Law School?.........................................................................229
III. Tracking a Pandemic Behind Bars: What We Got Right
and Lessons Learned..................................................................... 234
IV. The Gradual Fading of COVID Prison Data and the Pivot
to All-Cause Mortality................................................................. 241
Appendix..................................................................................................... 248
In January 2020, reports of a new, potentially lethal virus began
emerging from Wuhan, China.2 By March of that year, a global pandemic
had commenced, and millions of people in the United States and around
the world became overnight experts in the pathways of COVID transmission. Soon, every facet of American society was struggling to adjust


Professor of Law, UCLA School of Law; Faculty Director, UCLA Prison Law
and Policy Program; Director, UCLA Law Behind Bars Data Project (formerly
the UCLA Law COVID Behind Bars Data Project). Thanks to Scott Cummings,
Liz DeWolf, Grace DiLaura, Michael Everett, Keegan Hawkins, Aaron Littman,
Neal Marquez, Hunter Nagai, Sasha Natapoff and Kalind Parish for helpful
comments and for bringing clarity to my sometimes hazy memory; to Bennett
Stein, who first suggested that I write the Data Project’s origin story; to Lucy van
Oldenbarneveld, for early help clarifying my thoughts; and to Cecilia Bain, Emma
Maynard, and Jack Stephens for their research assistance. And especially deep
thanks to the staff members and more than 300 volunteers whose creativity, talent,
and awe-inspiring commitment during the height of the pandemic helped the
UCLA Law COVID Behind Bars Data Project fulfill its mission of collecting and
publicly disseminating as much data as possible concerning the impact of COVID
in American prisons, jails, and detention centers.
See Sui-Lee Wee & Vivian Wang, China Grapples With Mystery Pneumonia-Like
Illness, N.Y. Times (Jan. 6, 2020),
asia/china-SARS-pneumonialike.html?searchResultPosition=1 [https://perma.

© 2023 Sharon Dolovich. All rights reserved.





to the new reality, with those living in long-term congregate settings—
prisons and jails chief among them—understood to be at greatest risk of
COVID infection and death.
Responding to this outsized danger, legal advocates for the incarcerated immediately began drafting letters and motions to officials and courts
demanding action on behalf of their clients. Meanwhile, also understanding the heightened risk prisoners3 faced, countless journalists around the
country turned to investigating the situation in carceral facilities. In their
quest for information, reporters started calling the prison lawyers, who
were already overloaded trying to respond to the demands of the moment.
Out of this high-pressure situation, the UCLA Law COVID Behind
Bars Data Project (“the Data Project”/“the Project”) was born. The
effort began as a two-tab, crowd-sourced spreadsheet allowing advocates
for the incarcerated nationwide to share their work and save others the
trouble of reinventing the wheel.4 Almost overnight, that spreadsheet
became the go-to national clearinghouse for all available data on COVID
in detention. And things only escalated from there. By mid-2020, the
United States Centers for Disease Control (CDC) was populating its
prison COVID tracker with the national facility-level data the Project
collected each day. The Project was also collecting, cleaning, analyzing,
and making publicly available the COVID data out of U.S. Immigration and Customs Enforcement (ICE) detention centers,5 among several
other data points.6 In July 2020, in concert with partners at the Johns




In this Essay, I will at times refer to people incarcerated in prisons, jails, and detention
centers as “prisoners,” a term that squarely acknowledges the “extraordinary and
dehumanizing exercise of state power known as imprisonment,” Justin Driver
& Emma Kaufman, The Incoherence of Prison Law, 135 Harv. L. Rev. 515, 525
(2021), and foregrounds the experience of being held against one’s will with no
power to shape one’s own conditions of life. See also Paul Wright, Language
Matters: Why We Use the Words We Do, Prison Legal News (Nov. 1, 2021), https:// (“[When people are incarcerated, they] are forced into cages at gunpoint
and kept there upon pain of death should they try to leave. What are they if not
prisoners? They did not somehow magically appear there, and they stay there
based on violence and fear of violence.”).
Although the spreadsheet was conceived as a crowd-sourced effort, it did not
take long before the risks of that format outweighed the benefits. As early as
April 8, 2020, we began to discuss restricting editing privileges to members of
our team, and by April 17, the spreadsheet had been locked. E-mail from Hunter
Nagai, UCLAW ‘23, to Sharon Dolovich, Professor of L., UCLA Sch. of L. (Apr.
8, 2020, 13:00 PST) (on file with author); E-mail from Hunter Nagai, UCLAW
‘23, to Sharon Dolovich, Professor of L., UCLA Sch. of L. (Apr. 16, 2020, 11:55
PST) (on file with author).
As with our COVID prison and jail data more broadly, our collection of COVID
data from ICE detention began manually. Eventually, we built scrapers and
collected the data from ICE just as we did with the data from the 50 state
Departments of Corrections (DOCs), the Federal Bureau of Prisons, the D.C.
Department of Corrections, and several large jail systems. For more on this data
collection process, see infra Part I.
See infra Part I.

COVID Behind Bars Data Project


Hopkins Bloomberg School of Public Health, we published what became
the definitive findings of disproportionately high COVID infection and
death rates for people in American prisons.7 By March 2021, the Data
Project team had grown to more than 150 people, including 8 full-time
staff, 2 part-time staff, 11 volunteer team leads, 20 students working with
the project for credit, and over 100 active part-time volunteers. By September 2021, our work had been cited in more than 200 media stories,
along with countless petitions for release, sentencing memoranda, preliminary injunctions, and expert reports. We had published multiple op-eds,8
our data had framed joint testimony to the Senate Judiciary Committee
investigating the Federal Bureau of Prisons’ (BOP) mishandled COVID
response,9 and our data visualizations had been seen by nearly 50,000
Twitter users. In sum, over the first 18 months of the pandemic, we grew
from a two-tab Google Sheet to an effective and respected national
authority on the impact of COVID in American carceral facilities.10 Our
immediate mission was simple: to support the efforts of advocates, journalists, activists, and others to compel policymakers and the American
public to recognize and respond to the outsized danger COVID posed to
the roughly 2 million people then locked inside prisons, jails, and detention centers.11





See Brendan Saloner, et al., COVID-19 Cases and Deaths in Federal and State
Prisons and Jails, JAMA Network (July 8, 2020),
journals/jama/fullarticle/2768249 [] [hereinafter
COVID-19 Cases and Deaths]; see also Julie Ward et al., COVID-19 Cases Among
Employees of U.S. Federal and State Prisons, 60 American J. of Preventive Med.
840 (2021),–3797(21)00118–5/fulltext
[]; see also Neal Marquez et al., COVID-19
Incidence and Mortality in Federal and State Prisons, JAMA Network (Oct.
6, 2021), [https://] [hereinafter COVID-19 Incidence and Mortality]; see
also infra note 67.
The large number of op-eds produced by the Project and published nationally in
2021–22 was a tribute to the hard work and talent of Research and Policy Fellow
Amanda Klonsky, who spearheaded our work on this front.
See Joshua Manson, Data Project Submits Joint Testimony to Congress on
the Bureau of Prisons’ Mismanagement of the Pandemic in Federal Prisons,
UCLA Law COVID Behind Bars Data Project (April 15, 2021), https:// [].
In doing this work, we benefited greatly from ongoing conversations with the
other organizations who were also tracking, analyzing, and disseminating a range
of data concerning the impact of COVID on those incarcerated in American
prisons, jails, and detention centers during the pandemic. These valued partners
included the Marshall Project,
a-state-by-state-look-at-coronavirus-in-prisons [];
the COVID Prison Project, [
B46N-GYSX]; the Prison Policy Initiative,
index.html []; and the Vera Institute, https://www. [].
For more on the Project’s mission, see infra Part III.




This Essay tells the story of how a law professor, a clinical teaching
fellow, and a large group of students, researchers, and volunteers created
a social justice organization driven by legal scholarship, data, and crisis. Law schools do not tend to prioritize cutting-edge engagement with
current events, but the pandemic was a catalyst. It pushed our team to
leverage our knowledge of the carceral state to respond to an emergent
humanitarian crisis. We already knew a lot about American prisons; our
collective knowledge was key to our ability to identify and collect new
data, to work with (or around) familiar institutions, and to advance the
interests of incarcerated people—most especially by doing all we could
to support the push for decarceration as the single most effective COVID
response.12 For me personally, this effort was the practical manifestation of my own scholarly focus during the two decades preceding the
pandemic. Over my career, much of my work has addressed one basic
question: given the fundamental moral commitments of a liberal democratic society, what does the state owe the people we incarcerate? There
is a straight line between that theoretical inquiry, born of the ecumenical, interdisciplinary character of American legal scholarship,13 and the
emergence of a data-driven, public health-focused effort to compel all
branches of government to confront and mitigate the COVID crisis
in detention.
The Data Project experience, conveyed here in narrative form,
offers several generalizable lessons about institution building in the public interest and the unique value of doing such work in the law school
environment. It also offers an object lesson in efficiency and resource
management. For those of us in the thick of the enterprise, there was
barely time to think. Every day brought a new round of missives from
journalists, lawyers, activists, academics, governmental actors, and people
with loved ones inside, each seeking or offering help of all kinds. Just
maintaining our core repository of facility-level COVID data—scraping, cleaning, analyzing, posting—took an enormous amount of time and
energy. In addition, we were constantly filing public records requests,
crafting blog posts, posting our key findings on social media,14 writing
reports and articles for peer-review publications, launching new initiatives, and working with numerous partners to try to reduce the COVID
risk to those in custody. Apart from our eight full-time paid staff,


See Nat’l Acads. of Scis., Eng’g, & Med., Decarcerating Corrections
Facilities During COVID-19: Advancing Health Equity, and Safety (Emily
A. Wang, Bruce Western, Emily P. Backes, & Julie Schuck eds., 2020). I served as
a member of the National Academies panel that produced this report.
See infra Part II.
Our main social media outlets were Twitter (@uclaprisondata) and Instagram
(@uclaprisondata). The look of our data visualizations and other social media
output was crafted by our web designer hyperobjekt, in conjunction with Liz
DeWolf, our Senior Project Manager. Together, Liz and Josh Manson, our
Communications Manager, took the lead on designing the posts, with staff from
all corners of the Project contributing content.

COVID Behind Bars Data Project


everyone on the team—from the Director (me) and the Deputy Director
(Aaron Littman) to the team leads and countless other volunteers—was
doing this work as a sideline to other full-time pursuits, whether work or
school. Although we were by no means on the pandemic’s front lines, we
were nonetheless stretched to the limit.
Despite the constant pressure, and despite being fully remote, our
team collectively produced an enormous amount of data, research, and
analysis, all directed to conveying in real time the scale of the crisis of
COVID in custody. In a quest to make our findings as accessible as possible to a broad range of end users, we posted virtually all our output
on our website15 or on GitHub.16 In addition, by the end of 2023, every
aspect of our work—from the core quantitative data collection to our
many publications to the output of the several initiatives undertaken
during the almost three years we focused on COVID—will be fully and
permanently accessible on our project’s Dataverse and also in the ICPSR
database.17 Those interested in all we learned tracking the pandemic
behind bars will thus have many ways to access our work.18 The story
offered here concerns not so much the substance of that work or our data
findings and analysis,19 but rather the basic and somewhat more practical
question of how a team based in a law school during a pandemic-induced
lockdown managed to accomplish all we did in such a short time in a policy realm not known for transparency.
Part I of this Essay tells the origin story of the UCLA Law COVID
Behind Bars Data Project. Part II addresses the question of how an
effort like this, focused on data and policy, could have arisen in a law
school, and what our experience reveals about the role the legal academy and legal scholarship can play in the movement for social justice
and policy change. Part III highlights some of the organizational factors
that enabled us to do what we did despite significant time and resource
constraints. The focus here is on the process of institution-building and
lessons learned. Finally, Part IV briefly describes the denouement of our


See UCLA Law COVID Behind Bars Data Project, https://uclacovidbehindbars.
org [] (last visited Apr. 25, 2023).
GitHub is a platform for hosting open-source code and data. See UCLAlaw
covid19behindbars/data, GitHub,
data [–4HYT] (last visited Apr. 25, 2023).
See Inter-Univ. Consortium for Pol. & Soc. Rsch., https://www.icpsr.umich.
edu/web/pages [] (last visited Apr. 25, 2023).
Our output features a mix of quantitative and qualitative data. It should be of
interest to anyone seeking to understand the effects of COVID on the nation’s
incarcerated population and on their families and communities. For more on
this aspect of the Project’s mission, see infra Part III.
For substantive analysis of what our data showed about COVID behind bars
over the first months of the pandemic, see Sharon Dolovich, Mass Incarceration,
Meet COVID-19, U. Chi. L. Rev. Online *4 (Nov. 16, 2020), https://ssrn.
com/abstract=3766415 [] [hereinafter Mass




COVID data collection efforts and our decision to pivot to our current
focus on national, all-cause carceral mortality.


