Ruslan Grumble - Fotolia
A US state is pioneering technology that will ensure people do not languish for months or years in jail before the court tries their cases.
New Jersey is developing software that will automatically create risk profiles of people charged with offences, allowing judges to make informed decisions about whether those people are safe to release into the community.
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
The project, which is expected to be taken up by courts in other US states, will put an end to a system that has often released offenders based on their ability to pay bail, rather than their risk of absconding or re-offending.
The work is part of a major programme by New Jersey Courts to reform the state’s criminal justice system, backed by a $14m investment in IT.
The pre-trial assessment software is capable of searching multiple court and police databases to assess a suspect’s likelihood of absconding or failing to turn up to the court hearing.
It will reduce the time needed to produce risk profiles from hours or days to a matter of seconds, making it possible for judges to make fairer decisions about releasing suspects, said Jack McCarthy, CIO at New Jersey Courts.
“The fact that you can or cannot pay bail based on your financial situation has nothing to do with your likelihood to appear in court or your likelihood to get re-arrested. You may be a really bad guy but have all the money in the world and be able to get out,” he said in an interview with Computer Weekly.
New Jersey begins its judicial reform programme
But finding people with the programming skills to support software written using out-of-date programming was difficult, and the systems were becoming increasingly expensive to maintain.
“We wanted to take advantage of what modern technology has to offer, but also eliminate some long-term technology cost,” said McCarthy.
The IT team evaluated several IT suppliers before choosing software from Boston-based Pegasystems as a programming platform to develop its new IT systems.
Between 10 and 15 employees and external contractors worked on the project, which began in March 2015, rolling out a module every 3 months.
Automated risk assessment for offenders
New Jersey Courts worked with the Arnold Foundation, a US philanthropic organisation, to find a way of automating the foundation’s risk assessment process for offenders.
To conduct a risk assessment, a court officer would have needed to make manual database checks across a wide range of court and police databases, typically covering 40 million records.
The process could take four to five hours, or days in complex cases, said McCarthy, but the automated application is able to produce results in seconds.
The pre-trial assessment tool offers judges a recommendation for detention, release or release with a bracelet, but judges make the final decision.
With 80,000 risk assessments required in New Jersey a year, the app is expected to save hundreds of thousands of dollars a year, compared with completing the checks manually.
“The proud moment was when an individual from the Arnold Foundation saw it and said, ‘wow’. Then we knew we had something,” said McCarthy.
Speedy Trials application
A second application called Speedy Trials will monitor the time people who are remanded in custody spend in jail before a court hearing.
Until now, it has been possible for suspects, who may be innocent of the crime they are charged with, to spend years in jail before prosecutors bring their cases before the courts.
Under reforms coming into force in January 2017, prosecutors will be required to bring an indictment against a prisoner in 90 days, and move to trial 180 days later. If the case is late, a judge will release the prisoner, potentially with bail and conditions.
“If a prosecutor cannot move the case quickly enough to court, or the courts can’t hear it, our legislation is written so that that you will not be sitting in jail waiting for us. You will be able to go out on the street waiting for your trial,” said McCarthy.
The application, which will shortly go into user acceptance testing, will allow the courts to check whether prisoners are approaching the critical deadlines, enabling managers to prioritise cases.
Another priority for New Jersey Courts was to redevelop its compliant application, which is used by police officers to write warrants and court summons.
The project was a particular challenge, as the courts were in the process of introducing agencies and roles to support the judicial reform programme as the project was underway. “We needed a product that we could quickly change,” said McCarthy.
The team expected the application, which was built using the Java programming language on the Pega platform, to take nine months, but were able to complete it in five.
“We knew through the criminal justice reform that we had to modernise the application to make it extremely flexible,” said McCarthy.
How New Jersey managed its IT reforms
The project has allowed New Jersey Courts to standardise its IT applications on a single technology platform. It is expected to cut software development costs by a third and generate $5m in savings in IT development costs over time.
“As we took each of our legacy applications and re-wrote them using modern technologies, we expected to see our development time slashed by 30%, and we have achieved that,” said McCarthy.
New Jersey Courts worked with consultants from its IT team and Pegasystems to redevelop the court’s IT platforms in 6 phases, known as “slivers”.
The Pega consultants were able to integrate what McCarthy described as a “myriad of old technology systems” together. The Pega platform now acts as an “umbrella” linking diverse IT systems together.
Previously, for example, law enforcement officers used a separate application to create an arrest record and then re-key more than 50 data-fields into the court’s complaint system. Now they can send the data to the courts, and it is loaded at the push of a button into the court database.
“We have seen a lot of unexpected success. In certain cases, we thought we would have to go in and face a work around, but we have been happy with what we have seen,” said McCarthy.
New Jersey Court’s $14m IT project
- Early 2014 – New Jersey’s chief justice, Stuart Rabner, sets up a committee on criminal justice reform.
- March 2014 – The reform committee makes a series of 19 recommendations to improve criminal justice in New Jersey.
- August 2014 – The governor of New Jersey signed legislation to enact judicial reforms.
- March 2015 – New Jersey Courts begins work on implementing justice reform IT.
- 1 January 2015 – Original deadline for migrating to Pega was missed.
- March 2016 – Risk assessment tool completed, testing starts.
- 15 June 15th – Migration to Pega complete.
- 1 Jan 2017 – Automated risk assessment goes live in New Jersey Courts.
