Inaccurate sentencing condemns prisoners to serve longer than is lawful

“What if I just get up and walk out right now?” Michael McPherson posed this question to a friend one day last spring, when they were both inmates at Marcy Correctional Facility, a medium-security prison in Oneida County, New York, about 250 miles north of Manhattan.

McPherson didn’t go through with his escape plan, but in a sense, it would have been justified. His release date had come and gone, and he should have been a free man. But no one — not the prison counselors, corrections officers nor deputy wardens he approached on a daily basis, nor the lawyers, prison grievance committees nor judges he’d been writing to for months — would agree to help.

His story is not an isolated one. A product of byzantine penal codes and indifferent prison bureaucracies, so-called inaccurate sentencing could be affecting thousands or even tens of thousands of inmates nationwide. But while many aspects of the U.S. criminal justice system have come in for scrutiny recently, this problem is almost completely unknown to the general public as well as reformers. Even the few familiar with the issue know of no statewide nor nationwide initiatives seeking to quantify inaccurate sentencing, much less attempt to address it. Terence Davidson, a paralegal with the Legal Aid Society, in New York, who specializes in sentence calculations, says he helps correct 100 to 200 sentences each year. But when considering how common inaccurate sentencing may be across the country, he says, “There’s no way to tell.”

Like some of the inmates incarcerated beyond their rightful terms, McPherson was serving multiple sentences. He’d been arrested in the Bronx on Nov. 15, 2013, on charges of assault and gang assault. While awaiting trial at Rikers Island jail, an additional charge of conspiracy to commit wire fraud was brought against him following the conclusion of an FBI investigation unrelated to the alleged assault. Acknowledging his guilt in the federal case, McPherson took a plea deal and accepted a 15-month sentence. He also accepted a plea deal in the assault case, though he says he was innocent of those charges and only took the deal to avoid a longer term. If he’d gone to court, he would have faced a minimum of five years. Under the plea deal, he received two years for assault in the second degree, to be served concurrently with his federal sentence.

By McPherson’s calculations, the two concurrent sentences, reduced for time served in jail and and good behavior, should have been up on July 31, 2015. But upon receiving his time computation sheet, a report produced by prison officials that calculates release dates, he discovered that the state intended to hold him until Nov. 10, 2015. He immediately began lobbying to correct that.

With his letters ignored and his petitions to prison officials rebuffed, McPherson asked his mother for help. She sought out the Legal Aid Society, a Manhattan-based nonprofit that provides counsel to New Yorkers unable to afford representation. The letter she wrote to Legal Aid in July 2015 made its way to Davidson.

After looking into McPherson’s case, Davidson realized something was wrong: The New York City Department of Corrections had failed to credit the inmate for 307 days of jail time. With those days served, he should have been freed on Jan. 8, 2015, months earlier than even McPherson himself had calculated, due to a slip in his arithmetic. Davidson contacted the U.S. Marshals Service, under whose authority McPherson was being held for wire fraud, to confirm the completion of the federal sentence, and he contacted the New York City Department of Corrections to correct McPherson’s jail time, which was then forwarded to Marcy. A process McPherson had been working on for months was completed by Davidson in less than two weeks. McPherson was released on Aug. 7, 2015, his 27th birthday. It was 96 days earlier than if Davidson hadn’t intervened — but 212 days after the inmate’s sentence actually expired.

“I can’t understand that,” says McPherson, who is now living with his brother in Harlem and working for a window cleaning company. “Whose job is to make sure people don’t have to go through that?”

In New York state, much of the responsibility for calculating sentences rests with the New York City Department of Corrections. The department itself does not keep statistics on inaccurate sentences. But its administration of Rikers Island, the second-largest jail in the U.S., involves the DOC in the majority of sentences calculated throughout the state. The DOC must certify jail time at its facilities, creating a paper trail that, if not properly accounted for, as in McPherson’s case, introduces the first dimension of difficulty in calculating sentences.