The UCLA Law COVID Behind Bars Data Project: A Brief
From the moment the world learned of the novel coronavirus, it
was obvious to anyone paying attention that prisoners faced an outsized risk of COVID infection and death.20 Almost overnight, “social
distancing” entered the collective vocabulary, and public health officials
in the United States and worldwide were urging people to avoid coming within 6 feet of people not in their own households. But prisons are
crowded places, where people live in close proximity to one another
and social distancing is impossible.21 The poor ventilation found in
many carceral institutions meant that incarcerated people were constantly breathing recycled air.22 And while Americans in general were
cautioned to avoid gathering indoors and exhorted to get outside, people in custody were being locked down in their cells and dorms and
refused access to the yards.
This new reality dawned quickly. In response, advocates mobilized.
Much of the effort in the first weeks was directed at putting pressure
on decision-makers of various kinds—sheriffs, judges, district attorneys,
prison officials, etc.—to reduce the COVID risk, whether by reducing the
population density of the facilities within their sphere of authority23 or
by taking other steps in mitigation.24 To speak authoritatively about the
urgent need for action, lawyers for the incarcerated were being forced in
real time to figure out what was then known about this new lethal virus
in order to convey the grave risk it posed to the people they represented.
These efforts were facilitated by a simple yet surprisingly effective
channel of communication that has long linked the national prisoners’





Long before COVID, it was well understood that incarceration constitutes a
significant risk factor for exposure to communicable diseases. See, e.g., Paul L.
Simpson et al., Prison Cell Spatial Density and Infectious and Communicable
Diseases: A Systematic Review, BMJ Open (2019).
See Mass Incarceration, supra note 19, at *8-*11 (explaining the many reasons
why people in custody faced a disproportionate risk of COVID infection
and death); see also Saloner et al., COVID-19 Cases and Deaths, supra note 7
(reporting that, over the first 4 months of the pandemic, people in federal and
state prisons were 5.5 times more likely to be infected and 3 times more likely to
die from COVID than those in American society more generally).
Early on, there were also concerns about incarcerated people in crowded housing
units being forced to share showers, toilets, and sinks, which at the time were
believed to be prime vectors for COVID transmission. See Mass Incarceration,
supra note 19, at *8-*9.
See Sharon Dolovich, Every Public Official with the Power to Decarcerate Must
Exercise That Power Now, The Appeal (Apr. 10, 2020),
every-public-official-with-the-power-to-decarcerate-must-exercise-that-powernow [–957G].
For discussion of some of the steps state officials took to this end early in the
pandemic, see Mass Incarceration, supra note 19, at *11-*18.

COVID Behind Bars Data Project


rights bar: a listserv, started in the early days of email communication and
open to any legal advocates working on behalf of the incarcerated. The
list is always active, but in March 2020, it positively exploded as people
hastened to strategize the best ways to help their clients. From the start
of the pandemic, it served as a virtual hub for information about what
was happening in prisons and jails nationwide. Although mostly for lawyers, the listserv also includes a few academics. As one such academic, I
was able to watch the drama unfold moment by moment.
Among the first steps corrections officials took to try to keep the
virus from taking hold inside was to cancel visits into their facilities. The
move seemed epidemiologically appropriate given what we already knew
about COVID transmission. But it was also sure to cause a lot of pain
and suffering to people in custody, for whom visits from loved ones are
a lifeline. As reports of visit cancellations flooded into the listserv from
across the country, I felt the need to keep track. On March 13, I asked
my research assistant at the time, Keegan Hawkins, if he could create
a table noting which states had canceled visits, whether the new policy
also applied to attorney visits, and the date of the cancellation.25 Keegan
agreed and made a spreadsheet.
Although we did not realize it then, this step set the stage for all
that was to follow. The very next day, March 14, an email came into the
listserv from Corene Kendrick, then a staff attorney at the Prison Law
Office. Corene was responding to an email sent to the list earlier that day
from Sarah Grady, then a civil rights attorney at Loevy & Loevy, asking
whether anyone on the list had thus far “seen any jails and prisons taking
steps to decarcerate in light of the COVID crisis.”26 Corene answered
Sarah’s question, describing several formal requests advocates had been
making to that end and attaching PDFs of the relevant documents. Then,
in closing, Corene made a suggestion in the form of a question, asking
if there was “any chance anybody (perhaps some of our law professor
comrades) could set up a web-based clearinghouse that tracks all of
these requests?”27 Reading this, it occurred to me that the spreadsheet
Keegan had just created might be repurposed to the end Corene had
in mind—allowing advocates to post their various demand letters, court
filings and other public-facing documents seeking a decarceral response.
Corene liked the idea of a crowd-sourced database, so Keegan added
a tab labeled “Population Reduction Requests” and I shared the link
with the list.28

E-mail from Sharon Dolovich, Professor of L., UCLA Sch. of L., to Keegan
Hawkins, UCLAW ’21 (Mar. 13, 2020, 10:34 PST) (on file with author).
E-mail from Sarah Grady, C.R. Att’y, Loevy & Loevy, to Prisoners’ Rights
Listerv (Mar. 14, 2020, 14:26 PST) (on file with author).
E-mail from Corene Kendrick, Staff Att’y, Prison L. Off., to Prisoners’ Rights
Listserv (Mar. 14, 2020, 18:16 PST) (on file with author).
We also tweeted the link via the UCLA Prison Law and Policy Program Twitter
account. UCLA Prison Law (@uclaprisonlaw), Twitter (Mar. 21, 2020 8:58 PM), [https://perma.




After that, everything seemed to happen at once. People on the
listserv started posting their work product and sharing the link. Activists,
grassroots organizations, public defenders, and others began adding their
own documents. Litigators around the country, who had been drafting
petitions to courts on behalf of incarcerated clients, also started posting and sharing the link. As the spreadsheet made the rounds, people
from across the country began noticing gaps in the types of data it contained. Things were moving quickly and there were already many other
data points that would likely be helpful to all those constituencies—not
just lawyers and activists but also reporters, policymakers, academics, and
those with loved ones inside—who were trying to figure out what was
happening with COVID in custody. I began to receive emails from people I did not know, containing suggestions for additions and offers of
help. For example, in answer to calls from legal advocates and grassroots activists (many of whose written demands were already posted on
the Population Reduction Requests tab on our Google Sheet), public
officials around the country—including prosecutors, sheriffs and police
chiefs, trial court judges, and state Supreme Courts—had begun taking
steps to decarcerate in a bid to reduce the population density in their
facilities.29 Noting this development, Maddy DeLone, longtime director of the National Innocence Project, emailed to ask if we would like
someone to track these releases. I put her in touch with Keegan and
together they created a new Jails and Prisons Releases tab, complete with
coding categories. Maddy set to work assembling a team of volunteers,
which began populating the new tab with all the applicable reports, press
releases, and news stories they could find. Around the same time, Yasmine Tager, a juvenile justice lawyer with Berkeley Law School’s East
Bay Community Law Center, also reached out, in her case with an offer
to begin tracking the effects of COVID in juvenile detention. She and
Keegan designed a youth facilities tab, and Yasmine and the team of volunteers she mustered also got to work.30
In those early weeks, much of the pressure on corrections officials
to mitigate the COVID risk was coming from local activists and from
family and friends of incarcerated individuals, who were refusing to
remain silent about the grave threat facing their loved ones inside. Jordan Palmer, then UCLA Law’s Jane Kahn Prison Law Fellow, recognized
the remarkable work being done around the country by grassroots organizers and others engaged in ad hoc acts of protest and resistance. She


See Mass Incarceration, supra note 19, at *15-*18.
For each tab we added to the original Google Sheet, we crafted the coding
categories with an eye to likely end-users. The expectation was that different
categories of data would be useful to different constituencies, and we tried to
make each category of data we were collecting as accessible and user-friendly as
possible. This approach reflected the Project mission to support all efforts being
taken to address the crisis of COVID behind bars, from whatever quarter.

COVID Behind Bars Data Project


worked with Keegan to create two additional tabs: Grassroots and other
Organizing Efforts.31 and Fundraisers and Mutual Aid.32
As all of this was taking place, something was happening on the corrections side that was close to revolutionary for an institutional context
long defined by a rigid culture of secrecy33: Departments of Corrections
(DOCs) nationwide began putting up dashboards on their websites
reporting a range of COVID metrics for both residents and staff, including tests administered, infections, and deaths.34 The first of these DOC
dashboards came online in mid-March 2020, in the days just after the
launch of our spreadsheet. If we were to be a one-stop shop for all
data on COVID in custody, this was information we needed to capture.
In keeping with the pace at which things were moving, I barely had a
moment to give the matter some thought before I received an email from
one of my former students, Grace DiLaura. Grace had seen a post on the
UCLA Law website about what we were doing and reached out to see if
I needed any help. I responded that I did indeed need help and asked if
she could figure out a way to get the data from the DOC COVID dashboards onto our spreadsheet. Grace was game and worked with Keegan
to get yet another tab up and running—COVID-19 Prison and Jail Cases
and Deaths. At the time, I had no idea what I was asking of Grace, and
she (thankfully) had no idea what she was signing up for. Starting on
March 26, 2020,35 Grace began manually transferring from these dashboards to our spreadsheet all the COVID data then being posted by fifty
state DOCs, the BOP, and several large jail systems.36




I am especially proud of this database, the only one of its kind that emerged
during the pandemic.
This latter tab emerged after Jordan began tracking protests and noticed that
several campaigns had begun to solicit donations to help incarcerated people
protect themselves during COVID. She figured that if people were visiting our
spreadsheet, they must care about the health and safety of people in custody and
so might be inclined to contribute.
See generally Sharon Dolovich, The Failed Regulation and Oversight of
American Prisons, 5 Ann. Rev. Criminology 153 (2022) [hereinafter The Failed
Regulation]; Mass Incarceration, supra note 19; see also Andrea C. Armstrong,
Access Denied: Public Records and Incarcerated People, 19 U. St. Thomas L.J. 220
In this respect, corrections agencies joined many other public institutions
nationwide, which also began posting dashboards with data concerning
the impact of COVID in their spheres of authority. For discussion of this
phenomenon, see Damir Ivanković, et al., Features Constituting Actionable
COVID-19 Dashboards: Descriptive Assessment and Expert Appraisal of 158
Public Web-Based COVID-19 Dashboards, 23 J. Med. Internet Rsch. 2 (2021).
DOC COVID data first appeared in our dataset on March 26, 2020.
Readers with data science backgrounds will wonder why we didn’t just build
scrapers, which would have allowed us to achieve the same result with much less
effort. The short answer is that, at the time, I had zero data science knowledge
and had never even heard of web scraping. Eventually, we hired data scientists
and built and ran scrapers to do this work. But ironically, my early ignorance
on this score turned out to be one of the secrets of our success: as other savvier
players were taking the time to build their scrapers, we were already collecting




