Turning the Gears of Justice: A Conversation with Chidinma Ume and Erin Pilnyak

From 2013 to 2020, the New York City Mayor’s Office of Criminal Justice (MOCJ) implemented an array of strategies to reduce incarceration while maintaining safety. The average daily population dropped from 11,696 in 2013 to a low of 3.820 in April 2020, even as index crimes declined by 9.6%. Chidinma Ume and Erin Pilnyak, who had both previously served as state prosecutors, were leaders of a part of MOCJ dedicated to improving justice operations. Today, Ume is senior director of community justice at the Center for Justice Innovation. Pilnyak is now the chief operating officer at IntegrAssure, which is currently serving as the deputy consent decree monitor for the city of Aurora, Colorado, and deputy independent police auditor for Petaluma and San Leandro, California.

While in the mayor’s office, they led, successively, a multi-agency effort to reduce case delay among detained people in New York City. Dubbed Justice Reboot, the initiative joined the multiple decision-makers across both state and city agencies to look at data together and craft the fixes in operations that could shorten the length of time incarcerated people were in jail. The effort was chaired by the chief judge of the State of New York and the director of MOCJ (then Elizabeth Glazer, Vital City’s founder). It brought together teams of prosecutors, defenders, court administrators, the Department of Correction and others into working groups headed in each borough by that county’s chief administrative judge. The mayor’s office developed data for all the participants to look at and functioned as the overall convener of the initiative.

This conversation has been edited for length and clarity.

Vital City: Each of you worked across the city in trying to bring down the rate of case delay. You got deep in the details of what makes the justice system go or not go. Would you give a sense of what this initiative was and how it operated? 

Erin Pilnyak: I started my role as an executive director of justice operations in the Mayor’s Office of Criminal Justice, and on my team, I had a former indigent defense provider and policy associates and former prosecutors. This allowed us to be a little incubator of ideas, because we approached every issue that was given to us with diverse perspectives. My approach to what the potential cause of a delay was might be opposite of my teammates’. Everyone was looking at the same data and it was a way to get people literally on the same page. Because, as you can imagine, that’s not natural when you have the defense bar, prosecutors, courts, corrections, parole and probation who are not necessarily used to working together.

Evidence can provide a way for everyone to move from blaming to problem-solving.

VC: Chidinma, who were the partners and how did this work? 

Chidinma Ume: We worked with the courts, the defenders and the prosecutors, court administrative staff, the chief judge — so Judge Jonathan Lippman at the time. The chief administrative judge, Judge Larry Marks, along with all of the administrative judges in each of the five boroughs, worked with us as well. We had all five prosecutor offices; from each office, their first assistant was at the table. The defenders, each office had their chief of criminal in the mix. Then Probation, the Department of Correction and us. And then the Center for Court Innovation, now called the Center for Justice Innovation. 

First, we needed data to be able to understand what was going on with the cases that were in the system, who the people were, what types of cases they were charged with, how long they had been in and what could we learn qualitatively about the reasons for them staying as long as they did

The first thing that helped us with jail reduction efforts was to have a common goal. What Justice Reboot did was say, “We want to reduce this jail population.” Having a common goal helped us align our jail-reduction efforts: reduce half of the long stayers in a year, and then after that, come up with systemic reforms that would prevent incoming cases from aging as much as they did. 

VC: How did you know what steps to take? And how did you ensure that people would take those steps? 

CU: We started by cutting down the backlog. The plan was that, in each borough, each administrative judge would look at a list of all the cases that were a year or older and send those to a particular courtroom to be resolved or go to trial. Then it was really about figuring out how we would continue to work together to understand what was currently causing delay and what we would do in each borough that would reveal systemic solutions.

