Structured Intervention Unit Implementation Advisory Panel 2022 to 2023 Annual Report

Time for Meaningful Change

Second Annual Report (2022 to 2023) of the Structured Intervention Units Implementation Advisory Panel

January 11, 2024

Executive Summary

In late 2019, the use of administrative segregation by Correctional Service Canada (CSC) was legally abolished in federal penitentiaries following the implementation of Bill C-83: An Act to Amend the Corrections and Conditional Release Act and another Act (42nd Parliament, 1st session). Bill C-83 was a legislative response to a series of legal decisions in which the courts determined that the use of administrative segregation was a violation of prisoners’ Charter rights. Instead of segregating prisoners in isolating conditions of confinement somewhere in a penitentiary, Parliament specified that certain cell spaces would be designated as Structured Intervention Units (SIUs). CSC created these SIUs in 15 of Canada’s penitentiaries. Prior to the coming into force of Bill C-83, certain prisoners might be segregated from other prisoners for administrative or disciplinary reasons. After November 2019, prisoners placed in newly identified SIUs, and were, by law, to be returned to the mainstream population as quickly as possible. During 2nd Reading debate of Bill C-83, the Public Safety Minister informed Parliament that “Inmates in SIUs will be out of their cells for at least four hours daily with a minimum of two hours of meaningful interaction...."Footnote 1 Further, the use of the SIUs was to be reviewed by a new statutorily created oversight mechanism – Independent External Decision Makers (IEDMs). Taken together, Bill C-83 changes were meant to eliminate the use of isolating conditions that amounted to solitary confinement.

The findings in this report, and in previous reports issued by this Panel, are unambiguous. The intent of the government when the above changes were introduced has not been realized and the promises made to the people of Canada to end isolating conditions of confinement have not been kept. As outlined in the discussion of the evidence that follows, the SIU regime is not working as intended with the passing of Bill C-83. Practices that reasonable people reading the legislation might think would no longer exist, such as long stays in isolating conditions with little or no meaningful contact with other people, persist in federal correctional facilities.

CSC is currently able to justify some of these shortcomings with a strict reading of the letter – though not the intent – of the law. For example. CSC explains that the failure to provide the legislated minimum four hours out of cell is due to prisoners not “availing” themselves of the specific activity that CSC offers. While CSC may demonstrate compliance with the law through the act of making offers to prisoners to leave their cell for the 4-hour minimum, we believe CSC remains obliged to do whatever is within its ability to actually achieve the minimum of 4-hours spent out of cell. The inability to achieve this result is not minimized by attributing the failure to do so to prisoner refusals to accept the offers. Importantly, one of the over-arching purposes of the legislative response to Canadian court concerns about the use of segregation was to end social isolation of prisoners.

This is the second Annual Report of a PanelFootnote 2 established in 2021 by the Minister of Public Safety Canada to examine the implementation of the SIU regime. Prior to this report, there have been nine empirically based, and publicly available, reports on CSC’s SIUs beginning in October 2020 and carrying through until the spring of 2023.Footnote 3 None of these reports suggest that CSC’s implementation of the SIU regime has accomplished what the Government of Canada intended. In fact, all of these reports point to serious problems, both with the legislation, and its implementation. The problems that we have identified are consistent with the findings of others with an interest in the operation of the SIUs, including the Office of the Correctional Investigator (OCI).

We mention the fact that this is the tenth empirically based report on the operation of the SIUs for one simple reason: to make the point that these are conclusions that are grounded in facts. Many of these facts have been known for more than three years. CSC has not, to our knowledge, challenged the validity of this Panel’s findings, nor did it challenge any of the specific findings from the four empirical reports released prior to the creation of this Panel. We believe there is a very good reason for the absence of public criticism from CSC. The findings are based on CSC’s own data. Therefore, CSC ought reasonably to have been aware of these problems without external sources having to point them out. It is also possible that challenging these findings would amplify the attention given to the problems the Panel has reported.

There are at least two important multi-pronged goals of the SIU legislation, the close empirical examination of which demonstrates inadequacies of the SIU regime:

Although on any given day only a small proportion of CSC’s prisoner population are to be found in SIUs, many prisoners have experienced the SIUs during their terms of imprisonment, many are not receiving the minimum of four hours out of cell, and many stays are longer than the 15-day threshold established internationally as the maximum length of time one can be held in solitary confinement (without adequate meaningful contact with other people) before it is considered to be what we are describing as “prolonged” and therefore harmful. Conclusions drawn in this, and in previous reports, reflect the fact that CSC has not met reasonable goals with respect to the SIUs. The problems identified in the previous nine reports, and raised once again in this Annual Report include the following:

These bullet points highlight a few of the flaws of the operation of the SIUs and the flaws in the system of safeguards (e.g., the IEDM model) meant to differentiate the SIU experience from the former use of segregation. This is especially important in those cases in which the SIU experience closely resembles the harmful use of solitary confinement.

It is, of course, especially concerning when some of the flaws we have described here and in our previous reports are more likely to involve vulnerable groups such as those with mental health problems, and Indigenous and Black prisoners.

