Skip to main content
investigation

Federal inmates' risk assessments determine everything from where a prisoner is incarcerated to what rehabilitation programs they are offered. After controlling for a number of variables, The Globe found Black and Indigenous inmates are more likely to get worse scores than white inmates, based solely on their race

Open this photo in gallery:

Nick Nootchtai, shown near his Toronto home in August, was released from prison last fall after serving a 12-year sentence for manslaughter.Galit Rodan/The Globe and Mail


FIRST IMPRESSIONS

After a year spent staring at the walls of his cell at the Sudbury Jail, Nicholas Nootchtai finally had his sentence.

Twelve years. Manslaughter.

It had been less than a week since the judge had signed off on his plea deal, and on a Tuesday evening in November, 2007, a guard came to find him. Nick had a visitor.

They walked through the two-storey jail’s dimly lit hallways to a small room. A wall-to-wall metal table cleaved the space in two, with concrete stools and a door on either side. Across from Nick sat a parole officer with the Correctional Service of Canada (CSC), there to interview him ahead of his transfer to federal prison.

A reserved man, Nick spoke softly and didn’t catch her eye much, though the contraband joint he’d smoked earlier that day – easy to come by at the Sudbury Jail – might have been partly to blame. He was nervous, he told her, about going to prison.

They started with the basics: He was 28 years old and had grown up in the Ojibway First Nations community of Atikameksheng Anishnawbek, a 20-minute drive from Sudbury. They went over his allergies (aspirin), gang affiliations (none), childhood (traumatic) and struggles with alcohol. Nick’s parents had battled alcoholism, and he had his first drink when he was 10. By his early 20s, alcohol was a constant in his life, and when he was 24, he drank so much he fell down a flight of stairs and slipped into a coma.

As for the night of the killing, Nick couldn’t remember much, he said, but one image was seared into his mind: the young man he’d stabbed lying on the ground, bleeding.

The interview lasted nearly an hour. It went well: While he’d shown signs of depression and hopelessness, the officer wrote in her six-page report, he was honest, polite, remorseful.

An ominous paragraph dangled at the report’s end, however. “The subject has failed to develop strong ties to conventional settings i.e. schooling, employment, law and order or any form of pro-social activities.”

Nick didn’t know it, but that initial report – known as a preliminary assessment – would follow him for the next 12 years. The first report a parole officer reads, it would feed into all the assessments to come and every decision regarding his future: where he would serve his sentence, his access to rehabilitative programs, how often he would meet with officers, whether he’d be subjected to solitary confinement, his visitation privileges and, crucially, his odds of getting parole.

Risk assessments are the high prophets of the prison system, used to divine an inmate’s true nature through a mix of numerical scores on standardized tests and parole officers' raw judgments. They’re steeped in decades of research – and they are also fundamentally, powerfully biased against Indigenous and Black inmates, placing them in higher security classifications and assigning them worse odds of successfully re-entering society.

Simply put, Indigenous and Black inmates aren’t at the same starting line as everyone else – a conclusion featured in a host of government reports, academic studies and legal challenges going back years. In 2018, the Supreme Court of Canada agreed, ruling that the CSC had failed to ensure their psychological risk assessments were not culturally biased. The following year, a Senate report noted the security classification tools may be resulting in harsher incarceration terms for racialized people.

Late in 2018, The Globe and Mail set out to measure how race affected a prisoner’s time inside and filed a freedom of information request to the CSC. In response, the agency released nearly 750,000 rows of data, representing 50,000 federally sentenced people between 2012 and 2018.

The Globe has since spoken with more than 60 inmates, lawyers, psychologists, academics and correctional officers; reviewed dozens of academic studies on risk assessment; and read hundreds of pages of correctional files, including those of Nick and other inmates, who provided stacks of records several inches thick.

The interviews and documents reveal the vast influence of these pen-and-paper algorithms, and point to two key scores: an inmate’s security classification, which matters greatly early on in a sentence, and their reintegration score, which factors heavily near its end.

From there, The Globe set out to determine how much of a role race plays in those scores. The results of our statistical investigation – the first done at this depth – are striking.

After accounting for several variables, including age, gender, the severity of inmates' offences and past contact with the criminal justice system, the analysis shows that Black men are nearly 24 per cent more likely than white men to receive a “maximum” initial security rating, the worst possible score, which affects access to treatment programs. Indigenous men, meanwhile, are roughly 30 per cent more likely than their white counterparts to be assigned the worst possible reintegration potential score, which plays a significant role in parole decisions. As a Federal Court judge wrote in 2015, risk assessment scores are “like a branding – hard to overcome.”

Within the seven-year period for which The Globe obtained data, Indigenous and Black men are actually less likely than white men to commit a new offence, and our analysis suggests these scores often overestimate the likelihood they’ll land back in prison. According to experts, they create a criminal feedback loop, with negative scores ratcheting up the long-term odds an inmate will go on to reoffend, ultimately compounding the gross overrepresentation of Indigenous and Black inmates in federal custody.