From June of 2015 to June of 2017, we figured out what to do through a series of conversations — as a collective and in smaller groups, with each party. We needed to meet with everyone one-on-one to give a clearer picture of the data and understand the contributors of delay. We had to do this on many levels: we had our citywide meetings as a large group of over 30 people, and we also met with each stakeholder group separately — so prosecutors, defenders, DOC [New York City Department of Correction], the courts, etc. We did this because we had to understand their unique vantage points and where we could find efficiencies without compromising the process of case disposition. We had meetings by borough as well, because we learned that each borough was different — they had different paces at which they processed cases, different physical space needs given the difference in the courthouse facilities, and they had different volumes of cases as well. We also got qualitative data, through observing courts around the city and reviewing the city’s oldest cases.

VC: Were the kinds of facts that you were bringing forward surprising to the partners?

CU: Definitely. The idea of thinking about their case processing in terms of length of how long people were staying on Rikers Island was eye opening. I remember partners wanting to know, “How much is our caseload contributing to the Rikers population? How fast are we processing cases, and how do we fare as compared to the rest of the city?” It was extremely helpful for the Justice Reboot dataset to integrate court case processing data with the jail population data, so everyone could see how case processing was affecting the jail population. And to have this information for the whole city and be able to break it down by borough, the partners largely found it illuminating

VC: One of the things that is sometimes mentioned about why reducing length of stay is such a bear is, some observers will say, “Well, prosecutors want to delay, because they think that more incarceration may get them a better plea. And defenders want to delay, because they’re hoping witnesses will forget, evidence will get stale and they’ll get a better deal.” Did you find that there was resistance to this goal of reducing lengths of stay? And if so, what do you do about that?

CU: Not necessarily resistance, but we did have to align everyone around what Justice Reboot was — and what it wasn’t. We were trying to reduce unnecessary incarceration and avoidable case delay. It wasn’t about speeding up convictions or rushing people to trial. Once partners better understood that, we were able to get more buy-in around doing things to make the system more efficient in a thoughtful way. The efficiency approach resonated more with all the partners, because everyone wanted their efforts to translate into more productivity and more timely case resolutions. It also meant we could discuss process fixes that saved them time and energy, like better and faster communication and coordination.

Our discussions about efficiency also had to be very nuanced, because the partners were all approaching this work from different vantage points. Delay affected their work in different ways. For instance, from the prosecution perspective, we shared a study that showed that the longer cases linger, the less favorable the outcomes are for the prosecution. With defenders, waiting in jail was harmful to their clients. They were extremely concerned by this. And judges had to be concerned with the right to a speedy trial and overall case management. Their dockets had to move in a timely manner. 

Part of the benefit of meeting all together through this project was that partners could better understand these varied perspectives and find solutions that respected everyone’s main concerns. We focused on the process points where delay benefited no one. The data helped us really map this out. For instance, once someone is indicted for a felony, they’re waiting for the next step which, in New York City, is the Supreme Court arraignment. On average, this was a 30-day period, mostly for administrative — or calendaring — purposes. In some cases, people could be waiting 90 days for their first court appearance after indictment, which again was just bad for people, for their cases and for the courts. The partners agreed that this was an area where we could cut down on delay. 

These types of consensus moments proved that, even within an adversarial system, we could work together collaboratively.

Court can’t happen if the judge doesn’t know where the defendant is

VC: Those process points were also things as simple as the judge knowing that a bus had arrived from Rikers. 

CU: Absolutely, because that information was passed by paper, and it could get stuck downstairs in the basement of the courthouse where people were held before they would be escorted upstairs to the courtroom. If those papers didn’t make it upstairs to the court on time, or if the person at the desk downstairs didn’t call the clerk and affirmatively let them know that someone had arrived, there could be these miscommunications where people were in the building but not brought upstairs to the court simply because a phone call may not have been made or a paper wasn’t run upstairs. 

VC: Was this part of a larger technological problem?

EP: Absolutely. We evolved when email communication between the Office of Court Administration (OCA) and DOC was introduced, because that network did not exist at DOC at that time for that to even be considered at the beginning of this discussion. 

And the second part was really digging into why defense attorneys were waiving their clients’ appearances and why that couldn’t be communicated earlier.