CSC reports that on any given day there are fewer prisoners in SIUs than was the case when administrative segregation still operated. For example, we received summary information about everyone in an SIU on 23 March 2023. There were 167 prisoners in SIUs that day (plus four others who were recorded as being in a restricted movement cell). CSC has been quite open about the fact that the number of people in SIU cells constitutes only a small percent of the total CSC institutional population. Indeed, it could be argued that CSC sees the size of the SIU population, on an average day, as evidence of the fact that Canada’s inappropriate segregation practices have stopped since the SIU legislation came into force.

However, because our concern is largely with prolonged staysFootnote 6 in SIUs, the drop in the number of admissions to segregation/SIUs, though encouraging, is not as meaningful as is the reduction of prolonged stays in SIUs.

Comparing the number and rate (per 1,000 prisoners) of some of the most “prolonged” stays – those of 61 or more days and those of 120 or more days under the last two full years of administrative segregation (i.e. 2017-18 and 2018-19) to the three years under the new SIU regime (i.e. 2020, 2021 and 2022), it is clear that a reduction in these prolonged stays did not occur until 2022, the third year operating under the SIU regime. Indeed, Table 1 shows that the rate of the longest of these stays in the first two years of the SIU regime (rates of 13.10 and 9.12 per 1,000 prisoners in 2020 and 2021, respectively) is higher than in the last years under administrative segregation (a rate of around 6.5 per 1,000 prisoners). We do not know for certain the cause of the decline in the third year of the operation of the SIUs and it remains to be seen whether this change will be sustained.

Although there was some decrease in the number of overall admissions to administrative segregation beginning in 2015-16, there was no evidence of a decrease in the number of prolonged stays in segregation from 2017-18 onwards (see Table 1).Footnote 7 Hence by the time that the courts in Ontario in 2017 and 2019 and in British Columbia in 2018 and 2019 found constitutional violations that led to the creation of the SIUs, any decrease in the number of prolonged stays in segregation had effectively ended.

Table 1: Prolonged Stays in Segregation and SIUs
Year Number of Stays (Segregation or SIU) that were 61 or more days in length Stays 61 days or more) per 1000 Penitentiary Prisoners (“Rate”) Number of Stays (Segregation or SIU that were > 120 days in length) Stays of 120 days or more per 1000 Penitentiary Prisoners (“Rate”)
2017-18 (Segregation) 469 33.2 92 6.51
2018-19 (Segregation) 550 39.1 92 6.54
2020 437 34.1 168 13.10
2021 391 31.5 113 9.12
2022 313 25.2 55 4.42

Though we might be guardedly optimistic that, in 2022, there was less evidence of prolonged stays in SIUs, we believe that this conclusion should be tempered by three factors:

Prior to Bill C-83, CSC did develop policy guidance that responded to concerns about the segregation of prisoners who had a known serious mental illness, those engaged in self-injury likely to result in serious bodily harm, and those at elevated or imminent risk for suicide. What CSC in one instance referred to as “important changes” were publicized subsequent to the 2007 death of Ashley Smith. These "important changes" were announced only after the following events had occurred:

It would be hard to conclude that CSC is unaware of the many problems identified in this report. Two reasons can be offered for this view:

Specifically, the memo states that (bracketed identifiers added by us):

“There are currently 24 IEDM [Independent External Decision Maker], 11 SDC [Senior Deputy Commissioner] and 44 IH [Institution Head] decisions wherein there has been a decision to transfer the inmate out of the SIU and the decision has not been effected. This is a significant risk to the SIU scheme as well as government and CSC reputation if it is perceived CSC “ignores” the IEDM decisions. I suggest we need to start addressing some of the issues by looking toward more IH [Institution Head] and ADCCO [Assistant Deputy Commissioner, Correctional Operations] accountability – policy says that the ADCCOs are accountable to ensure implementation of IEDM decisions and recommendations, but I am not confident the ADCCOs are always cognizant of this requirement.”

Even though this problem is publicly known and detailed in our first Annual Report, there have been no apparent consequences for not implementing, in a timely fashion, decisions ordering the release of prisoners from the SIUs. The author of the memo appears to have overestimated the risk to CSC of ignoring or failing to implement IEDM decisions.

In interpreting the data in Table 1, we know from our first Annual Report that CSC does, at times, hold prisoners in isolating conditions of confinement without releasing them when ordered to do so. The Correctional Investigator notes that prisoners are sometimes placed in what he has labelled “hidden cells” rather than being released. This suggests that SIU counts may be artificially low. There do not need to be many “hidden” places of isolation in Canada’s penitentiaries to make the numbers look better. A few prolonged stays in hidden isolation cells in each of Canada’s institutions would create an “isolation rate” comparable to what Canada experienced under the former segregation model. Using the Office of the Correctional Investigator’s own words (from its 2021 to 2022 Annual Report),

“At another institution, the pressure to keep SIU numbers low and an increase in incompatible sub-populations (more on this to follow) resulted in the use of “hidden cells” (a term often used by staff at this facility), where prisoners are kept in solitary confinement-like conditions for weeks on end.”Footnote 10

This Panel’s concerns, and what we believe were concerns of the courts that led to the creation of the SIUs in the first place, are with the practice of what can be considered prolonged isolation (or solitary confinement). These concerns cannot be fully addressed by only looking at the SIUs. All isolating conditions of confinement must be brought out of the shadows and subject to robust oversight.

In conclusion, we think it is worth repeating what we said at the beginning: the SIU regime is not working as intended. For that reason, we believe that work should begin immediately to start moving the operation of Canada’s Correctional Service solidly in the direction contemplated by Parliament when the new legislation was passed.