This process proved exceedingly challenging and time-consuming. Fortunately, just as Grace was realizing that the job was too much
for one person alone, I received an email from Kalind Parish, who was
then on leave from a political science PhD program at the University of
Pennsylvania. Kalind was interested in the intersection between incarceration and health. He had heard of our work from Judith Resnik of
Yale Law School and reached out to see if we needed any volunteers.
Almost immediately, Kalind joined forces with Grace. The pair devised a
system for sharing the work of transferring the data from the DOC dashboards, dividing the states in a way that balanced those DOCs whose data
was especially challenging to access. Every day, Grace did one set and
Kalind did the other, and the next day they swapped. They continued
this time-consuming and tedious process seven days a week for almost
three months, until we hired data scientists and built our first web scrapers, finally allowing us to automate our data extraction and bring our
data-gathering process into the twenty-first century.
During the first weeks of the pandemic, prisoners’ advocates
flooded the courts, both federal and state, with petitions seeking releases
and other forms of COVID mitigation. Initially, we began posting these
filings on our “Population Reduction Requests” tab. But this tab quickly
became a grab bag of many different kinds of documents. In a quest to
streamline and make the contents of the spreadsheet more accessible,
we broke out two more tabs. First, we established a tab called “Population Reduction Responses.” Unfortunately, this tab also rapidly became
diversified, making it less useful than we had hoped.37 We also created a
separate court filings tab, to which we added all the judicial opinions we
could find bearing on COVID behind bars. This move proved more successful. The judicial decisions would soon start pouring in, and we now
had a place for them to live. Rebecca Fordon, then a reference librarian
at UCLA Law School’s Hugh and Hazel Darling Law Library, stepped up
to manage the tab and wound up serving as the lead of our court filings
team until mid-2022 when we finally closed the dataset.38



and disseminating facility level COVID data, which meant that people looking
for quantitative indicators of the impact of COVID in custody were already
coming to us. We therefore became the go-to site while others were still getting
up and running.
We eventually abandoned both Population Reduction Requests and Population
Reduction Responses. The contents of these tabs were among the datasets
we came to refer to as orphans and which we published collectively in July
2022. See Sharon Dolovich, Notes on Incomplete Datasets from the Project’s
Early Days, UCLA Law COVID Behind Bars Data Project (Aug. 4, 2022),’s-early-days.
Shortly after we broke out the Court Filings tab, we were approached by a team
of lawyers from Columbia Law School, the Bronx Defenders and,
who invited us to join forces with them to create a litigation hub for prisoners’
advocates containing as many cases bearing on COVID in custody as could be
coded and posted. Rebecca Fordon became our point person on this effort and

COVID Behind Bars Data Project


By mid-April of 2020, just a few weeks after we began, the spreadsheet had grown, as had the whole enterprise. Maddy DeLone was
running our releases team. Yasmine Tager was managing the data on
youth facilities. Rebecca Fordon was handling court filings. Jordan
Palmer had built a team of volunteers who were busy monitoring Twitter
and Facebook in order to populate our “Grassroots and Other Organizing” tab and our fundraisers and mutual aid tracker. Meanwhile, Grace
and Kalind had become our inaugural data team. In addition to effectuating the daily transfer of facility-level data to our COVID-19 Prisons
and Jails Cases and Deaths tab, Grace and Kalind were fielding daily
inquiries from journalists and others who had accessed our spreadsheet
and had questions about the data. Still, a month into the work, one
important gap had not yet been filled: the COVID data for ICE facilities. As a practical matter, immigration detention centers look, feel, and
function pretty much like prisons and jails. They are crowded, highly regulated, and those held inside have no control over their surroundings or
movement. Recognizing this gap in our data, I made a point of mentioning our need for volunteers to take on the immigration detention piece
of our COVID data collection efforts anytime I spoke publicly about our
work. Fortunately, two extremely motivated individuals answered my
call: Theresa Cheng, a JD-MD who was then an ER resident, and Joanne
Choi, then a student at Columbia Law School. I put the pair in touch, and
they began their work, strategizing how best to use the platform to track
the impact of COVID in ICE detention. Together they built our largest
and most ambitious volunteer-led team.39 Using co-team leads worked
so well that it became a model we would replicate many times during the
ensuing two years.
As the Project grew, I was often asked what had originally made me
decide to launch the enterprise. What I am trying to convey here is that,
at least in one sense, there was no initial idea: I did not consciously decide
to start a multi-million-dollar data project, which instead seemed to grow
rapidly and organically of its own accord. Yet the Project was also a natural outgrowth of the work I had been doing for two decades. I have


helped this consortium of partners to bring the Litigation Hub into being. She
was assisted on our end by Dylan Lee, UCLAW ‘22, who supervised the coding
on a part-time basis for the entire life of that initiative. The court filings database
now lives on our website. See Litigation Database, UCLA Law COVID Behind
Bars Data Project, [https://perma.
cc/28L2-TMYD] (last visited Apr. 25, 2023).
Among other things, under the leadership of Theresa Cheng and Joanne
Choi, and later of Theresa, Ishan Nagpal and Ben Woolley, the immigration
team tracked the COVID data inside ICE detention, tracked releases from
ICE facilities, collected court filings related to ICE detention and COVID,
followed and analyzed the impact of Fraihat before the Ninth Circuit reversed
the decision in Fraihat v. U.S. Immigr. & Customs Enf’t, 16 F.4th 613 (9th Cir.
2021), filed many FOIA requests, tracked the effect of the Omicron variant on
immigrant detainees in ICE facilities, and collaborated with multiple grassroots
organizations around initiatives concerning COVID in immigration detention.




written extensively about the dehumanization that occurs in American
prisons, especially when it comes to the health and well-being of people
who live behind bars.40 I have studied the opacity and intransigence of
our carceral institutions and the failures of regulation that shape the reality on the ground.41 I know how unresponsive all branches of American
government can be when confronted with the gross abuse and neglect of
people in custody.42 In other words, when the pandemic hit, I had already
been thinking for years about the legal, political, and humanitarian failures of the American carceral state and the basic failures of decency that
define daily life inside. I launched what became the UCLA Law COVID
Behind Bars Data Project because, in that historical moment, I felt compelled to do so and could not have imagined doing anything else.
This normative background still leaves open the question of how,
as a practical matter, the Data Project came together as an effective,
functioning organization. Later, when the emergency abated somewhat
and there was time to reflect, I came to recognize several factors that
made possible the emergence and ultimate shape of the Project, factors I
explore later in this Essay.43 But at the time, at least in terms of the actual
day-to-day, it felt as if one minute I was sharing a spreadsheet link with
the prisoners’ rights listserv, and the next minute I was collaborating with
an enormous team of people in a shared mission to gather as much data
as possible about COVID behind bars and make it available to whoever
wanted to access it.44



See, e.g., Sharon Dolovich, Prison Conditions, in 4 Reforming Criminal
Justice: Punishment, Incarceration, and Release 261 (Erik Luna ed., 2017)
[hereinafter Prison Conditions]; Sharon Dolovich, Two Models of the Prison:
Accidental Humanity and Hypermasculinity in the L.A. County Jail, 102 J. Crim.
L. & Criminology 965 (2012) [hereinafter Two Models]; Sharon Dolovich,
Strategic Segregation in the Modern Prison, 48 Am. Crim. L. Rev. 1 (2011)
[hereinafter Strategic Segregation]; Sharon Dolovich, Creating the Permanent
Prisoner, in Life Without Parole: America’s New Death Penalty? (Charles
Ogletree & Austin Sarat eds., 2012); Sharon Dolovich, Foreword: Incarceration
American-Style, 3 Harv. L. & Pol’y Rev. 237 (2009) [hereinafter Incarceration
See, e.g., Dolovich, The Failed Regulation, supra note 33.
See, e.g., id.; Sharon Dolovich, The Coherence of Prison Law, 135 Harv. L. Rev.
F. 302 (2022) [hereinafter Coherence of Prison Law]; Sharon Dolovich, Forms
of Deference in Prison Law, 24 Fed. Sent’g Rep. 245 (2012); Sharon Dolovich,
Canons of Evasion in Criminal Constitutional Law, in The New Criminal
Justice Thinking 111 (Sharon Dolovich & Alexandra Natapoff eds., 2017)
[hereinafter Canons of Evasion].
See infra Part III.
These initial efforts benefited hugely from the help and support of many staff
members at the UCLA School of Law. Ben Nyblade, then the director of our
Empirical Research Group, made himself available as an informal advisor from
the start. UCLA IT staff and website managers stepped in to help us manage
the spreadsheet and get our growing number of volunteer team leads into the
system. Here is just one example of the considerable contribution members of
the UCLAW staff made to our efforts during the initial days of the project: On
March 29, 2020, Judith Resnik reached out. She had been hearing from federal

COVID Behind Bars Data Project


It did not take long before we saw the need for someone with data
science expertise on the team. As was emblematic of that time, this idea
had barely formed when I received an email, dated March 25, forwarded
to me by my UCLA colleague Dave Marcus. Dave had a college friend
named David Menschel who now runs the Vital Projects Fund (VPF), a
foundation supporting criminal justice reform efforts across the United
States. David had seen our work and wanted to know if we needed any
financial help.45 The generous initial support we received from David
and VPF ultimately made it possible for us to grow into what we eventually became.46 Why had the email come to Dave Marcus and not to me?
When we initially created the link and I shared it with the listserv, I had
thought of it solely as a tool to help the advocacy community. Although
subsequent events made clear that we were reaching a much broader
constituency, I was too busy managing the effort while continuing my
day job of full-time law teaching to give much thought to what we were
building. As a result, it had never occurred to me to put either my name
or the name of UCLA School of Law on the spreadsheet itself. Because
Keegan—a UCLA law student—had created the Google Sheet, his email
was attached to it, and VPF used that clue to track us down.
Around this time, I had begun talking about our growing data-collection efforts with my colleague Aaron Littman. In the fall of 2019, after
several years litigating prison conditions in Georgia and Alabama with
the Southern Center for Human Rights, Aaron had arrived at UCLA as
a Binder Clinical Teaching Fellow and launched our first in-house prisoners’ rights clinic.47 Aaron has a deep well of subject-matter expertise



judges who were feeling at sea and uncertain of how to approach the many
petitions they were receiving from incarcerated claimants asking for relief of all
kinds in light of COVID. We had our case filings tab by then, but some judges
were finding it challenging to navigate. Could we set up a quick link for judges,
which would allow them easy access to the cases they sought? We immediately
mobilized, bringing together Ben Nyblade, Rebecca Fordon from the law library,
Dave Cappoli, who runs the law school’s website, and Scott MacKnight, who
heads up UCLA’s IT department. In less than a week, we had a shortcut on our
UCLA Law webpage that took judges to a page that allowed them to access
cases organized by state/federal, individual claims/class actions, petitions for
release, and petitions for mitigation. The speed with which this link was created
reflected the degree to which people at the law school, recognizing the urgency
of the moment, were prepared to step up and help make things happen.
E-mail from Marlena Williams, Outreach Dir., Vital Projects Fund, to David
Marcus, Professor of. L., UCLA Sch. of L. (Mar. 25, 2020, 21:37 PST) (on file
with author).
David Menschel is deeply committed to transforming the American criminal
legal system and achieving true criminal justice. He has supported innumerable
efforts to effect change in this space, and it is no exaggeration to say that, had
it not been for him and the Vital Projects Fund, we would never have lasted so
long or grown into what we eventually became.
Aaron is now Assistant Professor of Law at UCLA. In the very first weeks of
the pandemic, while I was launching what became the Data Project, Aaron had
undertaken his own data mobilization efforts. Recognizing the urgent need for
decarceration in the face of the pandemic, he compiled a 50-state survey of all




and also some knowledge of data science, which I lacked. He also, as
it happened, knew David Menschel and so joined our initial conversations about funding. We asked David for the money to hire two full-time
data scientists for an initial period of six months each. By the time we
had posted for these positions and were receiving what became a deluge
of applications (these were the early days of the COVID lockdown and
many people with data science skills appeared to be seeking meaningful
work), Aaron had become my partner in the project. Before long, we
had christened him Deputy Director and our growing effort had a name
that, thanks to my lack of marketing experience, was far too long and
unwieldy: the UCLA Law COVID-19 Behind Bars Data Project.48
With the seed funding provided by VPF, we hired Michael Everett
and Chase Hommeyer as our inaugural data science fellows. We had
intended to put the pair immediately to work building web scrapers to
free up Grace and Kalind, who were continuing to share the daily task
of transferring the COVID data from DOC websites to our facility-level
data page. This effort was delayed by a proposal that we merge our data
collection efforts with those of a team in Philadelphia, which had built
its own scrapers for this purpose. By the time the negotiations around
this merger fell apart, it was June 2020, and Grace and Kalind were completely tapped out.
So we mobilized. We put together an ad hoc web-scraper boot
camp team comprised of Chase, Kalind, Michael, and Aaron along with
two other talented coders who were serendipitously available to assist us
in this mission. The first was Cooper Mayne, a rising UCLA 2L, whom
we had hired as a summer research assistant to begin analyzing the abundance of case law decided in the first months of the pandemic. When
we learned that, before he came to law school, Cooper had worked as
a professional web scraper, we temporarily reassigned him to join the
boot camp web-scraper team. Cooper proved to be our ringer, developing at least 15 of the most challenging scrapers. Rounding out the group
was Isaac Dienstag, a talented high school senior with coding expertise
who had volunteered with us since the early days of the project. After a
week of intense work, we had our scrapers, and Kalind and Grace were
released almost completely49 from manual data transfer purgatory.50


existing statutory release powers enjoyed by various state actors with control
over decisions to incarcerate. This database also circulated widely in the first
weeks of the pandemic and became the first external resource we listed on the
Additional Resources tab in our initial spreadsheet. See Resources, UCLA Law
Spreadsheet of Statutory Release Powers in the Fifty States, UCLA Law COVID
Behind Bars Data Project, [https://].
Eventually, we dropped the “-19” and became the UCLA Law COVID Behind
Bars Data Project.
Some state DOCs posted their data in forms other than a scrapeable dashboard.
For these states, manual data transfer persisted.
At this point, Kalind joined Chase and Michael on our new data team, which
focused on running the scrapers, cleaning and posting the data, and fielding