As of the most recent census in 2016, Indigenous and Black people accounted for 4.8 per cent and 3.5 per cent of the Canadian population, according to Statistics Canada, but made up 25.5 per cent and 8.7 per cent of those in federal prison, respectively, according to the correctional database that same year. In 2015, Prime Minister Justin Trudeau made reducing the incarceration rate of Indigenous Canadians a priority for the minister of justice. But the disparity has only grown: The Office of the Correctional Investigator, a federal watchdog, estimates Indigenous people will account for a full third of inmates sometime in the next three years and that Black people are approaching 10 per cent of the prison population.

“There’s absolutely no doubt in my mind that there is systemic racism in federal corrections,” says Dr. Ivan Zinger, Canada’s Correctional Investigator. A psychologist by training, Dr. Zinger says the difference in Indigenous and Black inmates' scores amounts to “a question of opportunity.” By putting certain groups in higher security levels and focusing on risk, he says, the CSC limits their opportunities for programming, making it harder for inmates to receive treatment and learn the skills they need to get out of prison. “If the tools actually overclassify people, and the programs aren’t as available or effective at higher security, then of course it perpetuates systemic discrimination,” he says. “Absolutely."

Though the CSC declined The Globe’s interview requests, spokeswoman Véronique Rioux sent an e-mailed statement that referenced 22 separate internal research reports. Ms. Rioux says studies "continue to support their use with Indigenous and Black offenders,” and that the CSC is conducting research both on "some of its central classification tools” and “to better understand the experience of ethnocultural offenders under its care.” She also notes all CSC employees undergo mandatory diversity and culture training, and are required to consider inmates' cultural backgrounds during assessments.

The agency recognizes there is an “unacceptable overrepresentation" of Black and Indigenous people in Canadian prisons, Ms. Rioux wrote. “This takes time, and we know there is more work to do.”



‘HERDING CATTLE’

Mr. Nootchtai attended 4/5 elementary schools and 3 high schools. His attendance was poor but he said his grades were good. He was held back a year in grade 5…He was suspended many times and expelled from elementary school…He dropped out of school in grade 10 because he was getting too old for the crowd in his class.

—Psychological assessment dated March 6, 2009

In January, 2008, two months after his preliminary assessment, Nick was transferred to federal prison, where most inmates receiving a sentence of two years or longer serve their time. His destination was Millhaven Institution, a maximum-security facility near Kingston.

For Nick, it was a relief. Sudbury Jail had a reputation as an end-of-the-line assignment for reprobate correctional officers, and drugs flowed freely. It also stank – of old garbage, vomit, marijuana and feces. At Millhaven, he had a roughly six-by-10-foot cell all to himself – a luxury compared to his tight provincial quarters, where he was sometimes triple-bunked.

He traded in the jail’s orange jumpsuit for the jeans and T-shirt worn by federal inmates and joined the prison’s assessment unit – generally a prisoner’s first stop.

The next several months were a blur of tests, orientation sessions and meetings with the officers who would determine his fate over the course of his sentence. “It was just herding cattle,” Nick recalls. “Lines everywhere.”

The idea of forecasting inmates' criminal behaviour stretches back nearly a century. In 1928, Canadian-born University of Chicago sociologist Ernest Burgess built a statistical tool, later adopted by the Illinois Board of Paroles, that predicted whether an inmate would go on to reoffend. Some of the 22 variables used in the study were prosaic, like the nature of the crime. Others were more baroque. For “social type,” Burgess gave examples like ne’er-do-well, drunkard and hobo.

Another variable was the race of the inmate’s father.

In the 1980s and 1990s, using their own administrative data, the Correctional Service of Canada developed a suite of assessments to determine an inmate’s security level and potential for reintegration. Officers rely on these tools to create a correctional plan, the roadmap for an inmate’s rehabilitation and release. This document, updated throughout the sentence, summarizes the offence, progress on treatment programs, ongoing behaviour and all major risk scores.

Two assessments are particularly crucial.

The first is the Custody Rating Scale, meant to measure what kind of security risk an inmate poses inside. The 12 multiple-choice questions – focused on the severity of the current offence, any history of drug or alcohol use, and so on – are filled out by a parole officer and tallied up to set an inmate’s security level at minimum, medium or maximum. (Officers can override an inmate’s score if they think it’s too high or too low.)

Those designations roughly correlate to the security levels of Canadian prisons themselves: Get a “maximum” score and you’ll likely be sharing a facility, or at least a range, with other maxes, including people doing time for murder. A minimum rating, meanwhile, might mean you get to live in a dorm-style minimum-security prison, cooking your own meals alongside fraudsters.

Since certain treatment programs and jobs are only available at specific institutions, an inmate’s security level is arguably their single most important score. And though levels can change over time, that first one is decisive.

What a Custody Rating Scale looks like

This is a real inmate’s CRS assessment with their identifying information removed. It notes that the inmate has not tried to escape and has no history of “institutional incidents,” which would include things like assault, theft or the possession of prohibited weapons.

The second key assessment is the Reintegration Potential Score, which gains in importance toward the end of an inmate’s sentence.

The Parole Board of Canada uses that score – low, medium or high – to estimate how prepared an inmate is to re-enter society and whether they pose a public safety risk if released.