Introduction and Background

Structured Intervention Units (SIUs) were introduced in November 2019 to replace the practice of administrative segregation in Correctional Service Canada (CSC) institutions. At that time, court cases in British Columbia and Ontario had determined that administrative segregation was essentially solitary confinement and, as such, violated Mandela Rule 44, as well as the Canadian Charter of Rights and Freedoms (the Charter) sections 7 (i.e., right to liberty) and 12 (i.e., right not to be subjected to cruel and unusual punishment). Mandela Rule 44 defines solitary confinement as “the confinement of prisoners for 22 hours or more per day without meaningful human contact; prolonged solitary confinement occurs if such confinement exceeds 15 consecutive days”.Footnote 11 As appeals were underway, the federal government passed Bill C-83 which amended the Corrections and Conditional Release Act (CCRA) to eliminate disciplinary segregation and replace administrative segregation with SIUs. The intention was that transfers to SIUs would be a last resort and used “only if… there is no reasonable alternative…” (CCRA, s. 34.1); prolonged confinement would be avoided and prisoners would be offered a minimum of four hours out-of-cell per day, two hours of which were to involve “meaningful human contact”. The CCRA amendments also created independent external oversight in the form of Independent External Decision Makers (IEDMs) (CCRA s. 37.6(1)). The government hoped that Bill C-83 would bring CSC’s practices in line with the Mandela Rules and transform federal corrections and enhance the care and wellbeing of Canadian prisoners.

The court cases that led to the amendment of the CCRA and the establishment of SIUs had been launched by the British Columbia Civil Liberties Association and John Howard Society of Canada, and the Canadian Civil Liberties Association in Ontario. Considerations in these court cases and the subsequent legislation included the need for external review of decisions pertaining to prisoners’ confinement in SIUs. This led to the creation of two bodies intended to provide oversight to the implementation of SIUs and CSC decision-making with regard to SIUs: the aforementioned IEDMs and the non-statutory Structured Intervention Unit - Implementation Advisory Panel (SIU-IAP). The IEDMs are charged with reviewing SIU cases and making recommendations related to the conditions and duration of confinement. IEDMs have binding powers to order that a prisoner’s period in the SIU should end. The SIU-IAP has a mandate to advise the Minister of Public Safety and make recommendations to the Commissioner of Corrections in regard to the operation of SIUs.

The SIU-IAP was established in mid‐2019. Following CSC’s failure to provide data to the SIU-IAP, the original Panel’s one-year mandate ended without it being able to access data about the operation of the SIUs. Data on the actual operation of the SIUs were eventually made available to two independent researchers in late September 2020 after the first Panel ceased to exist. The first report on the operation of the SIUs by these researchers was released in mid-October 2020.

The current Panel was created in April 2021 and started with a two-year term. In June 2023, its mandate was extended to December 31, 2024, which will allow the final work of the Panel to be considered as part of the mandated 5-year review of the legislation that created SIUs. Over the past 3.5 years, the Panel has publicly released five reports outlining its findings. These five empirically based reports identify critical problems with the implementation of SIUs. Panel members have visited all of the SIUs (both male and female units) in every region of the country. These site visits have included meetings with administrative, operational, and professional staff, including health care practitioners and Indigenous Elders, as well as interviews with some SIU prisoners. The Panel also held several meetings with senior CSC officials, including the Commissioner. In addition, Panel members have conferred with a number of stakeholders with an interest in and knowledge of the SIUs.

It should be noted that the Minister has not yet appointed a Panel member with lived experience as a prisoner. While not specifically provided for in the Panel’s Terms of Reference, such appointment was discussed and agreed to as an important part of re-constituting the Panel in the Spring of 2021.Footnote 12 Several potential appointees were referred to the Minister for consideration and on December 16, 2021, one potential member was invited to submit a CV and complete the security clearance process.

Regrettably, there is little evidence that Bill C-83 has transformed federal corrections. Key aspects of the legislation remain poorly defined or operationalized in a way that undermines a full assessment of SIU implementation and performance. For example, it remains unclear what qualifies as either “meaningful human contact” or “structured intervention”.

Problems with the SIU System: A Case History

To set the context for this report, we would like to introduce you to Lawrence, (not his real name), a long-serving federal prisoner with experience in both the former segregation regime and the current SIU regime.

We have not included a full record of the time Lawrence has spent in custody to protect his privacy.

Lawrence is a middle-aged white male with a youth justice record and a significant history of mental health concerns, including self-injury. He has spent most of the past two decades within the federal correctional system. His first federal sentence began in the early 2000s and ended with his release at Warrant Expiry nearly a decade later. Within weeks of his release Lawrence committed the offence that led to his second federal sentence. We ended our review of this case in early 2023.

During his first sentence, Lawrence was involved in 190 “institutional incidents”, 70 of which were identified by CSC as “serious”. He was placed in administrative segregation 52 times and spent over 1,600 days in these isolating conditions. He was housed in isolating conditions of confinement in all five of the Service’s regions. A number of incidents of self-injury occurred resulting in transfers to Regional Treatment Centres. During this sentence, he was released, unsuccessfully, twice on Statutory Release (SR), spending less than 4 months total in the community.