COVID Behind Bars Data Project


The Vital Projects Fund continued to support our work, and we
received considerable in-kind support from the UCLA School of Law,
which, among other things, made it possible for us to onboard and administer so many new staff members so quickly. We also obtained funding
from Arnold Ventures which, along with facilitating new hires, allowed us
to begin constructing what became our interactive website.51 In addition,
in September 2020, we finalized a contract with the CDC, which had been
relying on our data to populate its COVID prison tracker. We had met
with the CDC COVID corrections response team multiple times in the
late spring and summer of 2020. When their team realized the limited
nature of our resources, they began the internal CDC process of seeking
financial support for us. With this funding in place, we were able to bring
on more full-time staff and our capacity grew, as did our output. By the
end of 2020, in addition to our core COVID data-gathering work, we had,
among other initiatives, launched our website,52 started our blog,53 begun
work on our data transparency scorecard,54 developed our own in-house
public records practice,55 and crafted a social media strategy.56 With our
growing team of data scientists—which by the end of 2020 had grown
to 4 full-time staff57—we found we had the bandwidth to provide data
science support to our five volunteer-led teams (releases, youth facilities,




the daily questions we were receiving from journalists and others. Grace
meanwhile turned to leading the team of volunteers coding the thousands of
COVID policies issued by corrections agencies since the start of the pandemic,
the capturing and saving of which Grace had presciently organized using a small
army of volunteers. That effort became our Prison Policy Index. See Grace
DiLaura et al., COVID Prison Policy Index: Version 1.0 [dataset]. Los Angeles,
CA, CPPI, 2022.
This initiative was spearheaded by Liz DeWolf. We had initially hired Liz as our
Stakeholder Relations Manager, but she quickly became our de facto Senior
Project Manager. Working with hyperobjekt, our web designer, to create our
own interactive website was just one of multiple complex projects Liz undertook
on behalf of the Project during her time with us.
See UCLA Law COVID Behind Bars Data Project, https://uclacovidbehind [].
See UCLA Law COVID Behind Bars Data Project: Blog, https://
For more on this initiative, see infra Part IV.
This effort was initiated and led by one of our first Data Science Fellows, Michael
Everett, early in his time with the Data Project. Over the years, it became central
to our work. Part of this effort was directed towards compiling as much data as
possible on all deaths in custody, whether from COVID or other causes. As
I explain in Part IV, this piece of the remarkable data repository Michael has
built through his dedicated public records work has come to assume national
importance. It also constitutes the foundation of our current focus on all-cause
carceral mortality.
See supra note 12.
By the end of 2020, the members of our data team were Michael Everett, Hope
Johnson, Neal Marquez, and Erika Tyagi. They were a truly remarkable group,
and it is hard to overstate the collective contribution they made to the success of
the Project.




immigration detention, court orders, and grassroots organizing). Each
“sub-team” (as we sometimes called them) was assigned a member of
the data team, who helped them effectively capture, manage, analyze, and
present their data. All these sub-teams also had as many part-time volunteers as they could manage. Without volunteers, we could have not
come close to doing all we did. With everything happening so fast, we did
not start keeping track of how many people were volunteering their time
on the Project until September 2020. At that time, in addition to our 11
team leads,58 there were 65 active volunteers working with us. By March
2021, that number had grown to exceed 100.59 Although the precise number of volunteers remained a moving target, our best guess is that, over
the life of the Project, the total was close to 325.60
As the pandemic evolved, so did our work. By early 2021, along
with the rest of the world, we expanded our focus to include the vaccine
roll-out.61 Among other things, we contributed to the vaccine education
effort directed at people in custody62 and collected signatories for a let58.



The leads of our sub-teams began as volunteers. Starting in the summer of
2020, we were able to provide each person in this position a monthly stipend as
recognition of their substantial contributions to the Project.
See Appendix (Project Org Chart, March 9, 2021).
Most of these volunteers were law students or undergraduates at UCLA.
Our volunteer corps also included people from all over the country—many
undergraduates and law students at other schools, along with, among others,
graduate students in data science and other STEM subjects, practicing lawyers,
high schoolers, and retirees. This number does not include the students at
Columbia Law School, who worked as volunteers coding the COVID cases for
the Health Is Justice Litigation Hub, on which we partnered with Columbia Law
School, Bronx Defenders, and See supra note 38. I say more about our
system for efficiently onboarding and deploying volunteers in Part III.
See, e.g., Joshua Manson, Who’s Getting the Vaccine Behind Bars?, UCLA
Law COVID Behind Bars Data Project: Blog (Feb. 18, 2021), https:// [];
Joshua Manson, Data Project Adds Vaccine Counts to Data Dashboard,
UCLA Law COVID Behind Bars Data Project: Blog (Mar. 9, 2021), https:// []; Erika Tyagi & Liz DeWolf, The Challenges of Interpreting Vaccination
Data Reported by Carceral Agencies, UCLA Law COVID Behind Bars Data
Project: Blog (June 23, 2021),
See Maya Chauduri, Sharon Dolovich & Aaron Littman, Urgent Need for
Vaccine Administration in Prisons, Jails, and Detention Centers, Prison Legal
News (Feb. 1, 2021), [https://perma.
cc/7VAK-Y4AL]; Maya Chaudhuri, Vaccine Guide in Prison Legal News,
UCLA Law COVID Behind Bars Data Project: Blog (Feb. 16, 2021), https:// [];
Marc F. Stern et al., Willingness to Receive a COVID-19 Vaccination Among
Incarcerated or Detained Persons in Correctional and Detention Facilities—Four
States, September–December 2020, 70 Morbidity & Mortality Wkly. Rep. 473,
473–77 (Apr. 2, 2021), [https://perma.
cc/8SFK-GNJH]. .

COVID Behind Bars Data Project


ter, ultimately signed by more than 480 experts in medicine and public
health, advocating vaccine priority for the incarcerated.63 We also tracked
the emergence of new variants,64 and when Delta and Omicron exploded
onto the scene, we put all our efforts into crafting a response.65 And as
the quality of the data reported by corrections agencies—never high to
begin with66—began to deteriorate, we did all we could to call attention
to that decline67 while pressing for improved data transparency across
prisons, jails, and detention centers.68






See Letter from UCLA Law COVID Behind Bars Data Project, to CDC’s
Advisory Committee on Immunization Practices, State Governors, and State and
Local COVID-19 Vaccination Program Planning and Coordination Committees,
COVID-19 Vaccination Recommendation (Dec. 17, 2020) (
edit) [].
See, e.g., Sharon Dolovich & Poornima Rajeshwar, SARS-COV-2 Variants Go
to Prison: What Now?, UCLA Law COVID Behind Bars Data Project: Blog
(Apr. 7, 2021),
See, e.g., Hope Johnson et al., As the Delta Variant Causes New Outbreaks in
Prisons, Now Is the Time for More Transparency, Not Less, UCLA Law COVID
Behind Bars Data Project: Blog (Jul. 30, 2021), https://uclacovidbehindbars.
org/delta-data-transparency [–4WZ8]; Amanda Klonsky,
Hope Johnson & Lauren Woyczynski, Failure to Decarcerate Jails Has Led
to Unnecessary Deaths and Widespread Infection During Omicron Surge,
UCLA Law COVID Behind Bars Data Project: Blog (Mar. 18, 2022), [];
Amanda Klonsky & Hope Johnson, As Omicron Surges in State and Federal
Prisons, Incarcerated People Remain Vulnerable, UCLA Law COVID Behind
Bars Data Project: Blog (Feb. 3, 2022),
omicron-surge [].
See Liz DeWolf, Poornima Rajeshwar & Erika Tyagi, Missing the Mark: Data
Reporting & Quality Scorecard, UCLA Law COVID Behind Bars Data
Project: Blog (Mar. 17, 2021),
[] [hereinafter Missing the Mark].
See Sharon Dolovich, Erika Tyagi & Neal Marquez, The States That Lead the
Nation in COVID Cases Are Hiding Their Prison Data, UCLA Law COVID
Behind Bars Data Project (Aug. 20, 2021),
delta-surges-hiding-data [] [hereinafter The States
That Lead the Nation]; Baji Tumendemberel & Bennett Stein, 21 States and
D.C. Have Stopped Reporting Active COVID Cases in Prisons, Despite Active
Community Outbreaks, UCLA Law COVID Behind Bars Data Project (Oct.
18, 2022),
[]; Lauren Woyczynski & Joshua Manson, What
We Won’t Know When the Next Surge Arrives, UCLA Law COVID Behind Bars
Data Project (May 9, 2022), [].
See Amanda Klonsky, Neal Marquez & Lauren Woyczynski, Jails Should Be
Prioritized for Surveillance of New COVID Variants, UCLA Law COVID
Behind Bars Data Project (Nov. 4, 2022),
jails-surveillance-covid []; UCLA Law COVID
Behind Bars Data Project Team, The Federal Government Doesn’t Know How
Many People Died in Prison Since the Pandemic Began, UCLA Law COVID
Behind Bars Data Project (Oct. 31, 2022),




Over the life of the Project, the data we collected helped to frame
the national conversation on COVID in custody. By the end of 2022,
it had been cited in more than 300 media stories and reports and was
receiving up to 5000 unique views per day on the CDC COVID data
tracker. As already noted, our data visualizations were seen by nearly
50,000 Twitter users. Public defenders relied on our data in countless
court filings. Activists used it as the foundation for legislative testimony
and open letters to governors. Academics, in collaboration with members of our team, used it to produce landmark research unmistakably
demonstrating the scale of the COVID crisis in American prisons.69 In
addition to this public facing output, many other facets of our work took
place behind the scenes: countless meetings with government agencies,
policymakers, and other nonprofits; collaborations with grassroots organizations; investigations into hidden COVID deaths in custody70; advocacy
around improved carceral data transparency; and our oral history project,
which created an archive of interviews with prison litigators about their
experience representing incarcerated people during the pandemic.71




PSI-committee-hearing []; Elif Yücel, Tracking
the Transparency of COVID Data in Juvenile Justice Facilities, UCLA Law
COVID Behind Bars Data Project (Oct. 3, 2022), https://uclacovidbehindbars.
org/youth-facility-transparency [].
See, e.g., COVID-19 Cases and Deaths, supra note 7; COVID-19 Incidence and
Mortality, supra note 7; Ward et al., supra note 7. This research was produced
in partnership with Brendan Saloner and Julie Ward of the Johns Hopkins
Bloomberg School of Public Health. This fruitful collaboration began in the
earliest days of the pandemic, when Brendan, having seen our spreadsheet,
reached out to ask about using our data in his own research. His overture and
subsequent work as the lead author of our first published study made it possible
for us to quantify the reality of COVID’s impact on the incarcerated, which in
turn helped give the issue considerable national traction.
This work was commenced in the first year of the Project by Victoria Rossi, our
investigator. See, e.g., Victoria Rossi, Prisons Mistreat Loved Ones’ Belongings
After Their Deaths, Some Families Say, UCLA Law COVID Behind Bars Data
Project (Mar. 3, 2021), []. It was
pursued over the life of the Project by Data Science Fellow Michael Everett, as
described in Part IV below.
See Eireann O’Grady, Capturing a Piece of History: Stories from Advocates Who
Lawyered on Behalf of Incarcerated Individuals During the COVID Pandemic,
UCLA Law COVID Behind Bars Data Project (Jan. 26, 2023), https:// []. Other
organizations—including, spearheaded by Keramet Reiter
and her team at UCI, and Mourning our Losses, an all-volunteer organization
dedicated to memorializing everyone living or working in carceral facilities
who lost their lives to COVID—were doing the vital work of capturing and
recording the experience of people incarcerated during the pandemic. Prison
Pandemic, [];
Mourning Our Losses, [
QH3G-UZUU]. Given our limited resources and our location in a law school,
we decided to focus on the lawyers, which would create a record of interest to
those who practice and study public interest law.