Part of this score is based on the Static Factors Assessment, which focuses on an inmate’s past convictions and behaviour. Another part depends on the inmate. Per regulations, if an inmate is Indigenous or a woman, the CSC uses the Dynamic Factors Identification and Analysis test, which measures the alterable aspects of an inmate’s life, including education, employment history, relationships, family ties, history of substance abuse, community attachments, and emotional and mental state. (Though all inmates have a Dynamic score, it only factors into the reintegration potential rating for Indigenous and female inmates.) For everyone else, there’s the Statistical Information on Recidivism scale, which has never been used on Indigenous inmates or women because of decades-old concerns about cultural bias.

Unlike the Custody Rating Scale and Statistical Information on Recidivism tools, which are math-based, “actuarial” risk assessments, the Static and Dynamic tests are based purely on the administering officer’s judgment – which allows all sorts of biases to creep in.

What the Static and Dynamic tests look like

These are redacted excerpts from the Static and Dynamic tests, respectively, from a male inmate at Millhaven Institution serving time for second-degree murder. They note a history of substance-abuse and mental-health problems, but no prior offences, gang affiliations or domestic violence.

Bernard Harcourt, a law and political science professor at Columbia University in New York, has spent years studying the risk tools used in criminal justice. In his book Against Prediction, Prof. Harcourt argues that when assessments rely on an inmate’s past – and particularly their criminal history – it becomes a proxy for race.

“You’re not predicting the actual essence of the person and their dangerousness – you’re predicting how likely it is that, because of policing disparities, they would get arrested again,” he says. “You’re not predicting crime. You’re predicting policing."

In other words, being Black or Indigenous becomes a risk factor in itself.

While the CSC’s assessments don’t explicitly ask for an inmate’s race, Prof. Harcourt says prejudice is baked right into the questions. For instance, the Custody Rating Scale asks a parole officer to rate an inmate’s “street stability” as “below average,” “average” or “above average” based on their employment, education, family, relationships and housing status.

“That’s so vague and open-ended,” Prof. Harcourt says. “You can just project all of your stereotypes right there.”

The assessment process isn’t entirely test-based – part of it involves the inmate retelling the story of how they ended up in prison. For Nick, that meant recounting the night he killed 20-year-old Gavin Nettagog.

The details and sequence of events that night vary across official reports and Nick’s own recollections, but the substance of it is simple enough: Late on Nov. 23, 2006, after an evening of drinking and smoking weed, Nick returned to his apartment in the Donovan, a hardscrabble neighbourhood near downtown Sudbury. Though it was near freezing, he poured himself some Wiser’s whisky and sat on the stoop of the squat, two-storey building.

A cigarette butt landed nearby. He looked up in time to see a young man – a kid, really – duck back into the apartment above. The people upstairs were trying to toss the butts into his cup, he realized.

He yelled at them and then went back inside, but he soon heard a knock outside his window. He chased the group upstairs, and a fight broke out. Nick lost. He retreated to his apartment, changed his clothes and, in a rage, grabbed a kitchen knife.

He found the group outside. A moment later, he stabbed Mr. Nettagog in the back. Nick went back into his apartment and hid in the bathroom until the police arrived. He didn’t know it yet, but he’d pierced a lung. Mr. Nettagog was pronounced dead at the hospital later that night.

It was a brisk, painful recital he’d be asked to perform for psychologists, therapists and parole officers again and again over the next dozen years.

“I said it a lot – that I probably should’ve gotten more time,” he says. To this day, he goes back and forth on whether his sentence was harsh enough. “Sometimes I do, sometimes I don’t,” he says with a long pause. “Yeah. Sometimes.”

Four months after he’d arrived, Nick’s time in the assessment unit was finally up.

“SECURITY CLASSIFICATION: MAXIMUM,” his correctional plan read. “REINTEGRATION POTENTIAL: LOW.”

When he received his prison assignment, a lump formed in his throat. Nick wasn’t leaving Millhaven. He was moving down the hall, to J-unit.



THE ASSESSMENT PROCESS

The Correctional Service of Canada uses several risk assessments to place, treat and release inmates. Here’s how the process works, and how scores affect their time in prison.

All inmates

Indigenous or female inmates

Everyone else

SENTENCING

Inmates receiving a sentence of two years or longer become the responsibility of the Correctional Service.

TRANSFER TO

ASSESSMENT UNIT

Most inmates undergo mental health screenings and a battery of risk assessments, including these tests:

Custody Rating Scale

This 12-question, math-based

“actuarial” tool classifies inmates into one of minimum, medium or maximum. Those ratings are then used to match inmates to a particular prison.

Static Factors

Assessment

Measures an inmate’s past involvement with the criminal justice system. Based on the parole officer’s judgment, this assessment has a final score of low, medium or high.

Dynamic Factors

Identification

and Analysis

Looks at the aspects of an inmate’s life that can be changed or improved. This assessment also relies on the parole officer’s judgment, and its final score is either low, medium or high.

Statistical

Information

on Recidivism

Another actuarial tool, this scale looks at an inmate’s criminal history and certain demographic factors such as age, marital status and employment status when arrested.

