Good Intent is Not Enough

Two researchers sit at a table with incarcerated people at the Louisville Metro Government Jail.

And other things we learned responding to overcrowding in a Louisville jail

Over the course of 2021, we’ll be writing at least one blog post per month — and hosting virtual conversations — about work we’ve done, things we’ve learned along the way, and what we’d like to do in our next 10 years.

This is the second part of the series. Learn more about the project here.

In 2015, the Public Policy Lab began working on a project in Louisville, Kentucky, with the innovation team (iTeam) in the mayor’s office and the Louisville Metro Department of Corrections (LMDC). Our brief, provided by the mayor’s team, was to explore how to deal with overcrowding in the city’s jail.

Louisville’s jail facilities had a capacity of nearly 1,800 people, but at the time they were holding more than 2,200 — a surge driven by arrests related to the opioid epidemic. Because of crowding, people arrested for nonviolent offenses were being housed with people arrested for violent offenses. We also learned that the jail was struggling to respond to people who were experiencing opioid withdrawal while incarcerated, and that many people released from jail swiftly returned on new charges.

In the end, we delivered research findings and policy concepts to inform Louisville leadership’s decision-making: solid work, of mixed effect — more on that below. But the project did powerfully transform PPL’s practices, from how we interact with research respondents to what we look for in our partners.

The Project Team

The team for this project included designer Jen Thibault (who was about to join 18F, the federal government’s tech-innovation practice), researcher Kate Krontiris (who had prior experience with justice-systems issues from time with the Center for Court Innovation), and policymaker Camilla Buchanan (then at the Design Council, the UK government’s advisor on design). PPL’s director Chelsea Mauldin served as strategy lead.

The project marked our first attempt at remote mentoring, using both-in person visits to Louisville and video sessions with the iTeam, whose staff came primarily from policy, planning, and management backgrounds. This was their first engagement with human-centered research and design.

Our goal was to not just deliver a project outcome, but also to advocate for human-centered methods for policy development. To that end, we asked the iTeam members to deeply embed in the project’s work: they managed recruiting and scheduling of research sessions (both in the jail and with external stakeholders), participated in research with us (and conducted additional interviews on their own), and took part in synthesis and concept generation.

As Ted Smith, the director of the iTeam and a master of sardonic understatement, later told us, we managed to “convert [him] from a total skeptic to less skeptical” about the value of a design approach. We count that as a win, Ted.

Some of What We Learned about Jail

Our team had a broad remit to look at conditions in the jail, as well as the pipelines leading to jailing, and the release circumstances that drove quick returns to jail — in hindsight, too broad a task for our allotted time and budget.

We also came to understand that our partners had two different theories about how to best address the jail’s challenges: LMDC leadership believed that the aging jail had to be replaced with a more modern facility, while the iTeam was interested in whether system and service changes could reduce the jail’s headcount.

Over the course of several months, the project team conducted research with nearly 70 stakeholders — in jail, in the community, and in City government. We explored what factors lead to addiction, how jail serves as an incubator for future problems, what makes a jail safe (or not), and what it means to be ‘release ready.’

The Difference Between Jail and Prison (and Why It Matters)

Louisville’s jail — like most city or county jails — holds people who’ve been arrested but not yet charged, people awaiting trial, or people convicted of low-level offenses. Many of those held in jail remain there purely as a condition of poverty; if they had funds to secure bail, they would be released. Prisons, by contrast, are run by state or federal agencies and hold people convicted of more serious crimes, typically with sentences of more than one year.

The situation in Kentucky was further complicated. State laws require people convicted even of more serious felonies to serve their sentences in county jails, as a way of alleviating crowding at state prisons. Meanwhile, the state was being ravaged by the opioid epidemic. In 2015, Kentucky had the third-highest rate of drug overdose deaths in the country. Our partners told us that drug-related arrests, ranging from public intoxication to burglary, were on the rise.

So the overcrowded jail where we conducted research was both holding convicted felons it was not originally designed to hold, as well as unconvicted opioid users going through painful withdrawal symptoms. And, since it was a jail, not a prison, it offered very little in the way of addiction, mental health, education, or other programs. Typically, those services are provided in prisons, where people are expected to be held for longer durations, not in ostensibly short-stay jails.

How Jails Filled with Opioid Users

As the jail director pointed out to us, jails don’t choose whom they incarcerate. Police bring arrested people to the jail, and courts decide who stays and for how long.

During research in Louisville, community stakeholders told us that police (nationwide, not just in Louisville) have a perverse incentive to arrest opioid users and take them to the jail, rather than using pre-arrest alternatives to incarceration, like diversion to treatment centers. Each arrest might lead to a court date, where the officers would testify and accrue overtime, increasing both their earnings and ultimate pension.

Further, we heard that most of Louisville’s treatment programs followed abstinence-only models that health experts we spoke to decried as ineffective for opioid users; most people who attempt total abstinence just relapse, we were told. Evidence supports harm-reduction approaches, where users switch to maintenance medications such as methadone or buprenorphine. Despite an influx of opioid users, in 2015 the jail could not prescribe treatment medications.