During his second sentence Lawrence was involved in more than 80 “institutional incidents”, 30 of which were identified by CSC as “serious”. He was placed in administrative segregation/SIUs over 30 times and spent in excess of 500 days in isolating conditions of confinement. He was transferred into isolating conditions in four of the Service’s five regions. Numerous incidents of self-injury occurred resulting in transfers to a Regional Treatment Centre. Lawrence was released, unsuccessfully, twice on SR, spending less than 8 months total in the community.

Lawrence has spent nearly two decades under federal sentence, mostly in custody interspersed with a number of unsuccessful periods of community supervision. He has displayed a behavioural pattern over this time period that has remained remarkably consistent, as has the response of the Correctional Service of Canada.

His time in an open or general population in a penitentiary was usually short lived, chaotic, and stressful. At times, his movements to administrative segregation/Structured Intervention Units were initiated at his own request; at other times it was the result of conflicts with other prisoners or staff, but more often than not, it was orchestrated by Lawrence himself through threats against others or acts of self-injury. Once placed in administrative segregation or an SIU he seldom appeared in a hurry to return to open population, often refusing to leave isolation or opting to remain until his Statutory Release date or until arrangements for an inter-regional transfer were finalized.

When we compare the first part of the second sentence to the second part, (i.e. pre- and post-SIU implementation), we saw no measurable difference in either Lawrence’s behaviour or the Service’s effectiveness in responding to him.

During the first part of this sentence, he was incarcerated for approximately 1,350 days, was placed in administrative segregation 19 times, and spent 240 days in isolation. There were multiple incidents of self-injury resulting in transfers to Regional Treatment Centres.

During the second part of this sentence, he was incarcerated for approximately 900 days, was transferred to the SIU 8 times and spent 256 days in isolation. There were multiple incidents of self-injury resulting in transfers to Regional Treatment Centres.

We found no evidence through this review, that the introduction of the SIU regime has enhanced the Service’s ability to effectively manage complex and high-needs prisoners such as Lawrence. The Panel is not particularly surprised by this finding, as CSC’s current decision-making approach to the SIU is narrowly focused on policy compliance, rather than addressing the underlying issues and circumstances that led to SIU transfers.

This failure to address underlying issues extends most noticeably to the management of prisoners with mental health concerns. The Service’s response to Lawrence over two decades reveals a cycle of placing him in isolation, followed by transfers to Regional Treatment Centres. It is evident that little-to-no progress has been made in the development of intervention and treatment programming designed to keep prisoners like Lawrence out of isolating confinement. The SIUs would appear to continue to be the placement option of choice for far too many prisoners with significant mental health concerns.

The expected positive impact of the Independent External Decision Maker (IEDM) on SIUs, specifically in terms of limiting the length of stays, was not apparent in this case. Nor were the internal National and Regional reviews, as currently conducted, of any assistance to the Institutions in effectively managing this difficult case. It should be noted that five of Lawrence’s SIU stays were 30 days or more, with two exceeding 60 days, whereas only one of his nineteen Administrative Segregation placements, in the three years prior to the implementation of SIUs, exceeded 30 days. This hardly looks like progress.

How Prisoners Experience an SIU

Canada’s penitentiaries have been challenged to eliminate the practice of isolating conditions of confinement (administrative segregation) since 2019 when Parliament required the creation of “places” to be known as SIUs. Part of the difficulty relates to the wording of the requirements in the Corrections and Conditional Release Act(CCRA). The companion problem relates to the manner in which Correctional Service Canada (CSC) has implemented this new practice. From a prisoner’s perspective, it matters little whether the challenges arise because of the legislation or because of its implementation. What is important is that prisoners receive treatment that approximates what Bill C-83 was drafted to accomplish as it responded to Charter violations identified by Canadian courts.

Length of Stay

The definition of an “acceptable” length of stay in isolating conditions of confinement should not be arbitrary. The dividing line internationally is 15 days pursuant to the Mandela Rules. The amended CCRA that created SIUs avoided the internationally accepted specificity in two ways. First, by stating in the legislation that “(a)n inmate’s confinement in an SIU is to end as soon as possible” (s. 33, CCRA), the CSC is permitted to determine the time frame by defining what is “possible” in their own terms. The second way in which Canada avoided a specified time limit was to put in legislation that there shall be “an opportunity to spend a minimum of four hours outside the inmate’s cell” (s. 36 (1)(a), CCRA) including two hours of interaction with others (s.36 (1)(b)). Setting a minimum of four hours out of cell (including 2 hours of meaningful interaction with people), it was argued, meant that this was no longer “solitary confinement” which has an international definition of two hours or less out of cell with others. But we know that legislating opportunities to be out of cell is not the same as ensuring prisoners are actually spending time out of their cells in meaningful contact with others. We will return to hours out of cell but will first examine the length of SIU stays.

Looking at the SIU stays that began in each of the 3 complete years for which we have data and allowing CSC some leeway in achieving the “15 day” limit, the data in Table 2 examine the proportion of SIU stays that are more than one month – twice the international standard.