Additionally, DOC officers are responsible for performing the actual arithmetic of calculating sentences at city courts and jails. The math seems simple, involving only the addition and subtraction of days from sentences. But the layers of federal, state and municipal statutes, rules and regulations are esoteric. Accurate sentencing requires knowledge of minute aspects of these laws and the ways in which they intersect.

The difficulties can begin right at the moment of sentencing. For example: Does the day of sentencing count as jail time or sentence time? From there, the questions become more particular to the sentence and inmate: Should the amount of time earned for good behavior be estimated at one-third or one-seventh of the sentence? That depends on the kind of sentence, namely whether it’s a determinate sentence — meaning it has a flat term, such as five years — or if it’s an indeterminate sentence — meaning it has a minimum and maximum, such as five to seven years. Besides considering this new sentence, the calculation also has to account for any previous sentences and any outstanding ones. Should any remaining post-release supervision time be added to the new sentence? Again, that depends: Yes, if it’s a determinate sentence; no, if it’s an indeterminate sentence. Questions arise from every facet of the inmate’s criminal history, and getting the right answers requires knowledge and application of very specific laws.

This complexity has been compounded by the devolution of New York state’s penal code in the last five decades, according to Frank Bress, a professor of law and director of clinical programs at New York Law School. “When it was passed, it was sort of an elegant statute, elegant in its simplicity,” he says. “The calculations were a lot easier.” That elegance was lost, according to Bress, through the introduction of new measures, like mandatory minimums, “three strikes” statutes and the Rockefeller drug laws, which lengthened sentences while also complicating their calculations.

According to the DOC, all corrections officers receive five days of “general office training,” which includes sentence calculation, when they’re hired. Captains and assistant deputy wardens receive additional training on discharge procedures, the department says.

Yet two veterans of the DOC who retired in 2005 after collectively serving 40 years say they did not receive any training in sentence calculation nor were they aware of such mandatory training despite their involvement in calculating sentences and overseeing discharges. In response, the department says it updates its training “periodically.”

Regardless, there remains a lack of expertise, which shows, according to Bress, who spent 44 years as a criminal defense attorney in New York. “Quite frankly I don’t think the personnel know what the hell they’re doing,” he says of DOC officers. “It’s frequently a battle with them over the calculation of the sentence.”

But even attorneys who practice criminal law receive little formal training in sentence calculation. It is not part of standard law school curricula, criminal law courses nor bar exams. Bress sets aside one day per semester for the study of sentence calculation as part of his criminal defense clinic, an elective course he teaches at NYLS. The results are not foolproof, he says. Even for relatively simple misdemeanor sentence calculations, 16 students may come up with 16 different answers.

To simplify the process, Davidson of the Legal Aid Society recently developed an online tool known as Sentence Calculator. Users input lengths, dates and types of sentences, along with jail time served and time earned for good behavior. That data is referenced against a database of statutes, rules and regulations related to sentencing in New York state to produce accurate release dates. Legal Aid has already licensed the technology for 800 of its lawyers. Davidson hopes the next step will be adoption by municipal and state governments.

Aside from the toll on inmates and their families, inaccurate sentencing also costs taxpayers. The DOC spends $100,000 per inmate each year. Extrapolating from this figure, Davidson estimates that he saves taxpayers roughly $4 million to $5 million annually by helping to correct prolonged sentences. In addition, inmates who serve overly long sentences sometimes file civil lawsuits against the city and state. McPherson has begun meeting with attorneys to prepare such a suit.

Three of McPherson’s close friends died during the seven extra months he spent incarcerated, he says. “If I would have come home in January,” he points out. “I would have been home for my goodbyes and I-love-yous and to show the families support.”

McPherson estimates that he heard from 150 other inmates at Marcy who say they were serving inaccurate sentences. These men banded together in groups to research relevant laws, file grievances and provide emotional support, he says. One friend, Brock, had already served 30 months on a one- to three-year sentence. After accounting for time served and good behavior, he should have been free.

“When I left he was still there,” says McPherson. “When I was going home he was saying, ‘Any day now, any day now they should fix it.’ ”