It started with a sample jail roster in January 2017 and a single question: Who’s in jail?
The roster wasadaily report generatedfrom the widely used Offender Management System software. The system is used in 340 correctional facilities across 40 states.
Revolving Door:How inmates end up back in jail
With theassistance of Middle River Regional Jail’s information technology officer Wyatt Brown, we were able to craft a request for a custom report to tell us more than the standard daily roster does.
That roster gave us names of those inmates for a day — May 9, 2017 — and what charges they were being held on, mugshots for each inmate, as well as other data including bond information and dates of entry and expected release.
Brown advised early on that the jail data might create more questions than it could answer.
In a March 1, 2017 email, he wrote, “A great deal of our 850-some guests are both serving time and awaiting new charges in multiple jurisdictions or the same one. If you want to get to the nitty-gritty on each one, it will require a review of each inmate’s case individually and compiling that info.”
Boy, was he right.
Seeing double — and triple — what we thought we'd see for re-offenders
We started by manually creating a spreadsheet that matched inmate names with the specific charges they faced. Inmates held on that day faced over 2,300 individual charges. Those charges we filtered by actual charge type, say PRB4850F6, a charge of felony probation violation. (There’s also a misdemeanor probation violation charge, PRB4851M9, and the somewhat curious PRB4839S9, "probation violation - type not clear.")
More:Explaining recidivism rates in Virginia, why the conversation around them is limited
For that day alone, wecharted over 300 specific charge types. For some offenses, like weapons charges, there were multiple separate charge types: 18 different types of weapons charges, 19 types of assault charges, 23 types of DWI charges, 26 types of fraud charges, 28 types of larceny charges, 39 types of narcotics charges, and … well, you get the picture.
Some of these charge types indicate that the person being charged has offended before. For example, NAR3087F9 is a charge for possession of drugs with intent to distribute, a “3rd or subsequent” time.
When westarted counting charges, it became clear thatfelony probation violation was by far the most-charged offense, with 455 people facing that charge. We went back and found all the other inmates who did not have probation charges but had charges indicating a return trip to jail, such as FTA (fail to appear in court) CON (often contempt charges or other charges resulting from failing to appear).
The number was astounding: 579 of the 838 inmates — 69% — fell into the “been here, done that” category.
The numberwas astounding because we wereaware that Virginia is recognized as a leader nationwide for having a strong evidence-based pretrial and probation assessment program, and that the state’s Department of Corrections regularly touted itsrecidivism rate, 23.4% as of their latest press release in early 2019, as the lowest in the country.
How to explain the gap between the state's 23.4% and the 69% we found? We decided to talk to the people caught up in that gap— inmates in jail who, despite their repeated trips to jail, did not show up in the state's official number.
Cleaning up the data just took ...more data... and more time...
By this time it was nearly a year into our investigation and we acquired another jail roster, for March 5, 2018. The names of 205 inmates appeared on both.
We were able to determine by their entrydates that some had remained in jail for those 10 months since May. Others had been released from jail and found their way back in, and so had a new entry date that was later than May 9, 2017.
What about the rest of the 838 inmates? I enlisted the services of our new government reporter, Julia Fair, to help us find them.
For several months, Fair used court records and Department of Corrections website data to find out where people ended up. In this way we were able to determine the outcomes that show up in the funnel chart in our main story.
"I've worked in three different states, each with different court website layouts," saysFair. "The first challenge was understanding how Virginia's system works, what's available online and what isn't — which required more reporting."
With Fair's help we performed additional analyses on the data. For instance, we looked at all the inmates who had larceny charges for racial or gender disparities compared to the local population. This searches did not yield any notable disparities.
"If there were racial or gender disparities, we would have found them," saysFair. "For weeks my only task was running queries on race and gender for each of the 300 charges."
It was important for us to understand as we worked thatthese are moments-in-time glimpsesofthe jail population, and not comprehensive tracking. So while 286 of the original roster's inmates ended up in prison by March 5, 2018, some may have gone straight to prison from jail, some may havebeen released and then violatedprobation and received a "state-responsible" sentence of a year or more that got them transferred into prison. Some may have been released and committed an entirely different crime that led to their prison sentence.
"We had to remember all we had was a snapshot in time. It was as comprehensive as we could get with the information available. And even that took us two years to analyze and package for our readers," says Fair.
In the same way, there may have been inmates from the original roster who got out of jail, violated probation and got put back in jail, served a few weeks or months and were released again on probation before March 5, 2018.
So the 120inmates who were back on the roster in March were the absolute minimum number of inmates who were released but ended up back in jail on either technical violations or a new charge. The real number is likely higher since even a six-month sentence is on the long side for the jail. In the ten months between our two rosters, it is almost certain that some inmates from the original date were released, came back after a violation for a few days or weeks or even months, and were released again.
Eighteen months after sending usthe first jail roster, Wyatt Brown rananother report and we found130 inmates from Nov. 9, 2018, who were also on the original roster.
"We were able to utilize Microsoft Excel and Access to do the heavy lifting for us," saysFair. "Simple tools within the softwareallowed us to compare our snapshot lists to see who kept returning to jail. We needed to find out why."
Three monthslater that number was around 120. All this time we'd been looking at data, some people had made multiple trips to jail and the community and back again.It was time to meet with some of those inmates.
'Revolving Doors' is built on data, but about people
While the story was built on data, it's important to remember that it leads to human beings. Members of our community.
Among our community there are those who have been convicted of horrible crimes, but those are not the ones who find themselves back in our area jail time and time again, and are not the subject of this story. Serious violent felony offenders are the ones sent to state-responsible incarceration in a penitentiary run by the state's Department of Corrections.
We met with jail officials in early March, who agreed to put aside a special time to visit with inmates who would be willing to speak with The News Leader. By doing so the inmates would not be using up their weekly visitation time with the interviews, which might have discouraged some from accepting invitations to be interviewed if it meant they would not get to see their family or friends that week. Inmates can only receive one 15-minute visit each week, according to the jail's website.
On the night of the interviews, one inmate decided at the last minute that he did not want to be interviewed.
All four inmates interviewed mentioned a lack of mental health and drug addictionprograms during their time in jail.
State-run prisonshave more funding for therapeutic programs than the local jail, which in some ways is a "lockdown" and not meant for long term stays, as Middle River Regional Jail's Major Eric Young said in a recent interview.
Young said that there are programs at the jail contracted through Valley Community Services Board, for both mental health and drug addiction, and all inmates are screened upon entry for eligibility to those programs. He says most inmates requesting help get some response in "a few hours to a few days."But most of the inmates are not in for fixed sentences which can make it harder to prioritize and schedule such help, according to Young.
Likewise, Blue Ridge Court Services director Megan Roane emphasized that the probationary periods her team supervises with offenders tend to last only 6 to 12 months, also making the time of contact between probation officer and offender relatively brief.
Roane says 553 of 727probation cases that were opened and closed in Fiscal Year 2018 (July 1, 2017-June 30, 2018) were closed successfully.
A 78% successful probation rate on the one hand in 2018. On the other hand, the jail doors opened over 5,300 times to take in aninmate last year, and about3,000 charges faced by those inmates were indicative of a return visit.
What is undeniable about the numbers is that of those people coming to jail, far more are coming back again than our area's more glowing statistics seem to imply.