Table 2: Length of Stay in the SIU
Year SIU stay began Number of Days in the SIU Total % Total Stays
1 through 31 32 through 552
2020 64.8% 35.2% 100% 2,152 stays
2021 53.7% 46.3% 100% 1,393 stays
2022 54.5% 45.5% 100% 1,424 stays
Total 58.7% 41.3% 100% 4,969 stays

There are two simple points made by the data in Table 2. First, the total number of person stays did decrease over this three-year period. Second, the proportion of stays over 32 days appears to have increased since 2020. In other words, a substantial number of prisoners had SIU stays that were considerably longer than the international standard. To express this in numbers, during these three years, there were 2,051 person stays in SIUs that were 32 days or longer. In the third year of operation of the SIU regime, there were 648 stays in SIUs that were 32 days or more in length, a number that is lower than the number for 2020. A focus on this apparent decrease misses the point. We are not willing to accept the implicit implication of these data: that allowing 45.5% of the person-stays in 2022 (or 648 actual person-stays) in the SIUs to exceed 2 months in length reflects the best possible outcome that we can expect in Canada’s penitentiaries.Footnote 13

There is a third issue that has repeated itself in almost everything we examined in relation to CSC’s implementation of SIUs: there were large regional differences in the way the SIUs operate (Note that only 2 of the 5 regions have multiple SIUs within region. Hence institutional variation within region is not relevant in all regions). The findings presented in Table 3 only examine length of stay (in the 5 regions) for stays in the SIUs that began in the year 2022 (the most recent year for which we had complete data when this report was being prepared). We see that the length of stays varied substantially across regions. Quebec, for example, had 22.2% of its SIU stays lasting 32 days or more whereas the Pacific region had 65.7% of its stays lasting that long.

Table 3: Length of SIU Stay in 2022 by Region
Region Number of Days in the SIU Total % Total Stays
1 through 31 32 through 552
Atlantic 47.1% 52.9% 100% 172 stays
Quebec 77.8% 22.2% 100% 419 stays
Ontario 54.3% 45.7% 100% 243 stays
Prairies 42.8% 57.2% 100% 409 stays
Pacific 34.3% 65.7% 100% 181 stays
Total 54.5% 45.5% 100% 1,424 stays

Is it reasonable that there is such variation? If long stays are to be avoided, what do these data tell us? These are not new findings, and CSC has so far not provided substantive explanations.

There appears to be significant differences across racial groups in the length of SIU stays. While there have been slight changes over time in the length of stays for various groups, most recently, in 2022, it appears (see Table 4) that Indigenous prisoners are more likely to have prolonged stays in an SIU of 91 days or more (11.7% stayed 91 days or longer compared to 8.4% of Black prisoners and 5.8% of White prisoners). We do not know why this has occurred nor has CSC suggested an explanation.

Table 4: Length of SIU Stay in 2022 for prisoners identified as Indigenous, Black or White
Group Number of Days in the SIU Total % Total Stays
1 through 31 32 through 90 91 through 552
Indigenous 49.4% 38.8% 11.7% 100% 605 stays
Black 59.1% 32.5% 8.4% 100% 203 stays
White 58.7% 35.6% 5.8% 100% 520 stays

Note: Data for other racialized groups and for those with missing personal identifiers are not included. They constituted 6.7% of person stays in 2022 (N=96 of 1,424 stays during 2022).

Finally, there is one other aspect of the “length of stay” estimates that cannot be ignored. One way in which the “length of a stay” can be difficult to assess is when stays are “split”, with one transfer closely followed by another transfer back to the same or a different SIU. This can have significant implications for timely reviews and recalls some of the abusive use of administrative segregation. Recall that the courts found the indeterminate nature of administrative segregation to be a violation of the Charter.

Early in 2023, this Panel received data from CSC on all those prisoners who were in SIUs on 23 March 2023. Multiple stays in multiple SIUs were common, just as multiple stays in a single SIU occurred quite frequently. One prisoner – whom we shall call André – experienced both.

André had three stays with the following dates. They were all in the same SIU.

Though arguably the second stay – separated by 11 days from the first stay – should count as a separate stay, it is hard to see how the third stay listed here is distinct from the second. André left the SIU on 28 February 2020 and returned to the same SIU later that same day. By splitting the second and third stays into two, André would be said to have experienced 59 days in an SIU in the second stay (and therefore would not have the length of this stay reviewed by an IEDM). The third SIU stay was for 48 days (28 February 2020 to 16 April 2020). Again, André’s case would not have been referred to an IEDM to review the length of time he was in the SIU. Had he not been released for a “re-start” of his SIU “clock” on 28 February 2020, André would have shown up as having stayed in an SIU for 107 days, in which case his file would have been – or should have been – brought to the attention of an IEDM for review.

But André had other experiences that illustrate this same problem. Consider the following record (Table 5) of his SIU experiences starting a bit later that same year (2020). In order to protect prisoners from being identified, we have, when describing individual experiences of prisoners, described institutions as being “East” (Ontario and eastwards) or “West” (Manitoba and westwards) and have given numbers to individual institutions within each of these two parts of Canada in the order in which they appear in a given table.