So it was really working closely with the institutional indigent defense providers and 18B panel [lawyers from the private bar appointed to defend people who cannot afford counsel] to give us an update as to whether the attorney is already engaged in a trial or a pretrial hearing and to let DOC know 24 hours beforehand, so they don’t have to produce this defendant who otherwise is just going to be woken up at 5:00 a.m., brought over to court without appearing and then not return until 7:00 p.m.

Increasing that level of communication had a tremendous effect and it also required the final approval of the judges for full transparency. 

The third part is increasing the touch point for defense attorneys to speak to their clients. What we heard from the defense bar was that getting to Rikers is challenging. So how could we then streamline that process for them?

One member of our team who had deep experience as an indigent defense provider suggested we as a city fund free transportation and then streamline the security protocols once that bus full of people gets to Rikers, so they’re not in line with everybody else going through multiple hours of waiting and delays, and allowing that process to be far, far more efficient. 

And that had a huge impact. And then using videoconferencing and getting it expanded to attorneys, and especially to indigent defense attorneys. 

So there are corrections officers who can facilitate escorting the defendants down to speak to their attorneys, to allow these additional slots to be utilized, and now there are meaningful contacts happening in between court dates. These kinds of very simple, minor changes had a huge impact in addressing the “not-seen” population, those who were produced by DOC but not seen by a judge. And really at the core of all of this is just increased communication and making sure that the right parties are communicating.

Fixing broken toilets: Daily impediments to appearing in court

VC: What percentage of defendants in your study had not been seen by a judge?

EP: It was more than half the people brought to court from Rikers each day! In our not-seen analysis, we identified a specific court part in Manhattan that had a larger percentage of not-seen defendants than other court parts. We spoke to OCA, we spoke to DOC. Ultimately the problem was an inoperable toilet in the holding cell behind the court part [which meant that defendants who were due in court had no place to be held before their case was called, and as a result, court appearances were canceled].

So then there was negotiating between DOC and OCA due to the location of the toilet [because depending on where the toilet was, it would either be on DOC’s tab or on OCA’s]: Who is going to pay for the plumber to come and fix this toilet so it doesn’t flood all the time?

Something that simple made that court part inoperable for the day. Everybody knew about this toilet as a problem, but nobody elevated it.

CU: Coordination is central. Everybody who dealt with the issue knew about it, but they just didn’t know how to fix it or who should fix it.

EP: Part of the mediation for the toilet problem was literally myself and another team member; we had the blueprint for that building to figure out who should be responsible for this toilet. Really, that’s what it was. 

CU: A lot of Reboot was getting cases ready for the next step in the process to happen. For example, there were some cases that were ready to go to trial, but the accused people only had the jail-issued jumpsuits — which they could not wear for trial, because it would be clear that they were in custody. That presents constitutional problems.

One of our team members found it so interesting to learn that jumpsuits could be the reason why people have to spend another month or more in jail. That inspired an idea: “What if there was a set attire, like in school? Why don’t you just have some clothing can even be worn in public?” And she suggested a white top and black pants, that was it.

Credit to DOC. The chief at the time said, “Okay, we can do that.” It took maybe two months for them to order the clothes and then get them in circulation. And then people started using their attire.

During one week, we spent a different day each week in each of the five courthouses. We saw the trial attire being used in the courthouses. And we said, “It’s working.” This seems to have solved that problem. And it makes you think, if there weren’t these trial clothes, these are people who would still be on the island waiting for a replacement for their jumpsuits to magically appear.

Addressing delay in the cases of people with mental illness

VC: One of the other projects you led related to reducing the length of stay for people suffering from mental illness.

EP: Every person who is facing trial has to be mentally competent to stand trial. And under New York State Law, if that competence is in doubt, a physician conducts an exam to declare the person to be either fit or unfit to stand trial. When we were looking at our long stayers, the big bulk of that population unfortunately had pending 730 issues, meaning many of the long-stayers were waiting for or were in the midst of a 730 adjudication [to determine their competence to stand trial] and/or they were repeating the cycle of being found unfit, then restored, then returned to Rikers and found unfit again.