Security

LEVEL

REINTEGRATION

score

Final scores and

prison placement

Inmates receive their Correctional Plan, which details their scores including their reintegration potential rating. The plan also assigns them a federal prison and lays out a treatment program.

Both scores continue to impact an inmate’s time in prison

Access to treatment

programs, activities

Inmates with a maximum security level have limited movement privileges, difficulty accessing programs or even obtaining arts and crafts supplies.

Parole and

statutory release

Inmates with a low

reintegration potential score may have more difficulty getting paroled, and if their scores are low enough they may be denied statutory release.

AFTER SENTENCE ENDS

Inmates who score poorly may have conditions attached to their release, such as checking in weekly with police.

THE ASSESSMENT PROCESS

The Correctional Service of Canada uses several risk assessments to place, treat and release inmates. Here’s how the process works, and how scores affect their time in prison.

All inmates

Indigenous or female inmates

Everyone else

SENTENCING

Inmates receiving a sentence of two years or longer become the responsibility of the Correctional Service.

TRANSFER TO

ASSESSMENT UNIT

Most inmates undergo mental health screenings and a battery of risk assessments, including these tests:

Custody Rating Scale

This 12-question, math-based

“actuarial” tool classifies inmates into one of minimum, medium or maximum. Those ratings are then used to match inmates to a particular prison.

Static Factors

Assessment

Measures an inmate’s past involvement with the criminal justice system. Based on the parole officer’s judgment, this assessment has a final score of low, medium or high.

Dynamic Factors

Identification

and Analysis

Looks at the aspects of an inmate’s life that can be changed or improved. This assessment also relies on the parole officer’s judgment, and its final score is either low, medium or high.

Statistical

Information

on Recidivism

Another actuarial tool, this scale looks at an inmate’s criminal history and certain demographic factors such as age, marital status and employment status when arrested.

Security

LEVEL

REINTEGRATION

score

Final scores and

prison placement

Inmates receive their Correctional Plan, which details their scores including their reintegration potential rating. The plan also assigns them a federal prison and lays out a treatment program.

Both scores continue to impact an inmate’s time in prison

Access to treatment

programs, activities

Inmates with a maximum security level have limited movement privileges, difficulty accessing programs or even obtaining arts and crafts supplies.

Parole and

statutory release

Inmates with a low

reintegration potential score may have more difficulty getting paroled, and if their scores are low enough they may be denied statutory release.

AFTER SENTENCE ENDS

Inmates who score poorly may have conditions attached to their release, such as checking in weekly with police.

THE ASSESSMENT PROCESS

The Correctional Service of Canada uses several risk assessments to place, treat and release inmates. Here’s how the process works, and how scores affect their time in prison.

All inmates

Indigenous or female inmates

Everyone else

SENTENCING

Inmates receiving a sentence of two years or longer become the responsibility of the Correctional Service.

TRANSFER TO

ASSESSMENT UNIT

Most inmates undergo mental health screenings and a battery of risk assessments, including these tests:

Custody Rating Scale

This 12-question, math-based

“actuarial” tool classifies inmates into one of minimum, medium or maximum. Those ratings are then used to match inmates to a particular prison.

Static Factors

Assessment

Measures an inmate’s past involvement with the criminal justice system. Based on the parole officer’s judgment, this assessment has a final score of low, medium or high.

Dynamic Factors

Identification

and Analysis

Looks at the aspects of an inmate’s life that can be changed or improved. This assessment also relies on the parole officer’s judgment, and its final score is either low, medium or high.

Statistical Information

on Recidivism

Another actuarial tool, this scale looks at an inmate’s criminal history and certain demographic factors such as age, marital status and employment status when arrested.

Security

LEVEL

REINTEGRATION

score

Final scores and prison placement

Inmates receive their Correctional Plan, which details their scores including their reintegration potential rating. The plan also assigns them a federal prison and lays out a treatment program.

Both scores continue to impact an inmate’s time in prison

Access to treatment programs, activities

Inmates with a maximum security level have limited movement privileges, difficulty accessing programs or even obtaining arts and crafts supplies.

Parole and statutory release

Inmates with a low reintegration potential score may have more difficulty getting paroled, and if their scores are low enough they may be denied statutory release.

AFTER SENTENCE ENDS

Inmates who score poorly may have conditions attached to their release, such as checking in weekly with police.



J-UNIT

NOOTCHTAI continues to blame this writer for not placing him in the required substance abuse program. He believes that completion of the program would stop him from making brew… This writer has no control over program assignments and the allocation of those resources.

—Correctional Plan dated Sept. 16, 2013

J-unit, widely regarded as the toughest maximum-security unit in Ontario, was a common destination for lifers starting out their sentence.

Because of his security designation, Nick’s movements were severely limited, and one of the only jobs he could get was sweeping the range. But he quickly found employment of another sort: It turned out he had a talent for making “brew,” contraband alcohol made by fermenting fruits in a dark, warm place before being distilled; his was often made in a footlocker.