Table 5: André’s later experience with SIUs 2020-2023
SIU Stay # SIU Start Date SIU End Date Institution of SIU
6 16 Dec 2020 20 Dec 2020 West-One
7 22 Dec 2020 21 May 2021 West-One
8 23 May 2021 29 Jun 2021 West-One
9 18 Aug 2021 16 Mar 2022 West-Two
10 20 Mar 2022 27 Apr 2022 West-Three
11 29 Apr 2022 13 Jun 2022 East-One
12 14 Jun 2022 05 Oct 2022 West-Three
13 08 Oct 2022 21 Nov 2022 East-Two
14 13 Feb 2023 [still in 23 Mar 2023] West-Two

André had 3 SIU stays in West-One (Stays 6,7,8) that, given the time between the end date of one and the start date of the next look remarkably like one long stay with some very short (2 day) breaks inserted. We don’t know what the purpose was of those two 2-day breaks between his SIU stays. However, starting at Stay #9 and ending with Stay #13, André had 5 SIU stays in 4 institutions (across 3 regions) with only a few days between each stay. Once again, a reasonable inference would be that these were all part of one SIU stay that included a cross-country tour of SIUs. That “consolidated stay” ran approximately 15 months.

André’s may be an unusual case in terms of the number of SIU stays he had and the timing of these stays; however, the problems that we see in his history – most notably the splitting of what is effectively, from the prisoner’s perspective, almost certainly a single SIU experience split up for administrative reasons into multiple stays – may not be unusual. We found, for example, that of all the people who found themselves in an SIU on 23 March 2023, 30% had experienced a “split stay” with four days or fewer between SIU stays.

There is another serious problem that must be addressed. For some prisoners – especially those in the SIU for their own safety – the SIU is perceived to be a better place to serve one’s sentence than in a cell in the mainstream population. We have repeatedly heard from CSC that prisoners do not want to leave the SIU for various reasons. However, prisoners’ preferences to remain in SIUs are not adequate justification for long stays in SIUs. Instead, it suggests that CSC should look carefully at its operation of the mainstream parts of its institutions to ensure that all prisoners have a reasonable likelihood of being able to safely serve their sentences outside of SIUs. In other words, it does not reflect well on CSC or Canada if a portion of Canada’s prisoners with complex needs, including the need for protection, prefer conditions of isolation and confinement to being in the mainstream population of a penitentiary. A prisoner should not have to forfeit retained liberty rights for their personal safety.

Our conclusion from the data is simple and clear. CSC has not achieved one of the key goals of the legislation: keeping the length of stays reasonably short. The data presented in Table 2 clearly demonstrate no appreciable improvement in reducing the length of stays over this three-year period.

Time Out of the Cell

The second (two part) promise of Bill C-83 was that prisoners would be offered adequate time out of cell – adequate to differentiate a stay in an SIU from the internationally defined practice of solitary confinement. This was expressed in the Bill as being a minimum of 4 hours out of cell each day, including “an opportunity to interact, for a minimum of two hours, with others, through activities” (CCRA s. 36(1)). Despite the comments of the Minister at second reading of Bill C-83 (see Appendix “B”) a key ambiguity of this promise is that prisoners are to be “offered” an “opportunity.” By implication, there is no requirement that a prisoner actually spend four hours out of their cell (including 2 hours of interaction with others.) We heard a number of explanations from prisoners about why they did not accept offers from CSC, including complaints about the physical setting that they were offered (e.g., a barren exercise yard on a cold day in winter), a lack of engaging activities, or personal safety concerns related to the social setting of the offer (e.g. with another prisoner whom the SIU prisoner was afraid of).

Time out of cell is particularly important for those who have long stays in the SIU. Hence in this report, we will often concentrate on long stays, in this discussion meaning those that are 16 or more days.

We have no doubt that providing prisoners with choices sufficiently attractive to make them want to accept an offer of time out of cell can be a challenge. It should be remembered that CSC was given substantial and permanent extra fundingFootnote 14 to have an enhanced staffing ratio and enhanced programming in the SIUs. This has not translated into achieving the desired time out of cell for SIU prisoners. In addition, the data show very clearly that “refusals” by prisoners to leave the SIU cell do not adequately account for the failure of prisoners to get the legislated minimum time out of cell.

For example, we looked at stays of 16 days or more and found that during the three-year period (person-stays starting between 1 January 2020 and 31 December 2022) there were 1,485 people identified who did not get their four hours out of cell on most days (i.e. 76% or more of their days in an SIU).

For 46% of these person-stays (689 person-stays) the prisoner refused no more than twice during these 16 or more days. But these prisoners did not receive their four hours out of the cell most of the time. Logically, if the prisoner did not refuse and the prisoner did not get out of their cell, it must mean that CSC did not offer the full four hours to the prisoner or CSC failed to properly record the operation of the SIU. However one interprets the “refusal” data, the conclusion (according to CSC’s own data) is clear: Many of these relatively long stay prisoners did not have the time out of cell that one might expect after having heard the Parliamentary debates on Bill C-83 or after having read the resulting legislation.

In any case, although the law indicates that prisoners must be offered time out of their cells, when they don’t leave their cells, it is a problem for the prisoner and society because it results in circumstances that led the courts to conclude that the treatment of prisoners did not conform to constitutional guarantees. Since most of these prisoners are going to be released back into our communities, long stays in isolating conditions are clearly not the best preparation for successful reintegration.

Looking at SIU stays that lasted 16 consecutive days or longer (see Table 6), we see that across the three years, in more than half of these SIU stays (52.7%), prisoners did not receive their four hours out of their cells on at least 76% of their days in the SIU.

When we look at another measure related to the prisoner’s experiences in an SIU, we see that close to half of the prisoners (45.8%) failed to receive their two hours out of the cell involving meaningful human contact on most of their days in the SIU.

