The frames of his glasses sat mangled next to him as he nodded in a trance. He had smashed the lenses against the floor with the palm of his hand. A crude and incomplete message to his family carved superficially into his forearms left his clothes smeared with blood. He then attempted to let the blood from his wrists. When that didn’t kill him, he wept dryly and then swallowed the shards of glass, leaving a sharp pain in his chest. A small packet of crystals he had found tucked under a blanket some time earlier provided momentary relief from his hunger and a boost of energy that was now gone. Drinking his own urine was unpleasant, he had decided, but also gave some relief to his dusty body -- an eventual husk, left to dry out in a closet like wildflowers.
During his five-day stay at the U.S. Drug Enforcement Administration (DEA) field office in Kearny Mesa, San Diego, Daniel Chong lost 15 pounds. They forgot about him.
Chong’s case is unusual. The DEA raided an apartment where Chong and eight of his friends were having a party in the University City neighborhood of San Diego on April 21, 2012. There, the DEA found marijuana, magic mushrooms, 18,000 ecstasy pills, firearms and ammunition. After being interrogated at the DEA field office, officers decided that Chong had been in the wrong place at the wrong time, and that he was to be released. Seven of Chong’s friends were taken to county jail, one was released, and Chong was left in his holding cell. In 2013, the DEA settled Chong's lawsuit for $4.1 million.
Chong’s case is unusual because it’s extreme, but administrative mistakes happen regularly in the justice system. In any system where humans are responsible for data entry, mistakes are bound to happen, and with a national inmate population exceeding 2.3 million, mistakes happen in U.S. jails and prisons every day.
Rodriguez Purnell, a suspect in a first-degree murder case in Maryland, was accidentally released in October 2014. Malik Erkins, who was imprisoned for breaking into cars in Cook County, Ill., was accidentally detained an extra five months beyond his scheduled release in December 2014. Also in 2014, a female inmate in Marion County, Ind., was placed in the same cell with a male inmate, whom the female inmate later accused of sexually assaulting her.
Cases like these appear in the news regularly because they’re titillating. They sound like premises for TV shows or movies. Mistakes surrounding incarceration upset the public’s sense that there’s a Big Brother who can account for everything, for better or worse. Most people assume that even if the U.S. justice system isn’t necessarily fair all the time, that most of the people in prison are there for a reason and that upon the hour of their scheduled release, a well-tuned grandfather clock will chime. The cases that appear in the news are just the ones the public hears about, but just how common these clerical errors are is a mystery, said Sharon Dolovich, a law professor at the University of California, Los Angeles (UCLA).
“It’s impossible to know,” Dolovich said. “The problem is that most people in custody are indigent and have nobody on the outside advocating for them. It’s not as if there are watchdogs on the system every day saying, ‘You have to release these people.’ … It’s a total black box.”
The United States blows away the rest of the world when it comes to incarceration. For every 100,000 residents, the U.S. has 707 prisoners, according to a 2014 report by the National Research Council. Russia comes in second with 474 per 100,000. For low-income Americans, and especially for low-income blacks, Hispanics and Native Americans, the prospect of prison is not far-fetched. That one might turn incarceration into a career seems to many Americans a reasonably likely scenario. No one acts surprised when they hear about a young black man headed to prison in America. Most of the country has decided it’s not a big deal.
And that, Dolovich said, is a dangerous attitude.
“I think the thing to remember is the undertaking of removing somebody from the community and locking them up is a really serious one with major implications, and we can’t do it in a cavalier way,” Dolovich said. “And to the extent that we’re doing it, to a degree and on a scale that makes these kinds of mistakes eminently possible should really give us pause.”
One of the ways prisons and jails are attempting to keep better track of their populations and to run their facilities more efficiently is by upgrading to modern software systems. There are dozens of companies in the prison and jail management software marketplace promising a broad range of functionalities and results. Tribridge is one of those companies.
Tribridge's jail and inmate management system, called Offender360, provides state departments of correction and regional jails a tool to manage their inmates and internal processes more thoroughly. Things like bed assignments, movements, sentencing, good behavior, grievance process, medical issues and nature of offense are all neatly catalogued in the system for each inmate. The system also manages inmates who have been released and those who have yet to enter the facility, said Josh Jaquish, the company's vice president of public sector and industry.