Gathering materials for his clandestine microbrewery – sugars, fruits, potatoes – gave Nick something to do. It also earned him respect and extra noodles from the canteen, and offered a measure of safety from other inmates. But it also fed his alcoholism. “I wanted to stop,” he says, "but no one else saw it that way.”

His maximum score also hindered his participation in Indigenous activities. In a 2013 progress report, an elder wrote that unless he was transferred to a lower-security range, Nick wouldn’t be able to attend sweat lodge ceremonies or talking circles. He even had a hard time accessing paint and pencil crayons, which the talented artist used to explore his feelings of remorse and desire for help.

Given his long sentence, the CSC was in no hurry to get him treatment for substance abuse and violence, either. “Closer to your release date, we’ll give you programs,” Nick recalls his parole officers telling him. “That’s when you’ll need them.”

But completing programs is one of the best ways for an inmate to “cascade down” during their time in custody, moving from maximum to medium to minimum before finally being paroled. Nick was stuck.

Open this photo in gallery:

Scenes from Millhaven Institution: The mental-health range at J-unit in 2020, left, and a cell's food slot and interior in 2013, right. In J-Unit, Mr. Nootchtai had little freedom and little to do but brew alcohol clandestinely.Office of the Correctional Investigator

This catch-22 is all too familiar to Rick Sauvé, who was handed a life sentence for murder in 1979 and paroled in the 1990s. Mr. Sauvé, 68, has devoted his life since then to helping prisoners: He’s won cases before the Supreme Court of Canada, testified before Parliament and received awards for pushing for human rights in the correctional system. Nonetheless, he’s still required to check in regularly with his parole officer.

Risk assessments, he says, are an inmate’s ticket out of prison. But because they’re so reliant on the person scoring them, they can feel arbitrary. It’s not uncommon for inmates to see a new set of scores when they switch parole officers, he notes. “It just doesn’t make sense.”

The system also focuses almost exclusively on the negative, says Mr. Sauvé. If you spend 29 days behaving well but break the rules on day 30, only that last day’s getting written up.

The reports can often be riddled with mistakes. Some of Nick’s documents call him by the wrong name; others say he isn’t Indigenous. (It’s common knowledge, he says, that some parole officers cut and paste passages across reports to save time.)

Open this photo in gallery:

In this passage from Mr. Nootchtai's preliminary assessment, another prisoner's name, Sutherland, appears to be pasted in.Galit Rodan/The Globe and Mail

They also note an impressive list of bootlegging infractions. “NOOTCHTAI’s continuous involvement in brew-making activities has been one of the primary reasons for his continued MAXIMUM security classification,” one correctional plan reads.

According to an analysis of the service’s data, Black men are 23.8 per cent more likely than white men to end up with a maximum designation at admission, after accounting for a litany of variables, including age, past contact with the justice system, offence severity and whether or not the inmate is serving a life sentence. (The analysis didn’t result in statistically significant findings for Indigenous men.)

Though the Correctional Service has a mechanism for appealing security-level decisions, figures show this doesn’t happen often. Between the 2015-16 and 2018-19 fiscal years, the CSC told The Globe it received a total of 722 grievances regarding security classifications, or 0.8 per cent of all grievances received during that period.

It can take years for scores to change. The Globe’s analysis found that more than a third of newly admitted inmates see no change in their security level within a three-year period, while two-thirds saw no change in their reintegration score within the same time-frame. (Mr. Sauvé teaches inmates to short-circuit the process by applying for prison transfers and day passes, both of which force automatic reassessments.)

Most non-lifers are eligible for full parole after serving the first third of their sentence, and are required by law to be released back into the community once they’ve served two-thirds, a form of supervised discharge known as statutory release. (This can be denied under certain circumstances, and if an inmate breaks the terms of their release, they can end up back inside.)

To get parole, it helps to have a good reintegration score – but The Globe found Indigenous men are 29.5 per cent more likely to receive the worst reintegration rating of "low.” (The analysis found that Black men, meanwhile, are actually 6.1 per cent less likely than white people to end up with the worst score.)

Yet, after controlling for the score, the inmate’s age and amount of time since their release, Indigenous men are 9 per cent less likely to reoffend than white men, and Black people a full 41.1 per cent less likely to do so. In other words, both groups receive worse reintegration potential scores than they should.

Zef Ordman, a parole officer and regional vice-president of the Union of Safety and Justice Employees, has no doubts about why these discrepancies exist.

“I don’t think these tests were designed for women, Indigenous people or probably other visible minorities,” says Mr. Ordman, who has spent a decade with the CSC at Bowden Institution, a medium-security facility near Red Deer, Alta., where he has administered many of these assessments himself. “They want a rigid tool that would apply to everybody.”

The tests simply can’t account for people’s drastically different experiences and upbringings, he says.

“If you grow up on a reserve with high crime, high fetal alcohol syndrome, high abuse, high historical socio-economic issues – well, maybe the person actually isn’t that bad, right?”

“It’s not comparing apples to apples,” Mr. Ordman continues. “That’s frustrating if you’ve got a tool that’s trying to push everybody into a square block or a round circle when they just won’t fit.”