Table 6: Long stays (16+ days) only. Percentage of person-stays in which the prisoner did not get 4 hours out of cell on most days, and the percentage who did not receive 2 hours of meaningful human contact on most days
Year Percent who did not receive 4 hours out of cell on most days (76% or more of their days in the SIU) Percent who did not receive their 2 hours of meaningful human contact on most days (51% or more of their days in the SIU) Number of person-stays
2020 63.2% 52.0% 1,076
2021 46.9% 43.8% 864
2022 45.5% 40.2% 880
Total 52.7% 45.8% 2,820

It could be suggested that between 2020 and 2021, CSC improved its ability to get people out of their cells. As shown in Table 6, the portion of (long stay) prisoners who did not get either type of hours out of the cell decreased between 2020 and 2021. Improvement is obviously to be applauded, but it is difficult to applaud “improvement” when, for example, the “improved performance is that “only” 45.5% did not get a benefit (4 hours out of cell) on most of their days during a long stay in the SIU. For many hundreds of prisoners, then, it is correct to describe their SIU experience as an experience that mirrors solitary confinement.

After that (2021 vs. 2022), progress stalled. However, there is dramatic regional variation and for some regions, no consistent improvement. For example, in the Pacific region in 2020, 89.2% of (long stay) prisoners did not receive their four hours out of cell in 76% or more of their days. In 2021 this declined to 71.5% and then increased to 80% in 2022.

The Experience of Solitary Confinement

Given that time out of the cell is the SIU feature that is meant to differentiate the SIU experience from the experience of administrative segregation, the importance of these findings cannot be ignored. Indeed, overall, for these (long stay) prisoners who did not receive their four hours out of cell on 76% or more of their days (N=1,485), 41.4% had, on average, up to one hour out of cell and another 51.6% received, on average, between one and two hours out of their cell – both of which reflect the practice of solitary confinement as defined internationallyFootnote 15. And to blame the prisoner – by saying that the prisoners did not “avail” themselves of the opportunity that was (sometimes) offered is beside the point. CSC has a positive obligation to ensure that prisoners do not experience long periods of isolating conditions of confinement. Furthermore, considering the worst-off person stays, we see in Table 7 that the proportion of these prisoners did not appreciably drop over the 3-year period. Specifically, Table 7 looks at the person stays with the following three characteristicsFootnote 16:

Table 7: Proportion of SIU person-stays in which the prisoner experienced three aspects of a solitary confinement
Year Percent of all person-stays that were 16 or more days long where the prisoner did not receive 4 hours out of cell on most (76%+) days, and didn’t receive their 2 hours of meaningful human contact most (51%) of the time Number of person-stays
2020 24.7% 2,152
2021 24.1% 1,393
2022 20.6% 1,424
Total 23.4% 4,969

Regional variation persists, and for some regions there is no improvement. For example, in the Atlantic region in 2020, 22.2% of prisoners stayed for 16 days or more and did not receive both their four out of cell and two hours of meaningful human contact. This proportion of prisoners declined to 19.4% in 2021 and then increased to 31.4% in 2022.

Finally, we have a set of findings that should be a starting point for CSC to improve its performance. Since the first external report on the SIUs was released three years ago in October 2020, it has been well established that there are regional and institutional differences in how CSC delivers on legislative intent in the areas of length of stay and time out of cell.

We have noted that Stony Mountain Penitentiary in Manitoba reports very few person-stays in which the prisoner is in the SIU for a long period of time, and low rates of prisoners who do not receive most of their time out of cell and most of the two-hour human contact hours. While under 2% did not receive their two and four hours on most days in Stony Mountain, roughly a quarter or more did not receive their hours in other institutions in the same (Prairie) region and in other regions (see Table 8).

Table 8: An example of institutional variation in experiencing solitary confinement in SIUs (2020, 2021, and 2022)
Institution Prisoner group Percent of all person-stays that were 16+ days and where prisoner did not receive, on most days, getting their 4 hours out of cell and their 2 hours of human contact.Table note i Total number of person-stays
Stony Mountain Non-Indigenous 0.8% 119
Indigenous 1.5% 394
Other Prairie SIUs Non-Indigenous 25.5% 345
Indigenous 22.8% 534
SIUs in the other four regions Non-Indigenous 24.7% 2,418
Indigenous 29.9% 1,159
Total 23.4% 4,969
Table note i

This index uses the same definitions as were used in the previous table.

Return to table note i referrer

Table 8 illustrates that it is possible to do much better than CSC’s overall performance. At first, when we saw the findings related to Stony Mountain, we thought that it might have something do with differential management of the Indigenous prisoners in this facility. When we looked at Indigenous and non-Indigenous prisoners separately, however, we see that each group fares much better (on this combined measure) in Stony Mountain than prisoners in other facilities in the Prairie Region and elsewhere in Canada. The question that these data demand an answer to is a simple one: if it is possible to perform considerably better in one institution, why can’t other institutions meet this standard?