“It manages all aspects from the time an inmate is placed under arrest until the time they’re released back into society and probation and parole after that,” Jaquish said. “It manages the intake process, so while that seems pretty simplistic, it’s a very important process. From making sure you’re doing an assessment of the offenders correctly, that you’re classifying them correctly.”
The software allows for administrators who don’t want nonviolent, first-time offenders to be lumped in, kept longer than necessary and possibly institutionalized, he explained.
Many institutions, Jaquish said, rely on computer systems installed in the 1980s that haven’t kept up with new laws and that can’t meet the needs of prison administrators. Finding efficiency through data-driven decision-making is very difficult for those with outdated technology. These days, he said, tracking metrics like use of force provides opportunities for external groups like the executive branch, law enforcement or the legislature to examine how their institutions are working. As high-profile cases of police brutality become more common in the news, governments become more interested in understanding data that might unlock answers to questions the public is asking.
“The people who are left to answer those questions can’t,” Jaquith said. “And they certainly can’t in a timely manner. What we’ve seen over the last three or four years is this massive technology upgrade in the justice and public safety sector. Part of it is just keeping up the commercialization of IT overall. They’re looking for a solution that can not only meet their current needs, but also give them a platform so that 10 years from now, they’re not stuck in the same position they are today.”
Illinois' Cook County Jail, the second-largest jail in the U.S., home to 100,000 inmates annually, upgraded to Offender360 this year. Before the upgrade, the jail had been using a nonenterprise system that was installed in 2009. Robert McInerney, CIO of the Cook County Sheriff’s Office, said the system was crashing frequently, it wasn’t secure, and retrieving data required contacting the vendor. Upgrading to Offender360 has helped them in several ways, he said, one of the biggest being that they now own their data.
“Our business intelligence team we stood up here at the sheriff’s office is actually one of the fastest-growing departments we have because we’re able to get data and match it with other data we have in our environment and show some unique capabilities, like where inmates are when they’re being released, where are they going, what services are around in those areas,” McInerney said. “We can use it on the police side, as well as for information on how people were when they were in jail and if they show up again. It’s based on Microsoft CRM platform, which we can find a lot of uses to develop, as well as it uses products like SSRS, which is Microsoft’s reporting services, natively. So we can do a lot of ad-hoc reporting.”
Personalizing inmate treatment is one of the common themes found around the current generation of prison and jail management software. The idea is that if software can make tracking inmates easier, then they’ll be treated more appropriately based on their criminal history and risk factors, and so be less likely to offend again.
“One of the biggest things on the sheriff’s mind is mental health,” McInerney said. “We have a lot of data here that we could never get to before. For example, if I can now know that you were in a jail and somebody came to visit you every week and they stayed for an hour, and for the whole four months you were here they were religious about coming in, well, I now know something about your support structure outside of here. So, when you’re released and I’ve got some instructions or services, I might be able to send that information to your support structure as well, and not just give it to you as you walk out.”
Cook County Jail plans to begin a program that gives inmates a personalized packet of information based on their unique histories, their living situation and home life, McInerney said.
“I want to be able to get it personalized so that person has something that can actually help them and, again, send it to other people in their support structure so that they actually adhere to it, and what is that impact on our recidivism? It’s got to be better,” he said. “It’s got to be reducing it. I just don’t have a percentage of how well we do.”
Reducing recidivism is the ultimate goal of many in the justice system, in academia and in social activism. Keeping people out of prison and in the workforce is one of the key indicators of a healthy society, and from a purely administrative standpoint, fewer inmates means fewer tax dollars wasted. Rehabilitating criminals is difficult everywhere, but especially in jails, where turnover is frequent. In Cook County Jail, the average stay is about 45 days, by McInerney’s accounting. Nonprofit Community Resource Services estimates that the average jail releases more than 75 percent of its inmates within 72 hours, which isn’t long enough to remedy athlete’s foot, let alone address serious behavioral problems.