Some stakeholders pointedly described increased interest by policymakers in opioid addiction and jailing as the effects of the epidemic came to be experienced by ever more white and upper-income families. It’s possible that they also understood our largely white research team in this context.

The Damage of Jail

Jailed people themselves described the destabilizing effects of detoxing in the jail, then of living in crowded dorms with dozens of bored, unhappy people and of the violence that sometimes flared, especially as people came back to jail after “a bad day” in court.

Some jailed people described challenging pre-jail lives of untreated trauma, mental illness, and poverty that slid into addiction. Others told us about solid careers derailed by injury, painkillers, and (when the prescriptions ran out) street drugs.

We heard how jail itself created conditions that led to repeat jailing: loss of jobs, health insurance, housing, family relationships, self-esteem. “Who’s going to hire me now?,” one person asked us.

We learned that jail did offer good opportunities to cultivate criminal skills and contacts. These resources often proved useful, since people were often released on short notice, lacking everything from proper clothing and a place to stay to a treatment plan for their substance addiction.

Jail Staff Weren’t Satisfied, Either

This project also confronted us with a common problem in intense frontline service environments: staff experience harms from the damaging systems in which they work.

Officers with an interest and training in security felt unprepared (and sometimes unwilling) to be healthcare providers. One outwardly stoic officer, when asked what one change to the jail would be most valuable, emphatically said “more mental health services” for inmates, and described his own stress and frustration.

It became obvious that for the jail to better serve the people it incarcerated, it had to also address the wellbeing of staff members. We later heard that the jail director said it was “the first time anyone was looking for a win for the officer and a win for the inmate in a project.”

Our Work Yielded Results?

If you’ve read this far, you can guess that we were in deep trouble. Our job was to provide concepts for addressing overcrowding in Louisville’s jail, but our diagnosed causes included a corrupt nationwide prescription-opioid scheme, America’s inadequate healthcare system, and the grotesqueries of the carceral state.

However, a rigorous design practice is a powerful tool: it will generate product, out of too much or nothing at all. Ultimately we delivered a set of findings, an assessment of the needs shared by multiple system participants, a sheaf of personas, and five policy-and-service concepts for reducing jailing, improving jail experiences, and limiting returns to jail. There’s a publicly available summary document.

Our concepts included models for providing support services and substance abuse treatment interventions keyed to duration of jail stay; reducing dangerous idle time in jail through educational and capacity-building activities; offering medically validated interventions, in jail and in the community, to stabilize people with substance-abuse issues; short-circuiting cycles of family damage by using data to pinpoint better services for families of jailed people; and providing incentives for police officers to divert non-violent offenders before arrest.

Ultimately, none of these five specific concepts were deployed as we proposed, but our recommendations provided groundwork for a series of changes over following years. Ted even said the work was “not entirely useless.”

The City began exploring how to make the jail facility and staff more responsive to people experiencing pain and stress, from environmental changes to mindfulness training. Some improvements were made to release procedures. The City and State governments began collaborating around pharmacy interventions to address people denied prescription medications, and in 2016 the jail began piloting provision of maintenance medications to jailed opioid users, an effort that’s continued to expand.

Some of What We Learned About Ourselves

Ted also gave us some blunt advice: “You shouldn’t really sell your product to people who can’t use it.” He meant that his team was not well-positioned to implement what we proposed.

We’ve taken this guidance seriously over the past five years, in two different ways. We’ve increasingly focused on undertaking projects with partners who have a demonstrated ability to implement policy and service changes, not only commission research and ideation. And we’ve continued to expand our own capacity to support system-wide implementation, from strengthening our development, piloting, evaluation, and scaling skills to getting smarter about having contracts include a ‘long tail’ of low-intensity implementation support.

Part of that commitment to implementation is also reflected in our now-routine practice of deep engagement with operations and public-facing staff. Public-service systems are made up of actors from policymakers to program administrators to members of the public. We are convinced that if you want to transform a system, then you need to support behavior change with and for all of these people.

Prior to the Louisville project, we had always engaged with service providers, but mainly as a means to an end — the end user, if you will. But going to jail made it super clear that in order for a system to meet the public’s needs, it has to meet staff’s needs, just as much.

Perhaps most profoundly, the Louisville project revealed that our ethical design practices were underdeveloped. We’d always tried to be good. But it wasn’t until we were doing research with people held against their will in jail that we had to assess if good intent was enough. (It’s not.) Now we’re on a forever project to better identify and change the power dynamics in our work.

Next Time: A New View of Consent

In our next post, we’ll show you how a pivotal moment in Louisville caused us to question one of our core practices. What does it mean to have informed consent from research participants?

The first time we joined the iTeam on the ground in Louisville, we went to visit the main jail. We expected to meet the jail director and maybe take a facility tour. We were unprepared when we were asked if we wanted to speak with some of the people incarcerated there.

We’ll tell you about how this moment made us reinvent how we seek consent from project participants — an approach we continue to use (and refine) years later. And we’ll make the case that consent offers researchers, designers, and policymakers a real-life tool to center agency, dignity, and honesty in the policy-design process.



Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Public Policy Lab

Public Policy Lab


The Public Policy Lab is a nonprofit innovation lab for government. We apply human-centered design methods to the challenges facing low-income people.