He says he finds it frustrating to administer the tests given his doubts about their fairness, but there’s not much he or other parole officers can do.

“It’s a little bit like being in a factory, and on the conveyor belt, inmates are coming through. There’s timelines,” he says. “Your hands are tied.”


Open this photo in gallery:

Some of Mr. Nootchtai's art is shown at his home in Toronto.Galit Rodan/The Globe and Mail


HOLLOW VICTORY

There have been no further instances of intoxication and you indicated you have been sober since that single incident. Based on your past, this period of sobriety is a significant achievement. However, the Board also canvassed with your apparent inability to view alcohol as a risk factor when you met with the psychologist earlier this year.

—Parole Board of Canada decision dated June 22, 2017

Prison hadn’t gone easy on Nick.

During the first seven years of his stint, he spent 212 days in solitary, averaging a month in the seg unit each year. He spent two years in the notorious Kingston Penitentiary before it shut down, at which point he was transferred back to Millhaven. This time, he was out of J-unit.

His statutory release date – November, 2015, the nine-year mark of his sentence – came and went; officers had argued to the Parole Board there was reason to believe he’d commit a new violent offence.

At least Nick now had a partner. During his time at the Sudbury Jail, he’d cultivated a friendship with Katheryn Wabegijig, a one-time high school girlfriend he phoned on a lark from behind bars. He called her often, and they wrote each other frequently. Within a year, they were in a relationship, limited as it was to the prison’s visiting room. By 2010, they were married.

Open this photo in gallery:

During his time in prison, Mr. Nootchtai rekindled a relationship with Katheryn Wabegijig, who would become his wife.Galit Rodan/The Globe and Mail

With Katheryn’s prodding, he began paying closer attention to his scores and to all the things they denied him, like conjugal visits with his wife in a private apartment inside the facility. He was always denied, in part because of his scores – a “high-moderate risk for future general and violent re-offence,” according to one rejection letter. It would be years before he and Katheryn could finally be together in private.

Nick’s frustration finally boiled over after a psychological assessment in February, 2017. When he saw the psychologist’s report, he couldn’t believe it: Several sections were repeated word-for-word from his previous two assessments – completed in 2009 and 2015 – and he’d received the exact same scores across a number of psychological tests for the third time.

He’d been in prison for more than nine years, and been denied statutory release and private visits. But he’d also gotten his high school diploma, found a loving partner and completed several programs, including for his alcoholism. What more did they want?

A few months later, during his annual statutory release review with the Parole Board, Nick finally got his break. In its report, the board said it believed Nick and his Indigenous elder adviser when they said he was maintaining his sobriety and only continued bootlegging because of pressure from other inmates. He was being released.

But there was a snag: Officers aren’t required to make living arrangements ahead of hearings for inmates with low reintegration scores, which meant Nick had nowhere to go.

Six weeks later, he was released to a halfway house in Toronto near Katheryn. A few months after that, during a routine check, a worker discovered a small empty vodka bottle in Nick’s room. He maintains to this day he’d found it empty the night before and had brought it home to sketch for an art project.

“I am still sober six months and counting,” Nick wrote in a series of letters pleading with the Parole Board to give him another chance. “I am sorry for putting my freedom in jeopardy. No matter the decision, I will remain focused on my goals, and continue sobriety one day at a time.”

In early 2018, the board sent him back to prison. He’d be serving out the rest of his sentence inside.


Open this photo in gallery:

Millhaven Institution, shown in early 2020.Office of the Correctional Investigator


WILLFUL BLINDNESS

Years before Nick Nootchtai killed Gavin Nettagog, the Correctional Service of Canada already knew its risk assessments were broken.

In 2000, a Métis inmate named Jeffrey Ewert filed a grievance with the CSC. Incarcerated since 1984 for the grisly murder and attempted murder of two women, Mr. Ewert argued that several of the psychological assessments used by the CSC had never been properly tested on Indigenous inmates. His own scores were so high, he said, they would prejudice any future parole applications.

Three years later, as Mr. Ewert’s complaint wended its way through the federal government, an employee at the CSC’s national headquarters e-mailed a colleague:

“This is timely in that we have already flagged this issue as a concern. In fact, the Research Branch has already begun some work on this – if only in a preliminary capacity. I suspect that the inmate will win his case and that this will force our hand as a Service. And rightly so! It has always been our position that the inappropriate use of actuarial scales and measures adversely affects our Aboriginal population. In fact, we contend that the use of these measures artificially inflates need and risk ratings.”

Nonetheless, the CSC’s Static Factors Assessment has never been updated. The Dynamic Factors Identification and Analysis test, used on Indigenous and female inmates, was modified once, in 2009. The Custody Rating Scale, the actuarial assessment used to calculate an inmate’s first security level, was also updated just once, in 2001, a year after Mr. Ewert filed his initial grievance. The CSC, for its part, says it conducts research regularly to ensure its tools are still reliable.

In 2016 – the same year Mr. Ewert lost an appeal at the Federal Court, bumping his case all the way to the Supreme Court – an analysis by the Office of the Auditor-General found that Indigenous inmates received higher security classifications more frequently than other inmates. In response to those recommendations, the CSC says it has adopted a plan to “transform Indigenous case management and corrections.”