One of the biggest challenges in the project was to ensure there was one approach to charging suspects across the state.
Each organisation had its own charging and statute table, and each was slightly different, said McCarthy. Subtle differences, such as a comma or a dash in a different place, presented a major challenge.
New Jersey solved the problem by creating a master statute table – which codes each offence in the same way – and sharing that across every criminal justice system across the state.
It took a huge effort for attorneys to go through the data and decide what was a valid charge and what was not.
“It was a big undertaking in New Jersey to take all the systems, and get them just to use the same charging table. Everyone had their own,” said McCarthy.
“We didn’t know we were going to do that when we started the project. But once we identified this must be done, we got that done, as well as hit our target date.”
Court plans to put criminal justice systems on platform
With the new platform in place, court officials will be able to work with the IT team to make changes to court applications far more quickly than in the past.
McCarthy plans to move the state’s complete range of criminal justice systems on the platform, including the IT systems to manage domestic, juvenile, family and indictable cases.
“The knowledge we have gained through this risk assessment process, and what we have already built, will be shared very quickly into other areas of the court system,” he said, adding that the result will be huge savings across all areas.
“We wanted to take advantage of what modern technology has to offer, but also eliminate some long-term technology cost.”
Jack McCarthy, New Jersey Courts
The next project is to develop an application that will monitor people who have been released on bail, which will record whether or not they have met their bail conditions.
This will include records of whether offenders have regularly reported into court, complied with court orders or committed any further offences.
Caseworkers will be able to present the judge with a report from the system, which will show whether someone has made every appearance or not, said McCarthy.
The courts are also planning a mobile application for probation staff, which will allow them to file reports and download information they need without returning to the office.
“Our probation officers currently complete a very manual process if they do a site or home visit,” he said.
Experiments in the state of Georgia have shown that mobile apps have allowed probation officers to work more efficiently, reducing the number of probation officers needed.
Staff learning curve delayed project
New Jersey’s original plan, said McCarthy, was to train the court’s IT staff in Pega technology and to reduce the number of consultants on the project with each phase.
But it took longer for the court’s own employees to learn how to use Pega, delaying the overall project by six months.
“The biggest challenge with Pega was the learning curve. We expected our staff to pick it up and work with it, and it took us a few months longer than expected,” said McCarthy. “We thought it might take six months to a year. It took more like 15 to 18 months.”
The IT department, however, had huge success hiring younger staff members straight out of college and training them to use the Pega software, said McCarthy.
He is aware that many of them are likely to leave for higher-paid jobs in the private sector after a few years, given that public sector pensions are not what they used to be.
“If I can bring in bright individuals and use them for three to four years, then they leave and go and make their millions, then great. I got a couple of good years out of them,” he said.
Customisation made work more difficult
Inevitably, in a project of this scale, there are a few lessons to be learned.
McCarthy said he would take a more ruthless approach to avoiding customisation if he worked on a similar project in the future.
“We allowed business to dictate customisation outside of the ‘guard rails’, without realising the severe penalty that you pay as you migrate to new versions [of Pega],” he said.
At that point, the courts had only developed one application, so the impact was less than it could have been, said McCarthy.
“By learning that lesson early, we started painting in the lines for any future development efforts, and that was really helpful,” he said.
He advised anyone embarking on a similar project to think carefully about how many software licences they will need, right at the start of the project.
Originally, New Jersey Courts bought 3,000 licences, but as the project progressed it ended up buying more.
“We purchased in several buckets. We probably could have got a better deal had we bought earlier,” he said.
Extending risk assessment to other states
New Jersey plans to extend its pre-trial assessment application to pull in data from neighbouring states – which are in the process of re-writing their databases – to standardise on XML.
Currently, the pre-trial application provides court workers with a case number, showing that data exists in another state.
Over the next four months, McCarthy plans to modify the system that will automatically download PDF reports of the out-of-state data.
McCarthy hopes to persuade New York, Pennsylvania, Connecticut and Delaware to adapt their charging tables to a master statute table that will allow them to share their records with New Jersey.
In the longer term, New Jersey is hoping to integrate its criminal justice system with data from other states.
“The integration of out-of-state charges has always been seen as a risk. If I can give them to you, and a human being can read them and put them in, that will be our first step. But if a human being can read them, decipher them and put them in, we should be able to do it with our software,” said McCarthy.
Project will make major difference to fairness of criminal justice
The project will lead to a sea change in criminal justice, said McCarthy, and is likely to lead to sentencing reforms. Releasing more people on bail will mean that people will not be under pressure to plead guilty just to get out of jail.
“An individual may spend six months sitting in jail and then plead guilty just to get out and do time served. If that individual is already out in the street, they are much less likely to take a term of incarceration,” he said.
The Speedy Trial project in particular could place a real strain on the court system in New Jersey.
“We know that we have a huge problem coming up if every individual goes to trial. We don’t have enough judges to try all the cases,” said McCarthy.
There will need to be some major changes, such as allowing more plea negotiations, if the court system is to continue functioning.
Read more about business process management
- Turkish bank ING uses business process management technology to offer financial services through 1,500 grocery stores, technology outlets and post offices.
- In this expert response, learn how business process management differs from ERP systems.
- The importance of business process management goes well beyond streamlining operations. Now, CIOs explain, they’re converting BPM to business value.
- Electronic giant Philips predicts a revolution in home healthcare thanks to internet of things and wearable technology.