COVID Behind Bars Data Project


The speed with which the Data Project exploded into being and the
relentlessness of the work made it impossible in the moment to reflect
too deeply on our organizational design, management, or strategy. But as
the crisis has ebbed, there has been time to consider how we were able to
do all we did in such a short time and with such limited resources. Much
of my motivation in writing this Essay is to share what I regard as the
factors that made our work possible, in the hope that others inclined to
launch social justice organizations, both inside and outside the academy,
might find useful lessons from our experience.


Why a Law School?
I begin with our location. Some observers of our work might
wonder: What was an initiative reliant on data science and dedicated
to gathering and analyzing data on prisons, jails and detention centers doing in a law school? Why did this enterprise not emerge from a
criminology department or a policy school, perhaps in conjunction with
colleagues in STEM?72
In fact, the Data Project arose and flourished, not despite its home
in the legal academy, but because of it. At every level, our undertaking
was infused with the insights a legal education provides. By studying
legal cases, law students develop a deep and organic understanding of
where power resides in the American legal and political system—knowledge essential to meaningful change. Lawyers understand political
institutions: how they operate, how they fit together, where the authority
comes from and how it gets exercised. Data science offers a set of skills
that may be mobilized in the service of understanding social and political
phenomena. First, however, it is necessary to know the questions to ask
and where to look for answers. Over the life of the Project, the lawyer’s
inherent grasp of the operation and arrangements of state power formed
the foundation for all our decisions: what steps we ought to take, what
inputs we required and where to find them, to whom we should reach out,
and how to best help those who reached out to us. Those who contacted
us for help included lawyers bringing all manner of claims, journalists
trying to convey what was happening in real time, and actors from every
level of government. Every single day, in all our work, we drew on the
deep institutional understanding a legal education brings.
Lawyers, moreover, care about facts. We instinctively seek not
just a vague, general sense of what is going on, but the specific, concrete
details. Given this ingrained instinct, it is no accident that it was lawyers
who saw the value of tracking multiple data metrics bearing on COVID
in custody. Lawyers are also trained to think laterally, to analyze with
rigor and precision and to recognize quickly which dynamics matter and

Supra discussion in Part I. As Part I explained, one part of our work was
focused on coding the judicial opinions bearing on COVID in custody. But that
undertaking was an artifact of the Project’s location in a law school, not the
driving reason for it.




which players outside the law are necessary partners in any enterprise.
We regularly deployed these skills day to day and in shaping the Project
as it grew. And we not infrequently had cause to be grateful for the legal
training of so many of our volunteer corps, on whom we grew increasingly
dependent for the success of many of our initiatives. Our law student
volunteers brought to the table more than a knowledge of the law; we
also relied on their analytical rigor, their attention to detail, and their
ingrained understanding of state institutions and how they work.
And there is yet another aspect of the legal academy that, although
seemingly arcane, played a significant role in the emergence, orientation,
and success of our efforts: the ecumenical character of legal scholarship. Law schools are famously dichotomous. They are professional
schools, training the next generation of legal practitioners. They are also
academic institutions, in which law professors produce abstract and theoretical work with no necessary direct relevance to legal practice. Today,
many American law professors have PhDs in the humanities and social
sciences, and interdisciplinary scholarship bearing on legal themes is not
only tolerated but affirmatively welcomed. My own writing is emblematic. Although I am licensed to practice law in two states, I have never
represented a client or tried a case. My graduate training is in political theory, and a good part of my work is highly theoretical and overtly
normative, focused on the fundamental question of what the state owes
the people we incarcerate. I have approached this inquiry from various
angles,73 and in each instance, I have been less interested in what courts
or legislators might realistically be convinced or expected to do74 and
more concerned with what the state officials who function within the various branches of government ought to do—an exceedingly impractical
question.75 Yet my scholarly work positioned me perfectly to lead the
Data Project—a very practical, entirely policy-oriented effort.


Methodologically, my approach has ranged from policy assessment to doctrinal
analysis to qualitative empirical research to what might be thought of as pure
political theory. See Sharon Dolovich, State Punishment & Private Prisons, 55
Duke L.J. 437 (2005) [hereinafter Private Prisons] (policy assessment); Prison
Conditions, supra note 40 (policy assessment); Sharon Dolovich, Evading
the Eighth Amendment: Prison Conditions and the Courts, in The Eighth
Amendment & Its Future in a New Age of Punishment 133 (William Berry &
Meghan Ryan eds., 2020) (doctrinal analysis); Sharon Dolovich, Cruelty, Prison
Conditions, and the Eighth Amendment, 84 N.Y.U. L. Rev. 881 (2009) [hereinafter
Cruelty] (doctrinal analysis); Strategic Segregation, supra note 40 (qualitative
empirical research); Two Models, supra note 40 (qualitative empirical research);
Sharon Dolovich, Legitimate Punishment in Liberal Democracy, 7 Buff. Crim. L.
Rev. 307, 314 (2004) [hereinafter Legitimate Punishment] (political theory).
Unless it is to critique current practice. See, e.g., The Coherence of Prison Law,
supra note 42.
See Legitimate Punishment, supra note 71, at 314 (defining as liberal democracies
those polities claiming a commitment to the “baseline liberal democratic values,”
including “individual liberty, dignity and bodily integrity, limited government,
and the primacy and sovereignty of the individual” and arguing that “on this
definition, the United States, the political life of which is routinely punctuated

COVID Behind Bars Data Project


There are two features of my academic writing that explain this
seeming paradox, both of which bear on the value of legal scholarship
and how it can enhance the potential of law schools and legal academics
to play a leading role in the push for social and political change. The first
is grounded in the lawyer’s commitment to facts. However abstract or
theoretical my work, the touchstone of all I have written has been the
concrete reality of life as experienced by the people we lock away. This
foundation in the actual is by design. As I see it, we cannot grasp the
implications of our current regulatory structure or imagine alternative
approaches unless we understand the ways incarceration dehumanizes,
degrades, and brutalizes the fellow human beings who live it every day.
The Data Project’s mission of gathering as much information as possible regarding the on-the-ground impact of COVID on those held in
carceral facilities naturally emerged from my longstanding commitment
to understanding and acknowledging the suffering of people inside and
to wrestling with how best to alleviate that suffering.
The second way my scholarly writing drove the Data Project stems
directly from its manifestly normative character. For me, the universal recognition of shared humanity is the defining moral imperative of
collective life. Its application to the prison context seems both straightforward and incontestable: the people we lock behind bars are fellow
human beings and fellow citizens who by virtue of this status alone must
be treated as within society’s moral circle. This commitment guides
everything I write. For more than two decades, I have mapped the way
a failure of such moral recognition is baked into the laws, policies, and
politics regulating the American carceral state.76 As a direct result of this
work, when the pandemic hit, I was already acutely aware of the many
ways callous indifference to prisoners’ humanity has long shaped the regulation and oversight of American prisons and jails. I was consequently
positioned to immediately understand the web of institutional and legal
forces sure to array themselves against any efforts to shield people inside
from the disproportionate risk of COVID infection and death they were
sure to face. This understanding, coupled with my long-standing commitment to the equal moral worth of all people, motivated every step I took
from the moment I first shared that two-tab spreadsheet and throughout
the life of the Data Project.
As I have explained, I was not guiding this effort alone. My partner
in leading the Project, Aaron Littman, came to UCLA as an experienced prison litigator with a practical skillset I lacked. Aaron’s deep
practical/institutional knowledge, wholly informed by the same moral


with the rhetorical invocation of these very values, qualifies as an aspiring liberal
See, e.g., Sharon Dolovich, Exclusion and Control in the Carceral State, 16
Berkeley J. Crim. L. 259 (2011); Cruelty, supra note 71; Failed Regulation, supra
note 33; Dolovich, The Coherence of Prison Law, supra note 42; Incarceration
American-Style, supra note 40.




commitments that ground my own work, proved vital to our success at
every level. Our complementary understandings and shared knowledge
base embody the combined strengths of the American law school more
generally and positioned us as effective co-stewards of the Data Project
as it grew.
In those early pandemic days, we were hardly the only ones who
mobilized. To the contrary, we joined a broad network of actors and
organizers—some ad hoc like us, others more long-standing—who collectively recognized a humanitarian crisis in the making and felt an
urgent need to respond. This was an all-hands-on-deck moment. Our
embeddedness in this network and the unique role we were able to play
in this massive shared effort helped to demonstrate what scholars of
law and social movements have long been arguing: law and lawyers can
have the greatest impact on the push for social justice if we understand
and recognize ourselves as partners in a broader community of actors,
working in concert with—and guided by the priorities and experiences
of—the people most directly impacted by the social and political forces
we aim to shift.77 Our experience also makes a strong case that the law
school approach of combining the theoretical/normative with the practical/hands-on can offer tremendous payoff in the service of social justice.
In the push for decarceration and other policy responses to the crisis of COVID in custody, the Project did not play the part of a traditional
lawyer. We were not in court litigating on behalf of incarcerated clients;
indeed, we had no clients. Instead, our role may best be understood
as a species of what Scott Cummings, in his rich analysis of movement
lawyering, has called “integrated advocacy.”78 Through this approach,
lawyers pursue all available strategies “to maximize political pressure
and transform public opinion.”79 As part of a national effort, our strategy
of choice was not litigation, but compiling as much data as possible to
support the litigators and others seeking to publicize the crisis and keep
prisoners safe.
But even here, our experience sits somewhat outside the model.
The fight to protect the incarcerated from COVID infection and death
was not the sort of strategically coordinated campaign, led by what Cummings calls “mobilized clients,” that typically comes to mind in discussions
of movement lawyers.80 Unlike, say, the push for marriage equality (per77.