Open this photo in gallery:

Clouds pass by the Supreme Court of Canada in Ottawa.Adrian Wyld/The Canadian Press

It still hadn’t done so by the time Ewert v. Canada wound up at the Supreme Court in 2018. In a 7-2 decision written by now-Chief Justice Richard Wagner, the court said the CSC had not done enough to ensure its psychological assessments worked for Indigenous inmates and ordered the CSC to study the issue.

Civil rights lawyer Jason Gratl, who represented Mr. Ewert, says the consequences of Canada’s colonial history, poverty and disrupted family structures all factor into an inmate’s score, but they’re often a harsh reality of growing up Black or Indigenous. “You’re classified to maximum at birth, which is close to a technical definition of discrimination: reliance on factors beyond a person’s control to determine their outcomes,” he says.

A year after the ruling, the Office of the Correctional Investigator excoriated the CSC for failing to address the concerns raised in the Ewert case.

In the meantime, senators have also been looking at risk assessments. In 2019, the Standing Senate Committee on Human Rights published an interim report looking at the human rights of federal prisoners. The overly harsh security and reintegration scores for Black and Indigenous men came up frequently during consultations and prison visits.

“It is pretty clear prisoners' rights are being violated every single day,” says Senator Wanda Bernard, who chaired the committee in 2019. “It’s creating the conditions for systemic racism, but it’s also further entrenching systemic racism.”

The CSC, she adds, has known about this issue long enough now that it ought to have a culturally relevant tool. Ms. Bernard hopes to issue a final report and recommendations in 2021.

Open this photo in gallery:

Senator Wanda Bernard is one of the parliamentarians who's been looking at reforms to the prison risk-assessment process in Canada.Senate of Canada

According to the experts The Globe talked to, simply focusing on risk assessments won’t cut it. Since they reflect societal biases, any long-term solutions will have to include reforms to the criminal justice system as a whole, and to the social conditions of Black and Indigenous people.

Jody Wilson-Raybould, the independent member of Parliament for Vancouver Granville and Canada’s justice minister from 2015 to 2019, says The Globe’s findings reflect a central problem with the justice system: Racialized people are judged without accounting for differing cultural or life experiences. “You’re making assessments about people based on what you think to be right," she says. "It’s very paternalistic.”

Ms. Wilson-Raybould says the time for more reports is over – it’s time for solutions, starting with the elimination of mandatory minimum penalties, diversion programs that would reduce the number of people being prosecuted, a greater emphasis on restorative justice, and allowing Indigenous communities to handle criminal cases themselves. Though she was justice minister for a little over three years, during her tenure the criminal reforms she described largely didn’t materialize. She tried, she says, but the government never had the will to make them happen. “We just have to have leadership show up.”

In an e-mailed statement, Mary-Liz Power, a spokeswoman for Minister of Public Safety Bill Blair, says he’ll be speaking with the CSC commissioner Anne Kelly “about how we can eliminate systemic barriers in our federal corrections institutions, and ensure that all inmates have access to appropriate and effective programming and assessments.”

“It is unacceptable that Black, Indigenous and all racialized people face significantly disparate outcomes at the hands of our justice system,” Ms. Power says.

Correctional Investigator Ivan Zinger is hoping for a complete reset. “Corrections has to develop tools from the ground up using Indigenous, Black or women’s samples, to make sure the tools they use bring in the factors that are relevant to that population,” he says. “The Service should’ve done it a long, long time ago.”

When asked about charges that the CSC has been dragging its feet, Ms. Rioux, the agency’s spokeswoman, says it is “collaborating with Canadian academic experts to explore the development of assessment tools that are premised on a full understanding of risks and related concepts associated with Indigenous cultures.”

Many experts who spoke to The Globe can’t understand what the holdup is.

Reports produced by the CSC’s own research department, responsible for analyzing inmate data, admit some of their assessments aren’t doing a great job. In a 2014 study, for instance, researchers found the Static Factors Assessment, used in part to calculate the reintegration score, predicted whether an Indigenous inmate would reoffend only slightly better than a coin flip.

Stephen Hart, a professor of clinical forensic psychology at Simon Fraser University and a world-renowned designer of risk assessments, says the CSC has been “willfully blind” to the problems for a very long time. The CSC “knows, has known, and ought to know now that this is a real problem,” he says.

“To this day, it’s still giving excuses about why they’re not doing that work and honouring that commitment,” says Dr. Hart, who testified for Mr. Ewert during his Federal Court case. “It’s profoundly disappointing.”

These risk scores put inmates under increased pressure while they’re inside, deny them treatment, extend their stays in custody and ratchet the conditions attached to their release, he says. In the end, having a negative score likely increases an inmates' odds of reoffending, too. “They become a self-fulfilling prophecy,” Dr. Hart says.

While controlled academic studies would be required to test that theory, the dataset obtained by The Globe does hint at it. Three years into their sentences, Indigenous inmates eligible to serve out their time in the community – through parole or statutory release – actually did so far less frequently than white or Black inmates. Those with a maximum security level were especially hard-hit: only 5 per cent of eligible Indigenous prisoners were in the community, compared with 34 per cent of white inmates.