Keeping track of prisoners would be easier if there weren’t as many of them. What works in reducing recidivism isn’t a settled matter, but there are some generally accepted tenets. In a paper called What Works in Reducing Recidivism? (PDF), published in 2006 by Edward J. Latessa, professor and director of the School of Criminal Justice at the University of Cincinnati, three key conclusions are reached. Through the analysis of two studies involving more than 26,000 offenders and more than 100 residential and nonresidential correctional programs, Latessa found that certain programs, if implemented correctly, tend to reduce recidivism. Others, however, are shown to either not to work at all, or if applied incorrectly, increase criminal activity among offenders.
First, Latessa wrote, programs should only target offenders who have a high probability of offending again. Putting low-risk inmates into programs alongside high-risk inmates can sometimes provide an unintended “education” in antisocial behavior. Compared to a control group, there was a 4 percent increase in recidivism among low-risk inmates placed in programs, with some programs showing increases as great as 36 percent. High-risk offenders placed in programs, by contrast, averaged an 8 percent lower recidivism rate when compared with a control group. The most effective programs for high-risk offenders saw recidivism reductions as great as 34 percent.
Second, Latessa found that just as important as who to treat is what to treat. Through scientific study, certain personality traits have been linked with crime, and they include things like antisocial peer associations, antisocial personality traits, substance abuse, lack of problem-solving skills and lack of self-control. Programs that effectively reduce recidivism address those traits. Addressing non-crime-producing factors, like self-esteem, physical condition, job skills, creativity, and understanding one’s culture or history, do not tend to reduce recidivism.
Latessa’s third conclusion is that the efficacy of a program can be linked to the nature of treatment. The most effective programs, he wrote, address the behavioral problems inmates face in their immediate circumstances using action-oriented treatment. Focusing on past events through talk-therapy or grief counseling does not reduce recidivism, he found, while teaching inmates pro-social skills to replace their antisocial skills through behavior modification techniques tends to work. This can be achieved through cognitive behavioral therapy and similar disciplines.
Scientists, like Latessa, have demonstrated a method that prevents people who were once criminals from continuing to commit crimes, or at least as far as the science is concerned, found a method that prevents criminals from getting caught again. Categorizing inmates by relevant risk factors and then assigning personalized rehabilitation treatment that has been scientifically proven to work could be facilitated though the use of current generation corrections management software. It could greatly reduce the number of people incarcerated in America. Unfortunately, said Faye Taxman, professor of criminology at George Mason University, we do very little of any of that.
“There are a lot of good programs that are effective,” Taxman said. “They’re infrequently implemented. … It’s our culture. We as a culture think people who are involved in the justice system are second-class citizens, and we ask people to be treated poorly who are involved in the criminal justice system.”
There’s a lot of good software can do, but it’s just as prone to human error or neglect as any other system, Taxman said – ultimately, if American culture does not leave anyone in power inclined to help those who need help most, software won’t do anything.
“To me, the more important part is how an organization reinforces certain values,” Taxman said. “And that value is that people are citizens and they have certain rights, even if they’re incarcerated. … Most jails and prisons already have a rap sheet of someone. They can look at someone’s history. They can look at the nature of the offense, but they don’t, because they’re managing large numbers of people and the person isn’t always sitting there. If you’re a [corrections officer] and you take in 200 people a day, how are you going to remember?”
The American corrections system is a human mill. More people go in every year. Corrections spending has increased from $10 billion to $50 billion in the past 20 years. Across party lines, the American public says that’s too much. According to a 2012 survey by Public Opinion Strategies and the Mellman Group, 84 percent of Americans agree that it would be a good idea to shift some of the money used to incarcerate low-risk, nonviolent inmates toward community corrections programs like probation or parole. The public doesn’t want the prisons filled. It’s expensive and there are ways of rehabilitating prisoners that work. But people don’t do what works, Taxman said – they do what aligns with their values.
“It’s sort of like, we know what works for diabetics is people getting insulin and working on exercise and eating programs, but we don’t provide eating programs. We give people medication, but we don’t pay for people to go to nutritionists or to go to exercise programs. If we had really effective treatment, you would deal with the whole person," she said. "The same thing happens in criminal justice. Software can’t be a cure-all. It’s about the people we hire and what we expect people to do and that’s where most of these systems fall down.”
Colin wrote for Government Technology from 2010 through most of 2016.