See, e.g., Gerald P. López, Shaping Community Problem Solving Around
Community Knowledge, 79 N.Y.U. L. Rev. 59, 60 (2004) (describing “the
importance of community knowledge to effective community problem solving
of all sorts, seen from diverse perspectives—nonlegal, legal, and both.”).
Scott Cummings, Movement Lawyering, 2017 U. Ill. L. Rev. 1645, 1695–96
Id. at 1696.
Id. at 1660 (“By aligning with mobilized clients, movement lawyering embraces
a strong version of lawyer accountability to democratically led collectives that
themselves claim to stand in for broader constituency interests.”); id. at 1691
(explaining that movement lawyering “depends on lawyer accountability to

COVID Behind Bars Data Project


haps the archetypal instance of such coordinated efforts),81 the response
to COVID in custody was far more hastily constructed, with nodes of
advocacy independently bursting forth around the country. The immediate goal of all this activity was plain: to minimize as much as possible
the harmful impact of COVID on incarcerated people. Any coordination
among disparate efforts occurred only once people had begun pursuing
their chosen strategy and somehow found the bandwidth to look around
and begin building connections with others rowing in the same direction. And even then, it was less about coalition building among the many
players—advocates, activists, journalists, family members, grass roots
organizers, and others—who had independently carved out a role in the
fight, and more about supporting each other as best we could while trying
to keep our heads above water. In this collective effort, the Data Project
played a dual role. We were one advocacy shop among many. And we
were also enablers of the whole emergent network, as the urgent need for
data defined virtually every strategy pursued on behalf of the incarcerated during the pandemic.
The experience of the Data Project highlights the possibility of a
more expansive role for lawyers and law schools in responding to urgent
crises of social justice. But more than this, it reveals the potential for
crisis mobilization that can emerge from an institutional confluence of
legal training, resources (both financial and human), and subject-matter
expertise. Today’s law schools are frequently sites of such potential, often
crystalized in programs, projects, or centers. Enterprises of this sort have
proliferated in American law schools in recent years, as academics with
expertise in real world issues have oriented their efforts towards knowledge dissemination and specialized training for students in the service of
policy change.
Just such an initiative lies at the root of the UCLA Law COVID
Behind Bars Data Project. Over the past decade, the UCLA School of
Law has become nationally recognized for our rich programming and
deep curricular offerings bearing on the law, policy, and operation of the
American carceral state. The UCLA Prison Law and Policy Program, a
hub for multiple initiatives at the law school bearing on these issues, has
made us a destination for JD, LLM, and MLS candidates who care about
the treatment of people caught up in the criminal legal system, both
while they are in custody and once they are released. As a result, as I
had hoped when I first launched the Program, a commitment to issues of
carceral law and policy has become encoded in the DNA of UCLA Law,
with the effect that all members of the community—students, faculty, and


mobilized clients that play a leadership role in social change campaigns”).
See, e.g., Scott L. Cummings & Douglas NeJaime, Lawyering for Marriage
Equality, 57 UCLA L. Rev. 1235 (2010); Nan D. Hunter, Varieties of Constitutional
Experience: Democracy and the Marriage Equality Campaign, 64 UCLA L. Rev.
1662 (2017); Tom Watts, From Windsor to Obergefell: The Struggle for Marriage
Equality Continued, 9 Harv. L. & Pol’y Rev. Online S52 (2015).




staff—recognize this set of concerns as integral to the life and mission
of the school. Having laid this programmatic groundwork, we were able
to mobilize quickly and efficiently when COVID hit, confident we had
the institutional and cultural support required to make the gambit a success. Other law schools will have different signature issues that could
drive similar efforts in other contexts. But if any law school was likely to
become a site for an energized collective response to the specific aspect
of COVID on which we focused, it was UCLA.82

III. Tracking a Pandemic Behind Bars: What We Got Right and
Lessons Learned
The clichéd descriptions of pandemic response—drinking from a
fire hose, building the plane while flying it—very much applied to the
Data Project. There was much we did that was less than ideal, not least
having no dedicated executive director, which forced our full-time staff
to wait until Aaron or I could carve out time from our other institutional
responsibilities to provide needed supervision and guidance. This problem often slowed us down, causing the members of our staff unnecessary
frustration and almost certainly costing us in terms of opportunities lost,
initiatives stalled, and timeliness of responses compromised.
Still, there were many things we got right, which allowed us to
grow rapidly yet efficiently and to accomplish all we did in such a short
time. In this Part, in no particular order, I identify some aspects of
our approach that enabled us to succeed in those components of our
work that met or exceeded our expectations. Some of this success was
no doubt a function of the pandemic itself—had there not been a society-wide lockdown, it is unlikely we could have recruited the army of
part-time volunteers that wound up providing crucial support. But on
reflection, it is possible to identify several features of our organizational
approach that were not pandemic-dependent and which may have traction more generally for other efforts not undertaken during a global
viral emergency.
1. A Clear Mission: As Part I indicated, it was not obvious in the earliest days of the pandemic what shape the Project would eventually take.
Even so, from the moment the spreadsheet expanded, I found myself with
a clear twofold mission. The first aim was to gather in one place as much
data as possible concerning the impact of COVID behind bars for the use
of any party wanting to help protect—ideally through decarceration—the
people then locked inside. The second purpose was more long-term. It
was clear from the start that COVID was a world-historic event on the
order of the world wars of the twentieth century. And as with those other
global crises, people would be studying and analyzing every aspect of the

I thank Scott Cummings for helping me recognize the way the UCLA Prison
Law and Policy Program and its long-standing role in the law school ecosystem
seeded the ground for the Data Project, making it possible for us to mobilize
quickly and effectively from the very start of COVID.

COVID Behind Bars Data Project


pandemic for decades to come. Our repository, I felt, should be a locus
for future research, a rich source of as much real-time data as possible for
those researchers and journalists who, once the pandemic was over, would
be looking to make sense of what happened in carceral facilities during
COVID.83 Keeping this two-track mission top of mind was a source of
great clarity as the Project grew and changed. It allowed us to readily see
which opportunities and proposals of the many that emerged were ones we
should pursue and which we should leave for others.
2. A Guiding Ethos: As has already been noted, from its inception,
the Data Project was grounded in fundamental normative commitments
regarding the shared humanity of people in custody and society’s consequent obligation to protect the incarcerated from unnecessary harm.
These moral commitments constituted our North Star, shaping our work
at all levels, including our insistence on enhanced data quality and transparency and our commitment to decarceration as the most appropriate
way to reduce the COVID risk inside.84 But there were other, more
practical guiding principles that emerged from the work and helped us
make decisions on the fly, trusting that we were getting things right. One
central example concerned the data itself. It is standard practice for university research labs to treat their data as proprietary, so that those who
collect or compile it will have exclusive access to the analysis and publication of any findings. But we were not in this work for scholarly credit.
What mattered was that the data be shared and analyzed, and its implications widely understood as quickly as possible. From the start, we had a
mantra: the data wants to be free. Anyone who reached out to ask if they
might use any of our posted data in their own research got an immediate
answer in the affirmative. And once our web scrapers were up and running, we made it our practice to clean and post all our quantitative data in
CSV files on GitHub,85 accessible to anyone wanting to study it. Our data
mantra, born of our mission, was wonderfully clarifying during a period
offering so little time for reflection.
3. Horizontal Structure: As Part I explained, our organization was
built on the fly. There was so much to do, and it was all coming at us so
quickly, that we could only function if the work was broken into chunks
and parceled out to people who would take the lead on each front. As



It is for this reason that we have taken steps to ensure that every piece of output
we produced over the life of our COVID work will be fully and permanently
accessible on our project’s Dataverse and also in the ICPSR database. UCLA
Covid Behind Bars,
[]; Inter-Univ. Consortium for Pol. & Soc.
Rsch., [].
Many on the team, myself included, were already committed to the goal of radical
decarceration—i.e., the drastic reduction in the size and scope of the American
incarcerated population. This goal dovetailed with our judgment as to the most
effective means of protecting people in custody from the disproportionate risk
they faced from COVID.
See UCLAlawcovid19behindbars/data, GitHub,
covid19behindbars/data [–8UK4].




the Project coalesced, our organizational structure remained largely horizontal, with each sub-team running its own operations and supervising
its own volunteers.86 There was a downside to this approach, especially
for the team leads, who in many cases were also working or in school fulltime and so were not always able to effectively leverage their volunteers
and complete their projects. What’s more, Aaron and I often had only a
skeletal sense of what was going on, and perhaps it might have been better if we had been more plugged in.
But this structure had a strong upside. Each team lead and each
staff member enjoyed a high degree of autonomy and consequently a
sense of ownership over their piece of the work. This situation seemed
to promote a tremendous amount of creativity in the crafting of projects and strategies to further the broader mission. I came to enjoy
watching team members build out their own corners of the Project in
exciting and meaningful ways precisely cohering with our goals. This
was the case with Michael Everett, one of our first Data Science Fellows. Noticing the shortcomings of the data being posted on agency
dashboards, Michael launched an in-house public records practice to
try to fill the gaps. Begun as a collaboration with our investigator Victoria Rossi, this effort eventually grew under Michael’s supervision to
comprise a central piece of our work. It allowed us not only to enhance
our data but also to build out a host of independent datasets87 that—
unaccountably—had never before been compiled.88 Among other
contributions, Michael’s remarkable leadership and tenacity in this
space has enabled our more recent focus on all-cause carceral mortality, on which more below.89
Then there was the demanding work of building and running our
scrapers. By the fall of 2020, this job primarily fell to our Data Team, then
comprised of, along with Michael, data scientists Hope Johnson, Neal
Marquez, and Erika Tyagi. Together, day in and day out, the Data Team



See Appendix (Project Org Chart, March 9, 2021.
These datasets include: a comprehensive list of every facility in our COVID data,
crosswalked to the most recent BJS census data and HIFLD prison boundaries
data; nationwide agency and facility-level population data; PREA facility and
physical plant data, also containing the most recent PREA audit report for
each facility for which we collect COVID and all-cause mortality data; and our
prison all-cause mortality database. We are also in the process of building out
a comprehensive dataset containing agency and facility-level demographic data.
None of this dataset construction would have been possible without Michael’s
ingenuity, exceptional data management skills, and remarkable facility with the
public records process.
The Bureau of Justice Statistics (BJS) and the Department of Homeland
Security (DHS) have produced some datasets in this space. However, several
of the datasets we have created did not exist previously, and those that existed
in some form are in our versions more comprehensive than any of the BJS and
DHS datasets. Our data also allows for facility-level analyses that were not
previously possible.
See infra Part IV.

COVID Behind Bars Data Project


handled the scraper challenge with remarkable efficiency and ingenuity.
Although our summer 2020 web scraper boot camp had launched our
data scraping effort, the agencies were forever altering the presentation
of the data on their dashboards, not to mention degrading the quality of
the data they were posting. As a result, Erika, Hope, Michael, and Neal
were constantly forced to come up on the fly with creative new ways to
rebuild broken scrapers and to capture and clean the data itself. In keeping with our horizontal structure, this was their bailiwick. We left them to
it, and they outdid themselves.
Here is another example of the way good things happened when
everyone ran their own initiatives: at one point, I vaguely knew that our
immigration team was speaking to a New York Times video journalist for
a report on COVID in immigration detention. When the documentary
came out,90 I watched with delight as data team member Neal Marquez
spoke on camera at considerable length, talking reporters through the
COVID data for ICE facilities displayed on our website. Our immigration team leads at the time, Theresa Cheng and Ishan Nagpal, had been
working assiduously with a large team of volunteers to excavate the realities of COVID in ICE detention and to digest their findings for public
consumption. Through a collaboration with Neal, their designated data
scientist, the immigration team had been able to create the visualizations that made their findings instantly accessible.91 This work formed
the foundation of the Times video report and accompanying news story.92
Our horizontal organizational structure—born not of long reflection but
of necessity—gave them, and all our team leads, the license and autonomy that enabled many successes of this sort.
4. Bi-Weekly All-Team Meetings: Partly because of our horizontal
structure, and partly because of the pace at which things moved, the Project suffered from the start from a chronic insufficiency of communication
among team members. Put simply, people always felt like they had no
idea what was going on outside their corner of the work. I even felt that
way. Better communication was plainly needed. Among other things, it
would ensure that Aaron and I knew what steps people were considering
so that we could weigh in, discuss potential issues, and redirect efforts
where necessary. In an effort to improve things, in early May 2020, we



See Isabelle Niu, Emily Rhyne & Aaron Byrd, How ICE’s Mishandling of
COVID-19 Fueled Outbreaks Across the Country, N.Y. Times (Apr. 25, 2021), []; Maura Turcotte, Virus Cases Are
Surging at Crowded Immigration Detention Center in U.S., N.Y. Times (Sept. 6,
html [].
Also with Neal’s help, the immigration detention team was able to build and run
the scraper that extracted the COVID data for ICE detention from the DHS
See Niu, Rhyne & Byrd, supra note 87; Turcotte, supra note 87.




instituted bi-weekly all-team Zoom meetings.93 Although the structure
of these meetings varied over the life of the Project, the core agenda was
the same: each team lead reported on what they were doing, and together
we explored in detail any issues, prospects, or concerns raised by their
undertakings before moving on to the next team.94
These meetings served many ends. They helped me and Aaron stay
up to date on what was going on. They allowed all parts of the team
to learn what others were doing. They also enabled collaboration and
learning across teams. People who were contemplating some strategic
initiative might discover that other teams had already tried the same
thing. Or we might find that two teams were thinking the same thing at
the same time, which would make it possible for them to work together.
Or someone might propose some idea or approach, only to learn from
our data scientists that they couldn’t do it quite that way, but that perhaps it would be possible to do it some other way. These meetings were
intense and often draining, but they always left me feeling exhilarated
and in awe of all that was being accomplished. They constituted a vital
channel for supervisory input and for the cross-team communication and
collaboration necessary for a Project that was fully remote and growing rapidly.
5. Shared Credit and a Culture of Collaboration: From the start, a
central tenet of the Project was that everyone involved had expertise
and insight from which we could all collectively benefit. Another was an
openness to what I thought of as creative gambits: everyone was welcome
to pursue their own initiatives and to own their own successes. Proposed
innovations were collectively vetted, and anyone contemplating significant new projects needed to run them by me and Aaron. But so long as
the proposals were consistent with our mission and seemed doable given
available resources, the answer was always yes. And once the projects
were launched, we did all we could to promote their success, while making sure that those who spearheaded them received the credit for what
they accomplished.
After we started our blog, we encouraged team members to write
up any findings, insights, or issues they believed warranted attention and
to publish them on our platform.95 We rarely published anything without

This being the first months of the pandemic, we were fully remote. All our
meetings were on Zoom.
There were also weekly meetings just for staff, which took a similar and equally
fruitful form.
My experience with the Project left me perpetually amazed at how much a small
group of dedicated, motivated people can accomplish when they combine creative
thinking with a realistic sense of what is possible. Those who might wish to get a
feel for what came from this dynamic could do no better than to scroll through
our blog, which became the repository for many investigations and analyses
conceived by team members who noticed gaps or patterns that bore pursuing.
COVID Behind Bars Data Project: Blog,
blog [].