With Dr. Hart, risk assessments take on a philosophical quality. The field is split between those using tests to predict the future, he says, and others, like him, who think tests should be used to shape it by identifying treatment. “The people who think you can predict the future think the future is already written.”

This is no longer a question of fairness, Dr. Hart says; it’s about ethics. “I can’t sit back anymore and try to quietly nudge things along. Sometimes you’ve got to blow up the system if it’s not going to change on its own.”


How frequently do people with a given risk score get released?

All federal inmates eventually become eligible to serve their sentence out in the community, be it through parole or statutory release. But three years into their sentence, release-eligible Indigenous people were doing their time out in the community less frequently than white or Black inmates. Those with a “maximum” security level were released seven times less often than white inmates with the same score.

Percentage of inmates released three years

into their sentence, by race and security level

White

Black

Indigenous

MINIMUM

MEDIUM

MAXIMUM

80.7%

82.1

57.4

43.1

40.9

34.3

30.8

27.3

4.9

Note: Excludes people serving life sentences.

tom cardoso and JOHN SOPINSKI / THE GLOBE

AND MAIL, SOURCE: correctional service

of canada

How frequently do people with a given risk score get released?

All federal inmates eventually become eligible to serve their sentence out in the community, be it through parole or statutory release. But three years into their sentence, release-eligible Indigenous people were doing their time out in the community less frequently than white or Black inmates. Those with a “maximum” security level were released seven times less often than white inmates with the same score.

Percentage of inmates released three years

into their sentence, by race and security level

White

Black

Indigenous

MINIMUM

MEDIUM

MAXIMUM

80.7%

82.1

57.4

43.1

40.9

34.3

30.8

27.3

4.9

Note: Excludes people serving life sentences.

tom cardoso and JOHN SOPINSKI / THE GLOBE AND MAIL,

SOURCE: correctional service of canada

How frequently do people with

a given risk score get released?

All federal inmates eventually become eligible to serve their sentence out in the community, be it through parole or statutory release. But three years into their sentence, release-eligible Indigenous people were doing their time out in the community less frequently than white or Black inmates. Those with a “maximum” security level were released seven times less often than white inmates with the same score.

Percentage of inmates released three years into their sentence, by race and security level

White

Black

Indigenous

MINIMUM

MEDIUM

MAXIMUM

82.1

80.7%

57.4

43.1

40.9

34.3

30.8

27.3

4.9

Note: Excludes people serving life sentences.

tom cardoso and JOHN SOPINSKI / THE GLOBE AND MAIL,

SOURCE: correctional service of canada


A WOODEN CANE

Mr. Nootchtai has informed this writer that he believes he has been rehabilitated. A review of file information demonstrates that gains have been made and Mr. Nootchtai has not only matured but appears to have a clear sense of direction. He is encouraged to continue on, and incorporate his wife into his healing journey plans. It is hoped that Mr. Nootchtai can move forward with a new skill set to live a productive, pro-social and fulfilling life.

—Correctional plan dated July 9, 2019

Nick spent the last 21 months of his sentence back in prison, where his scores never budged beyond medium security, low reintegration potential. By the end, he’d racked up close to 70 institutional charges, mostly related to brew.

He walked out on Nov. 13, 2019, 12 years to the day after he was sentenced.

As the one-year anniversary of his release nears, a final prophecy still looms: One of the scores on his last psych assessment estimated he had a 57 per cent chance of reoffending within a year.

Nick has no intention of letting that happen. “I’m not a number,” he says. “Nothing is sending me back.”

It’s hard not to wonder whether a different set of scores might have changed things. There’s a pang of anger in his voice – or perhaps it’s regret. “In my case, I ended up staying longer,” he says. “It’s not right. People change. People can turn around.”

Now 41, Nick lives with Katheryn in Toronto, where he attends vocational school and plans to find a job in home renovation.

When he thinks back to everything that’s happened since that cold night, he remembers the six words carved into a wooden cane an elder once gave him. They’ve become a mantra, a way of steadying himself when setbacks feel too great. The words appear throughout his documents, scribbled on the back of correctional records chronicling half his adult life.

“One door closes, one door opens,” they read.

Open this photo in gallery:

Galit Rodan/The Globe and Mail

Credits

Digital presentation: Evan Annett

Art direction: Ming Wong

Data visualization: Tom Cardoso, Murat Yükselir, John Sopinski and Chris Manza

Data verification: Chen Wang and Jeremy Gray

Editing: Dawn Calleja


More: Tom Cardoso on the investigation

A Globe and Mail investigation found that Black and Indigenous inmates in federal prisons are more likely to get poor scores on risk assessments that determine everything from where a prisoner is incarcerated to what rehabilitation programs they are offered. Tom Cardoso joined Dakshana Bascaramurty for a Facebook Live discussing the findings and process.

The Globe and Mail

Follow related authors and topics

Authors and topics you follow will be added to your personal news feed in Following.

Interact with The Globe

Trending