A Culture of Learning Can Defeat a Culture of Violence

I direct the NYU Marron Institute of Urban Management, a think-do tank that takes on critical challenges of urban living by working with governments and nonprofits to improve health, safety, mobility and inclusiveness. We are interested in how we can make it easier for organizations — especially those in the public sector — to learn, whether that’s standing up better data infrastructure and learning from it or figuring out what’s broken and trying something new.

Several years ago, the Pennsylvania Department of Corrections (PADOC) invited members of my team to work together with their staff, clinicians and residents in their custody to try approaches to solving the problems they were grappling with. They wanted to reduce violence and the use of solitary confinement and to improve the conditions of confinement as well as resident and officer health and well-being. We asked staff and residents to contribute their thoughts on how to make things better. We sat with them to mull over their ideas (with the PADOC general counsel at the table to quickly rule out strategies that were not permissible before they moved to the subsequent vetting and planning phase), and we advised the PADOC team on how they might pilot test these new approaches (with timelines ranging from only a few months to a few years). 

The more pilots that were launched, the more people — both staff and residents — felt confident that their ideas for improvement might turn to action. Even more ideas began to roll in. These were their ideas and their pilots, not ours. Some were stunningly inexpensive. Soothing sounds in living areas. Air plants and fish tanks in high-security units. Inspirational signage. Aromatherapy (liked the lavender, hated the cherry). Different-color bed linens (nice colors cost the same as grim ones, and almost nobody prefers grim). Everything was on the table, from trivial changes like these to much more substantial lifts such as crisis-intervention training, body-worn cameras and new programming curricula. No intervention was a silver bullet, but collectively, they ignited a culture of learning. And this culture has persisted.

PADOC had (and continues to have) leadership that enabled a learning environment, but there was otherwise no secret sauce that makes this approach to learning unique to them. Empowering people to solve their own problems has broad appeal. Over the last eight years, we have worked with systems in over three dozen states and across several domains (including policing, prosecution, corrections and reentry). No pilot was itself the end goal, but rather the process — whether changing living environments or programming — ingrained an expectation and then the normalization of joint problem-solving. 

Advocates have long argued that the conditions at Rikers Island are so bad that fixing them isn’t an option. I have visited jails and prisons across the United States and internationally over the last two decades. Rikers is indeed terrible. It’s hard to reconcile how a city with as much wealth, talent and creative problem-solving capabilities as New York could allow the conditions at Rikers Island to exist. New York City should — and often does — represent the best of America, yet its flagship correctional facility remains a stark symbol of a failed justice system and a city’s tolerance for subjecting vulnerable populations to deplorable conditions.

Rikers faces a classic chicken-and-egg dilemma in corrections: Poor conditions of confinement breed more violence, and the rise in violence shifts the focus away from improving these conditions to mitigating the immediate threats. This cycle perpetuates itself, leaving the underlying issues — underlying trauma, a lack of programming to keep people productively engaged and an environment that does not foster productive staff-resident interactions — unaddressed. The root causes of the violence are ignored, the poor conditions continue to fuel disruptive behavior, and escalating conflict leads to more aggressive responses from both residents and staff, resulting in higher levels of violence and stress. Demoralized and fatigued staff, meanwhile, struggle to manage the ongoing crises, and the result is a dangerous and toxic atmosphere.

Rikers is violent. The final quarter of the fiscal year (April through June 2024) saw 1,408 written infractions for fights and 1,688 instances of use of force (that’s a rate of about 27 per 100 people). And it’s not for lack of spending: Last year, the New York City Department of Correction spent $507,000 per person-year. That far exceeds the person-year cost of other large U.S. jurisdictions (more than five times Los Angeles County and Cook County (Chicago)) and even outstrips the cost of the Norwegian Correctional Service, which is widely regarded as the most humane in the world.

The City is legally mandated to close Rikers by August 2027, but delays with the borough-based jails intended to replace Rikers make meeting this deadline unlikely. The borough-based jails are intended to be safer, fairer and more humane, and most residents would be housed closer to family. Hopefully, this bodes well for New York City’s correctional populations of the future. But in the meantime, the refrain that Rikers is too bad to fix is cold comfort for the roughly 6,000 people incarcerated there now and the even larger number of people who work there — and for the loved ones of both. It is possible to improve conditions at Rikers for these people without impeding progress on the borough-based jails or sentencing reform, but jail administrators and advocates will need to allow these changes, and philanthropists and community members need to step up.