Fixing this required working in partnership with our core partners, but also bringing in New York City Health + Hospitals’ Correctional Health Services and others. And really trying to understand what happens from A to Z when a defense attorney, upon their assessment of their client, requests a 730 exam. And then what happens from that point until there is an exam produced to the court and then the judge deems an individual fit or unfit to proceed. 

VC: The problem that you were trying to solve was that when people who had been found unfit, it turned out that they stayed for longer periods, significantly longer periods, and it was unclear why?

EP: We wanted to shorten the processing time, not just for folks who were found unfit, but also just for people who were being requested for a 730 exam. Because for that latter population, just to come to a resolution was taking an average 43 days. And what we were trying to aim to do with the pilot was to shorten that time to seven days for misdemeanors and 14 days for felony cases. For the population already deemed unfit, we had to understand why there was such delay in finding them an appropriate place to go to the state hospital system, and this was especially impactful for misdemeanor cases because if found unfit, they would be sent to the hospital and not be returned to custody because the case will be dismissed.

Ultimately, the delays were because there was paper being traded between the city and the state, and because much of this was information protected under the federal Health Insurance Portability and Accountability Act (HIPAA). So one innovation was having a uniform HIPAA waiver for all New York City public hospitals. Because prior to this implementation, the defense attorney and their client had to fill out a different HIPAA waiver for every single hospital.

So that was a huge change. But that obviously took a lot of discussion with health providers and their legal counsel to really understand what appropriate boundaries are put in and how this can be done with sufficient safeguards in place to protect HIPAA privacy rights.

Then there was a state-versus-city challenge. Once someone is found unfit, they have to be transported to the state hospital system. But all of these communications were happening by paper. So some records could be sitting in some property clerk’s office for weeks on end without there being a registration that this information was received, or that the state hospital should be expecting this information. 

The other piece of this was implementing electronic transmission of medical records, which was also a very thorny issue. Again, just kind of making sure that both legal counterparts felt comfortable with how the information was being transmitted and how it was being safeguarded on both ends. Technology is great, but there needs to be communication between people in the system, saying to one another, “This is how we should be preparing together to make sure that this is a streamlined process.”

VC: Talk about other city-state coordination you had to unstick.

EP: When a person receives a state sentence from the courthouse, the city DOC has to transport that person to the state prison system to serve that sentence. There are a variety of different reasons why the state may reject that inmate, such as lack of some documentation that was not in the initial batch sent from DOC. But that rejection, what that ultimately did was cause a huge delay and backlog of state-ready population that was not being transported in an efficient way. So the DOC came to us and said, “We are experiencing delays in getting this requested information from OCA.” A lot of it was case-related information that DOC did not have.

We mapped it out for OCA and DOC and the state, “Here is the information flowing from point A, this is how ultimately one gets to the state.” 

That in itself reduced that state-ready population significantly.

VC: There are these repeating themes: You have to have a trusted convener who doesn’t have a stake or certainly is not perceived as taking one side or another. It really helps to have data that no one has seen before, that everybody is looking at the same data. Then just understanding the nuts and bolts and the operations. 

CU: I just want to add a few more ingredients. You took the words out of my mouth with the neutral convener: someone who is interested in protecting the virtues that are fundamental to the process, but who’s not attached to that process staying the same.

I’d also add a combination of patience and urgency. The urgency is: we have this common goal. We still need to hit a certain population by a particular date. I remember that in those early days, people weren’t shy about letting us know that we added a slew of meetings to their calendars, that we were sending them all this data to review, and dashboards and performance indicators often. But because we had that urgency—“No, we have to get this done”—it motivated us to stick to the process. And all along the way, as we maintained, the group saw the value of it themselves.

The other side of that is patience, because every issue we came across had about four or five sub-issues. 