COVID Behind Bars Data Project


editing and comments from me or Aaron (as well as our Communications Manager Josh Manson), but in terms of topic choice, framing, and
argument, authors largely made their own calls. Our horizontal structure
also encouraged intra-team collaborations. At internal staff meetings
(which Aaron and I did not attend—who had time?), members of the
staff floated ideas and pursued them in ever-changing configurations,
leveraging their different skill sets and expertise to best serve each idea.
And credit was accorded where credit was due: every piece was edited by
several sets of eyes, but it was those team members who conceived, investigated, and/or drafted each post or report who received co-authorship.
It is a source of pride to me that many people on our staff left the Data
Project with resumes full of publications and accomplishments sure to
enhance their future opportunities.
6. Effective Volunteer Management: Almost immediately after the
Project launched, two things happened: we began receiving emails from
people all over the country who were keen to help, and we became
aware that, without volunteer help, much of what we wanted to do
would be impossible. But managing volunteers takes time and effort,
and we knew that, if we were to successfully leverage these many offers
of assistance, we needed a feasible strategy. The system we developed
was extremely simple, but it worked. First, we designated a volunteer
coordinator (VC).96 Then we instituted a rule: any staff member or
team lead in need of volunteer help had to send the VC a brief written
description of the work their volunteers would be tasked with performing. Most of the emails from people offering help came to me, often
in response to general calls for volunteers I sent out to the law school
community as a whole. I would connect anyone interested in working
with us to the VC, who would send the prospective volunteer a Google
form listing the project descriptions the team leads had provided. The
volunteers would be invited to indicate their preferences on the form
and, relying on those preferences, the VC would connect them to the
staff member or team lead they would be working with. That was it. At
that point, the volunteers were folded into the work and supervised by
whoever initially requested their help.97



Rebecca Fordon graciously agreed to play this role as we launched. She
eventually handed it off to Danielle Flores, who joined the staff as Project
Coordinator in August 2020. Danielle served in the VC role during the height
of the Project’s size and ambition, and without her otherworldly administrative
abilities, the whole volunteer operation would have fallen apart. Danielle stayed
with us until she went to graduate school in the fall of 2021. At that point, the
task fell to Cece Bobbitt, who took over from Danielle as Project Coordinator.
If any issues arose, the VC served as the point person. These issues tended to take
one of two forms. Either people who committed to the work would disappear, in
which case the VC would find others to take their place, or else volunteers came
back to the VC indicating they would prefer work of a different sort, in which
case the VC would reassign them to another team.




It is hard to overstate the effectiveness of this system. It is the reason why our org chart, continually updated, was so jam-packed, and the
explanation for how we were able to do so much on so many fronts with
so few resources in such a short time.98 Our volunteer management system also allowed us to easily leverage the particular skillsets of the people
who reached out, for example, routing volunteers with data science skills
directly to the Data Team.99 In several cases, people who signed on as
volunteers wound up authoring reports on topics we had hoped to investigate but lacked the staff resources to pursue.100 Others who started as
volunteers eventually joined the Project as team leads101 or staff.102


Our org chart dated March 9, 2021, a year into our work, clearly illustrates
just how great a contribution volunteers made to the Project. See Appendix
(Project Org Chart, March 9, 2021). It shows twelve volunteers working with
Grace DiLaura on the Prison Policy Index, and twenty-one volunteers working
with the data team on a host of projects, including data collection, populating
facility crosswalks, filling in the backlogged historical death data, conducting a
gender analysis, helping with data visualizations, filling out our SQL database,
fixing the scrapers, assisting with our public records work, and investigating
data patterns in private prisons and in psychiatric facilities. Our “qual team,”
comprised of Liz DeWolf (Project Manager), Josh Manson (Comms Manager),
Poornima Rajeshwar (Policy Fellow), and Victoria Rossi (Investigator) had three
volunteers helping with social media, three investigating state testing initiatives,
four working on our oral history project, and ten working on public records
requests, family outreach, and medical data entry. Another forty volunteers
were working with our sub-teams: three on releases, three on court filings, eleven
on youth facilities, nine on the grassroots team, and the rest on various projects
with the immigration detention team.
99. The Google form shared with prospective volunteers had a place for people to
indicate any particular skills or expertise they would bring to the work. We paid
careful attention to this information.
100. See, e.g., Alix M.B. Lacoste et al., “Horrible Here”: How Systemic Failures of
Transparency Have Hidden the Impacts of COVID-19 on Incarcerated Women,
UCLA Law COVID Behind Bars Data Project (Nov. 23, 2021), https://
[–8WFF]; Aparna Komarla, COVID-19 Vaccination Data
in California Jails: Lessons from an Imperfect Model, UCLA Law COVID Behind
Bars Data Project (Nov. 17, 2021), []; Alix M.B.
Lacoste, Erika Tyagi & Hope Johnson, Fast, Frequent, and Widespread: COVID-19
Outbreaks Inside Federal Prisons, UCLA Law COVID Behind Bars Data
Project (Nov. 2, 2021),
[]; Andrea Allen, Minali Aggarwal & Neal
Marquez, As Long as There Is COVID-19, ICE Detention Centers Will Be Poised
for Disaster, UCLA Law COVID Behind Bars Data Project (Jul. 27, 2021), [].
101. Ishan Nagpal and Ben Woolley joined as volunteers with the immigration team
and ended up as team leads.
102. Cece Bobbitt joined the grassroots team as a volunteer and wound up as team
lead and then as Project Coordinator. Cece’s energy, organizational capacity, and
ability to handle any challenge were legendary.

COVID Behind Bars Data Project


Our volunteer management system allowed us to greatly expand
the work of the Data Project. And it had a further positive effect that
may in the long run prove even more consequential: it exposed hundreds
of people, most of them students in college or law school, to the daily
experience of people in detention. Some of our volunteers were drawn
to the work by a previously developed sense of the injustice of mass
incarceration and a desire to shrink the state’s carceral footprint and
improve the lives of those people remaining behind bars. But many had
no prior knowledge of the issues and were drawn simply by the chance to
do meaningful work during the COVID lockdown. Whatever corner of
the Project they worked on, they will have been exposed to something of
the nightmare experience of being in custody, whether during COVID or
otherwise. For some, their work with the Project may well have seeded
a lifelong commitment to work on behalf of the incarcerated. And even
for those who never again work on carceral issues, this exposure will have
sensitized them to a site of state power that no one concerned with social
justice can ignore. I count this effect as one significant long-term benefit
of our work.

IV. The Gradual Fading of COVID Prison Data and the Pivot to
All-Cause Mortality
In the first weeks of the pandemic, corrections agencies began to
voluntarily report their COVID data. This was a dramatic departure
from the usual way prisons and jails manage information. The standard approach to data-sharing by prison and jail administrators has long
been to treat all information concerning life inside as proprietary, to be
accessed only by corrections officials.103 Whatever drove the DOCs to
depart from this practice and create public-facing COVID dashboards,
the official reporting of COVID data for carceral facilities helped reveal
an otherwise hidden world and made it possible to confirm in real time
the outsized harm the virus was causing those in custody. More than anything else, it was this real-world development that enabled our work, and
helped journalists and others force onto the public radar a realistic picture of the inhumane conditions and consequent suffering experienced
daily by the incarcerated.
Yet if the COVID prison data provided a rare and welcome window into what was happening inside carceral institutions, the data itself
was deeply inadequate across the board.104 A year in, in March 2021, the
103. As I have argued at length elsewhere, this culture of secrecy is wholly
inappropriate in a liberal democratic society. See Mass Incarceration, supra note
19, at *28–*30. Prisons and jails are public institutions, operated on behalf of
society as a whole. Corrections officials are not sovereign over the people in
their custody. They are public servants whose sole job is to administer carceral
facilities in ways consistent with the public interest. In order to ensure that the
public’s interests are served, we need a full and complete accounting of what
goes on inside.
104. The lack of data transparency was especially pronounced as to youth detention:




Data Project published our first scorecard, grading corrections agencies
on the quality and transparency of their published COVID data over
the first 12 months of the pandemic.105 Scores were compiled for twelve
metrics—eight addressing which data were being reported106 and four
bearing on the quality of that data.107 The results were not inspiring: of
the fifty-three agencies we assessed,108 one jurisdiction (West Virginia)
earned a B, eight jurisdictions received a C, five got a D, and the rest—
forty-one in total—failed.109 Over the next thirteen months, we regularly
revised our metrics to take account of changing conditions, for example,
adding reporting variables for vaccine data. Over time, agency performance either remained equally poor or got worse.110
Performance on our scorecard was only one measure of the progressively worsening quality of the data being reported and the increased
opacity of corrections agencies concerning COVID in their facilities. In
August of 2021, as the Delta variant surged nationwide, multiple DOCs
simply stopped reporting their COVID data—including prison systems in
states hardest hit by Delta.111 By October of 2022, twenty-one states and





nationwide, only 16 percent of youth facilities reported their COVID data,
and of county-run youth facilities, which constitute more than half of all youth
detention centers, only 1 percent reported their data. See Yücel, supra note 66.
There was also extremely limited COVID data bearing on county jails. There
are roughly 3200 county jails nationwide. Of these, scrapeable data was available
for 155, of which 121 are in Texas, where some COVID data was made available
through standardized reports from the Texas Commission on Jail Standards, a
state monitoring agency. See Tex. Comm’n. on Jail Standards, https://www.tcjs. []. But for the most part, the largest
county jail systems had scrapeable dashboards, which meant that, although the
vast majority of jails did not report COVID data, the jail data we were able to
collect covered a substantial percentage of the national jailed population.
Missing the Mark, supra note 64. This effort, and the ongoing work of updating
the scorecard and analyzing changes, was collectively undertaken by Senior
Project Manager Liz DeWolf, Policy Fellow Poornima Rajeshwar, and Senior
Data Scientist Erika Tyagi. See Update: Data Reporting & Quality Scorecard,
Rounds 1 – 6, UCLA Law COVID Behind Bars Data Project: Blog, https:// [].
Five of the transparency metrics in our first scorecard related to data concerning
prisoners: cumulative cases, cumulative deaths, cumulative tests, active cases,
and total population. Three of the metrics related to data concerning staff:
cumulative cases, cumulative deaths, and cumulative tests. See Missing the Mark,
supra note 64.
The data quality metrics asked whether the posted data were machine readable,
regularly updated, clearly defined, and contextualized historically.
Our scorecard rated the data quality and transparency of 50 state DOCs, the
BOP, ICE and the DC DOC. See Missing the Mark, supra note 64.
Missing the Mark, supra note 64.
By the sixth and last update in April 2022, West Virginia’s score had dropped
to a D, and Pennsylvania was the highest scorer, with the lone C. Update:
Data Reporting & Quality Scorecard, Round 6, UCLA Law COVID Behind
Bars Data Project (Apr. 7, 2022),
march22scorecard [].
See Dolovich, Tyagi & Marquez, The States That Lead the Nation, supra note 65.