In our work on institutional conditions, we have found a strong willingness to engage in improvement efforts when residents, clinical and security staff, and families are invited to contribute ideas. It’s time for a Rikers in-custody think-do tank to kick-start action, like the effort in Pennsylvania. The think-do tank would focus on short-term initiatives to improve conditions until the borough jails open.

How would this work? City administrators would need to champion the approach and create the enabling environment for ideas to be curated, vetted, implemented and assessed (in an agile fashion). Trusted community partners and educational institutions can step in to set up the think-do tank, working with each other and with staff, residents and families. All participants in the think-do tank, including residents, should have equal standing.

Nothing productive can happen if frontline security staff aren’t engaged early and actively in processes that affect their work. Correctional institutions are hierarchical and, while directives can flow from above, it’s practically impossible to implement and sustain change unless staff believe in it. Staff engagement efforts can help ward off early resistance by raising awareness that improving living conditions also improves working conditions, including in matters as consequential as safety. During a field visit in India, a senior member of the Indian Police Service told me, “We hire for their heads but then treat them like hands.” If their input is requested and valued, staff will offer ideas for improvements.

Empowering staff and residents to look around themselves and ask, “What if we did something else instead?” — tied to a meaningful process for implementing those recommendations — is good for morale, something in short supply at Rikers. The think-do tank ideas that move to implementation would be vetted for security and have the support of residents, security and clinical staff, and families. Importantly, results from some of these pilots could generalize to the borough-based jails and beyond.

Collectively, we can decide what “good outcomes” are, but these would likely include reductions in incidents of violence, disciplinaries, complaints and improvements in quality of life for residents and staff. Not all ideas, even among the well-intended and well-thought-through ones, will yield good outcomes. New York City colleges and universities can be important learning partners. If Rikers invites them, they can unleash the talents of their faculty, researchers and students to assist with figuring out what is moving the dial in the right direction. Promising ideas can be expanded and optimized through iteration, while those less promising can quickly be retooled or stopped. 

We also need to unleash learning from the data already maintained at Rikers to support this work. A good place to start is to review data on rule violations and filed grievances, the system by which incarcerated people can file complaints; not only the neatly structured data on grievance type, but also the complex and messy open-form responses that can give granular insights into what is causing distress. Is it unfair practices, wait times for health care, vermin, the food? Grievance data can shine a light on which areas to prioritize.

In prior learning partnerships, we have tested changes to spaces, but also programming, systems of incentives and even technology. We have tested the use of virtual reality (as a training tool, as an ancillary treatment for substance use disorder and for recreation) and are seeing many novel applications of the technology. In a pilot in Nebraska, we tested what was at the time novel electronic cell-check technology, which digitally records a cell check when an officer is at the cell, and found it dramatically increased officer compliance with the cell-check policy. Absent the technology, officers were routinely skipping the checks, which, in a high-stakes environment, can have deadly consequences; Jeffrey Epstein’s suicide in Federal Bureau of Prisons custody in Manhattan is a noteworthy example. We have found that many new technologies can easily be tested and doing so is also a good check on vendor promises, helping to quickly distinguish those that work well from those that are useless or have serious flaws or unintended consequences.

Among my favorites of the many dozens of projects we were involved with in Pennsylvania was a small pilot in a women’s high-security mental health unit. The unit faced high levels of violent misconduct and together, the residents, clinicians and security staff designed an approach to reduce it. The residents involved coined it the “Chill Plan.” The plan shares many elements of a birth plan, where a birth mother creates a plan so she will know what to expect and that the health care professionals will follow her stated preferences. With the assistance of clinical staff, residents could file their own Chill Plans, which laid out how they wanted the facility staff to respond, and which they could invoke if they felt that they were escalating into a behavioral health crisis.

There are many hardworking people at Rikers and many well-established relationships with community-based organizations that work to help our most vulnerable, particularly in supporting reentry. Recognizing and leveraging these existing strengths can help efforts to address conditions within the facility. If invited, the community will step up. When we have asked architects and designers to assist with our work to improve conditions, they have been eager to contribute their talents. New York City is swimming in talent. Bringing together jail residents and their loved ones, correction officers, practitioners and researchers can unleash a host of ideas that can be quickly triaged and applied to improve conditions at Rikers for residents and staff alike. There’s still time.