So we had to have the patience to say, “We’re going to fix these issues” even when they reveal more steps to understand and address. So, we had to then say, “Oh wait, we have to take a step back. We can’t do that until we look at the whole process.” 

EP: I think where we really excelled as a team was the constant relationship-building and affirming that relationship-building, because it really is just about people to people. Because we can all have high hopes and high goals and get there, but we’re really asking people, on a daily basis, to come out of their comfort zone and do something extra that they probably don’t perceive as part of their job that they’re being paid to do. So we’re asking them to be aspirational, to come out and do extra work. 

And that also meant that we, as a team, were available to all of our partners 24/7. So when there were issues, we were on the call; it didn’t matter if it was 1:00 a.m., 2:00 a.m., on the weekends. If there were bail payment problems, we were emailing parole.

I remember one specific weekend when we were emailing parole from 2:00 a.m. on a Saturday to like 5:00, 6:00 a.m. because there were issues with paperwork and making sure the DOC’s paperwork was getting to the right person at DOC to make sure that this person was going to be able to be discharged.

VC: So here’s the $64,000 question for you. What do you think it takes for changes to stick from year to year, from administration to administration?

EP: After working in the Mayor’s Office of Criminal Justice, I went to the NYPD as an assistant deputy commissioner in the Risk Management Bureau, and we were the primary liaison for the federal monitorship for the NYPD. I think what I learned, kind of seeing how the largest law enforcement agency in the world operates, is that it’s not enough to say a policy has changed. What it really takes for culture change is to make sure that internal stakeholders understand the goal of this change.

A lot of reforms that happen are coming from an external agency to the agency. It’s a lot of times being mandated upon. And unless that communication internally is done with thoughtfulness to explain why certain things are being done and are not just detrimental and negative, the change will not stick.

VC: Chidinma, your thoughts? 

CU: I think first is memorializing the change. I remember when we changed those court case milestones for the indictment to [state] supreme court arraignment, and then supreme court arraignment to first appearance after that, and then conviction to sentencing needed to adhere to the 14 day period, we worked with the partners to announce these changes in writing. We worked together on a memo for that, and the chief administrative judge circulated it, as did the project partners, so it had some weight.

Then second is measuring progress. And specifically, I think measuring that performance from the lens of people, the actual doers. So if we’re talking about changes that judges need to make on the ground or that prosecutors need to make on the ground, they have to come to view it as “Me doing this thing is tied to my success.” It becomes an incentive for them. We used to send out these monthly dashboard reports, which visualized the data about how things were going in each borough.

VC: Sometimes it really comes down to just bread-and-butter issues. It would be great if you’re an extraordinary leader, et cetera, but let’s just make sure that you’re able to do these three things, because that’s your job.

CU: I also think there’s value in tightening up the bread-and-butter issues, because you show that you’re credible. Coming in and changing, saying the DOC is now going to transport people on spaceships, that is so out of touch. It’s so hard to get people behind something that is so far from their daily reality. But if you’re fixing these actual process changes, that makes sense to people. The bread-and-butter issues to me are what gives credibility with the changes we were trying to make.

VC: Such a rich conversation. Any last thoughts?

CU: Yes. Resist the urge to start with technology as a fix.

I remember someone saying, “Oh wow, so you were just making the trains run faster? Why can’t you just automate this?” I know we’re in the technology age, and a lot of folks think you can just kind of jump your way to tech. But a lot of this is all about how you keep the humanity of the person while you solve these problems, so I would do people. People first.

EP: I think piggybacking off of that is my takeaway from my collective experience: There isn’t just one correct answer. Having an open mind and willingness to listen to those who are on the ground, to really understand what is happening and why certain things are happening is crucial.

Then number two, I think there’s always some tendency to try to implement changes all at once. And what that really does is produce unintended consequences. So sometimes while fixing it from top to bottom is necessary, sometimes doing that in increments is helpful as well. Because with the first step, if it goes wrong, you always have naysayers in the room that’ll tell you, “See, this is why this is never going to work.”