COVID Behind Bars Data Project


DC had ceased entirely to report active case numbers in their facilities,
despite the presence of community outbreaks.112
The progressively poor quality of the data as the pandemic unfolded
meant that our data-scraping efforts produced diminishing returns. And
as in society in general, despite persistent COVID mortality in the prisons,
the increasingly endemic nature of the virus also suggested diminishing
returns for policy efforts focused exclusively on COVID. This situation,
which evolved over time, eventually signaled that the moment had come
for the Data Project to change course.
The question of our post-COVID future was one we had long wrestled with. The success of the vaccine and its rollout in early 2021 widely
raised hopes that, come summer 2021, the pandemic threat would entirely
recede. In the spring of that year, there was a brief period of calm, which
we used to backfill our data, focus on initiatives in progress, and contemplate the winding down of our work. Then things changed again, and
Delta, Omicron, and widespread vaccine denial made our efforts as relevant as ever. The summer and fall of 2021 proved to be some of our
busiest months. Still, the question of the Project’s post-COVID future
continued to loom.
In the end, the answer was obvious, with two pieces of our work
together pointing the way. The first piece was our emergent in-house
research agenda. Starting in 2021, Neal Marquez began digging into the
scale of COVID deaths in custody. Neal has advanced training in epidemiology and quantitative methods and was the perfect person to lead our
Project’s efforts to quantify the mortality impact of COVID behind bars.
With Neal at the helm, our team produced two studies, each of which had
awful, jaw-dropping results.113
The first focused on Florida. In it, we found a striking disparity
during the first year of COVID in the life expectancy drop among people
incarcerated in Florida prison as compared with Floridians in general.114
Specifically, while life expectancy across the state’s population fell by 2.7
years between the calendar years 2019 and 2020 (itself a catastrophic
decline), the drop among those incarcerated in the state over the same
period was 4 years, fully 100 percent of which was attributable to COVID
deaths.115 The second study addressed the question of racial disparity in



As we reported at the time, this trend was most pronounced in those states in
which Delta-driven infections and deaths were especially elevated, most notably
in Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, Nevada, and
Texas. See id.
See Tumendemberel & Stein, supra note 65.
See Neal Marquez et al., Racial and Ethnic Inequalities in COVID-19 Mortality
within Carceral Settings: An Analysis of Texas Prisons, 41 Health Affs. 1626
(2022); Neal M. Marquez et al., Life Expectancy and COVID-19 in Florida State
Prisons, 62 Am. J. Preventive Med. 949 (2022).
See Marquez et al., Life Expectancy and COVID-19, supra note 110, at 951.




COVID mortality in custody.116 Here, we focused on Texas, and found
that, of those people in Texas state prison who died of COVID over the
first year of the pandemic, mortality was considerably higher for Black
and Hispanic populations than for Whites: specifically, 1.61 times higher
for Blacks and 2.12 times higher for Hispanics.117 We viewed these studies as pilots. The obvious next step was to see if the findings also applied
to other states.
Meanwhile, in collaboration with successive teams of UCLA Law
students, Michael Everett had been quietly building out a remarkable
dataset on deaths in custody from 2015 forward.118 The database, painstakingly constructed from information obtained through public records
requests, contains death data for fifty state DOCs and the BOP.119 For
forty state prison agencies and the BOP, Michael also collected individual-level records that provided a specific date for each death, along with
detailed demographic data for the prison populations of twenty-eight
states.120 This data, which had never before been compiled, will allow
for a range of investigations into rates and disparities concerning deaths
in custody.121
Between Neal’s pilot research studies and Michael’s data repository, we realized that, going forward, we had the potential to play a
significant contributing role in the growing national conversation
around deaths in custody. Reinforcing this sense was the fact that, in
spring 2021, the Department of Justice’s Bureau of Justice Statistics
“quietly announced” that it would no longer “collect [] data on deaths of
incarcerated people.”122 In early 2023, we published our data repository
116. See Marquez et al., Racial and Ethnic Inequalities in COVID-19 Mortality, supra
note 110.
117. Id. at 1631. In our paper, we speculate as to the causes of the racial disparity we
identified. See id. at 1631–32. Possible explanations include racial disparities in
housing with greater population density, in medical co-morbidities, in the quality
of medical care provided to people in prison and in the way people interact with
a prison’s medical system. All these possibilities require further study.
118. See Michael Everett & Lauren Woyczynski, UCLA Law Releases New Database
to Monitor Deaths in U.S. Prisons with Funding from Arnold Ventures, UCLA
Law COVID Behind Bars Data Project: Blog (Feb. 18, 2023), https:// []
(mapping the dataset and identifying initial findings); Sharon Dolovich et al.,
UCLA Law Covid Behind Bars Data Project: Prison Morality Dataset,
[] [hereinafter Prison Morality Database].
119. The dataset also contains pre-2015 ICE custodial death data collected by
Muckrock. Detainee Death Reports – 2016, MuckRock, https://www.muckrock.
120. Michael has also begun the task of accessing data bearing on mortality in county
121. The BJS dataset on carceral mortality only reported aggregated data. As a result,
it could not be used to analyze rates or disparities concerning deaths in custody.
122. M. Forrest Behne et al., When It Comes to Reporting Deaths of Incarcerated
People, Most States Break the Law, The Appeal (Mar. 2, 2022), https://theappeal.

COVID Behind Bars Data Project


on deaths in custody,123 and turned our attention away from COVID to
a new focus on all-cause carceral mortality. Our first step in this direction, taken before the public release of our dataset, was to work with
investigative reporters at the New York Times, who performed the first
comprehensive analysis of our data.124 The findings were at once chilling and unsurprising: during the first year of the pandemic, “[d]eaths
in state and federal prisons across America rose nearly 50 percent . . . ,
and in six states they more than doubled.”125 In keeping with what was
widely predicted from the start of the pandemic,126 the Times found a
marked disparity in outcomes for people in custody as compared with
the American population as a whole, with the jump in carceral mortality
in 2020 representing “more than twice the increase in the United States
overall,” greater even than the “percentage increase at nursing homes,
among the hardest-hit sectors nationwide.”127
These initial findings, published on the front page of the nation’s
paper of record, make the case for the power of data collection, analysis,



org/when-it-comes-to-reporting-deaths-of-incarcerated-people-most-statesbreak-the-law []. Under the Death in Custody
Reporting Act of 2013 (DCRA), all states receiving federal funding under
the Omnibus Crime Control and Safe Streets Act of 1968 (i.e., all states) are
required to report to the Attorney General certain basic information pertaining
to “the death of any person who is detained, under arrest, or is in the process of
being arrested, is en route to be incarcerated, or is incarcerated at a municipal
or county jail, State prison, State-run boot camp prison, boot camp prison that
is contracted out by the State, any State or local contract facility, or other local
or State correctional facility (including any juvenile facility).” Death in Custody
Reporting Act of 2013, Pub. L. No. 113–242, 128 Stat. 2860, 2860 (2013). Until
2019, it fell to the Bureau of Justice Statistics (BJS) to collect this data. Starting
in 2019, DOJ’s Office of Justice Programs (OJP) transferred this responsibility
to the Bureau of Justice Administration (BJA). Unfortunately, for a variety
of reasons, the data being collected by BJA is sufficiently unreliable that it
is unlikely to be released publicly. See U.S. Senate Permanent Subcomm. on
Investigations, Uncounted Deaths in America’s Prisons & Jails: How the
Department of Justice Failed to Implement the Death in Custody Reporting
Act 12 – 13 (2022) (“When compared to [BJS], BJA’s data collection did not
capture any state prison deaths in 11 states or any local jail deaths in 12 states
and the District of Columbia . . . BJA’s collection included only 38.9% of local
jail deaths and 66.3% of state prison deaths” that BJS collected . . . . There were
[also] various data quality concerns with BJA’s collection, such as inaccuracies
and missing fields. For example, 56 of the deaths reported to BJA as deaths
during arrest had actually occurred in jails and prisons when reported to BJS.”).
See supra Prison Morality Database, note 115.
See Jennifer Valentino-DeVries & Allie Pitchon, As the Pandemic Swept America,
Deaths in Prison Rose Nearly 50 Percent, N.Y. Times, (Feb. 19, 2023) https://www. [].
See, e.g., Mass Incarceration, supra note 19, at *4 (“From the earliest days of the
pandemic, it was clear that the novel coronavirus posed an outsized danger to
the more than two million people locked inside America’s prisons and jails.”).
Valentino-DeVries & Pitchon, supra note 121.




and dissemination as a strategy for “maximiz[ing] political pressure
and transform[ing] public opinion.”128 In our new focus on all-cause
carceral mortality, we are supported by a generous grant from Arnold
Ventures and by the UCLA Office of the Vice Chancellor for Research.
Again, we are working in coalition with many partners, including Professor Andrea Armstrong, whose leadership on this issue has sparked
a national effort to track deaths in prisons and jails.129 And again, our
aim in this new venture will be to gather and disseminate as much data
as possible on a matter of vital policy import, while continuing our own
efforts to analyze and thus to make sense of what is happening on the
ground to the fellow human beings held in American carceral facilities.
As we do this work, we will be drawing on, and implementing, some of
the organizational strategies that helped us accomplish all we did over
the almost three years of the UCLA Law COVID Behind Bars Data
Project.130 And in so doing, we will continue to model a central claim
of this essay—that not only lawyers but also law schools and legal academics have a central and perhaps unique role to play in the push for
social justice.
A crisis has a way of laying bare realities that may otherwise be
obscured by layers of institutional myth and misunderstanding. Among
other things, the work of the Data Project helped to expose as indefensible the culture of secrecy that has long reigned in American carceral
spaces—and to reveal the urgent need for openness and transparency
around what happens behind the walls. Where darkness and obscurity are allowed to persist, state actors may abuse their power and
thereby inflict gratuitous pain and suffering on the very individuals they
are sworn to protect. In a constitutional democracy, there can be no
greater anathema.131
Although there is no way to know for sure, I like to think that the
work of the Data Project helped to save lives. I hope too that our work
128. Cummings, supra note 76, at 1696.
129. See Incarceration Transparency,
[]. Armstrong’s path-breaking work in this space
can be viewed at Focused primarily
on deaths in custody in Louisiana and South Carolina, it provides a model that
may be (and is being) replicated in other states.
130. Indeed, one of these strategies—a policy of making our data freely available to
anyone who asks—has already borne fruit. The New York Times investigative
report on deaths in custody during COVID, published in February 2023,
represented the first analysis of the all-cause carceral mortality dataset that
will form the foundation of our work going forward. See Valentino-DeVries &
Pitchon, supra note 121. That report, with its blistering findings, was possible
only because we (readily) shared our data with the Times.
131. See Sharon Dolovich, How Prisoners’ Rights Lawyers Do Vital Work Despite
the Courts, 19 St. Thomas L.J. 435, 442 (2023) (discussing the work of Judith
Shklar, who “argue[d] that the preservation of liberal democracy, always a work
in progress, requires that we strive to minimize as much as possible the ability
of state officials to act cruelly or to instill the fear of cruelty in those over whom
they wield power”).

COVID Behind Bars Data Project


helped people living inside, and their loved ones on the outside, feel less
alone and abandoned during such a scary time. Of one thing I am certain:
through our support of journalists, advocates, and other stakeholders, our
efforts helped to force courts, policymakers, and the public as a whole
to recognize and acknowledge the outsized threat COVID posed to the
close to two million people then being held in America’s prisons, jails,
and detention centers.
Institutions can be slow to change. And American carceral institutions in particular are shaped and weighed down by a long history of
racism, brutality, and callous indifference to the human beings trapped
inside. At the best of times, the experience of imprisonment is degrading, traumatizing, and brutalizing. Ultimately, my deepest hope is that,
by contributing to efforts to expose the desperate experience of people
in custody during COVID, we have helped to shift public sentiment in
the direction it so urgently needs to go: toward recognition of the shared
humanity of those society has chosen to lock away, and the duty of care,
protection, and concern we collectively owe every incarcerated person as
a result of that choice.


(Last updated March 9, 2021)


11- - 1
I - 11--1
I ·--·..,......,..,..,..






PLN Subscribe Now Ad
Advertise Here 3rd Ad
Stop Prison Profiteering Campaign Ad 2