It bears repeating that CSC has a positive obligation to achieve the legislative goal of the SIU regime: that prisoners should spend at least the minimum time out of cell established in law, and that part of that time must involve meaningful human interaction. This Panel has not conducted a comprehensive study of why certain facilities are considerably more able to achieve the goals of the SIU legislation than others. We believe this responsibility falls squarely upon the shoulders of CSC. In response to questions from the Panel, CSC officials told us that they have held town halls to solicit opinions about regional and institutional differences like those described above and have invited community organizations to become involved in offering meaningful activities to SIU prisoners. It is this Panel’s view that these efforts are not sufficient. Some of CSC’s own institutions have demonstrated that it is possible to come much closer than most of its other institutions have to fulfilling the intent of the legislation. CSC must do the work necessary to improve its record in providing SIU prisoners time spent out of cell and must then immediately implement the changes required to achieve performance that will clearly differentiate SIUs from administrative segregation and the abuses of solitary confinement.

CSC’s Approach to Dealing with Challenging Prisoners

Publicly available information from CSC claims that SIUs “provide more structured and effective interventions to inmates that address their specific needs and risks, with the goal of facilitating their reintegration into a mainstream inmate population as soon as possible.”Footnote 17 This goal is incorporated into one of CSC’s “Key Performance Indicators” which is the percent of transfers from an SIU that result in the prisoner remaining in the mainstream population for 120 days. Said differently, CSC’s interpretation of its mandate regarding SIUs is that they have an obligation to help prisoners in SIUs change such that they can be reintegrated to the mainstream population without returning to the SIU.

On 23 March 2023, the arbitrary date of the data extraction by CSC, there were 167 prisoners in an SIU. For 125 of these prisoners (75%) this was not their first time in an SIU. If the “interventions” that take place in the SIU are meant to keep prisoners from returning to the SIU, it was not successful in these 75% of cases. The Panel acknowledges that some prisoners are difficult, and that those who have been in an SIU once are more likely than others to return to an SIU, therefore the issue of repeat visits isn’t necessarily a failure.

What is of greater concern, however, is that of these 125 prisoners with more than one SIU stay, most (77%) had been transferred into two or more different SIUs. In fact, the most common pattern for those who had experienced multiple SIU stays was that they had been placed in different regions for their stays (73 of the 125 prisoners, or 58% of the multiple SIU stay prisoners that day). Not surprisingly, then, most of these prisoners had stayed in multiple SIUs in more than one of CSCs five regions. Indeed, 9 prisoners had experienced life in an SIU in all five of CSCs regions.

As noted above, we are not suggesting that CSC has failed because some prisoners are not transformed with a single stay in an SIU. However, we do find it difficult to understand how the instability and unpredictability of transfers to another SIU, quite often in different regions, makes correctional sense. Of note, prisoners are supposed to continue with their correctional plans while in an SIU, but it is difficult to see how correctional plans can be implemented in an effective manner with this degree of disruption and lack of continuity. Furthermore, we know that a significant proportion of SIU prisoners have identified mental health problems. Therapeutic relationships are difficult to forge in a prison, and it seems unlikely that frequent transfers would make this any easier.

While tracking these 125 “multiple SIU stay” prisoners, we noticed something else. Prisoners identified as being Black were more likely than other prisoners to experience multiple SIU stays in different regions. Of the prisoners with multiple stays, 77.8% of the Black prisoners experienced their stays in different SIUs in different regions compared to “only” 58.3% of Indigenous prisoners and 53.1% of white prisoners. Once again, we have no explanation for this finding.

Of the 125 prisoners in SIUs on 23 March 2023 with two or more SIU stays (since November 2019), 40% (N=50) had one or more split stays – stays that were counted as multiple stays but only had four or fewer days between each stay. Black prisoners were more likely than members of other groups to experience “split stays”. Of those with multiple SIU stays, 55.6% of Black prisoners experienced a “split stay” compared to 38.3% of Indigenous prisoners and 31.3% of white prisoners. As before, we have no explanation for this finding. If stability in the delivery of programs and treatment is important, then the uncertainty and/or unpredictability of stays that are split (often involving moving to another institution or another region) appear to us to be at worst an obfuscation and, at least, counter productive.

Looking more broadly at the administrative data we have on SIU prisoners, there were 2,257 different people who began an SIU stay between 1 January 2020 and 31 December 2022. Almost half of them, during this period experienced multiple stays (see Table 9).

Table 9: Prevalence of multiple SIU stays among those prisoners with at least one SIU stay
Prisoners entering SIUs one or more times – 1 January 2020 to 31 December 2022 Number Percent
Only 1 SIU stay 1,176 52.1
Multiple stays in one SIU in one region 474 21.0
Multiple SIU stays in different SIUs, but same region 201 8.9
Multiple SIU stays in different SIUs in different regions 406 18.0
Total 2,257 100.0

Note: Of those with multiple stays (N=1,081), 37% (N=406) spent their stays in different SIUs in different regions.

The data suggest that CSC continues to manage difficult and challenging prisoners with inter- and intra-regional transfers. However, we doubt very much that the movement of a prisoner to SIUs in different regions is part of their correctional plan or that it is helpful in resolving the underlying issues that make them “difficult”. In the days of administrative segregation, transfers of difficult prisoners to other institutions and across regions were frequent. Offloading challenging cases has contributed to tragic outcomes (e.g., the death of Ashley SmithFootnote 18). Another aspect of the concern with all this movement is the operation of the policy that governs transfers. Institutional managers report that the approval process is “constipated” because there are so many requests. This frustrates attempts to implement timely transfer decisions to move a prisoner who is ready to integrate into the general population out of an SIU.

Two separate but related issues need to be examined more closely. First, the issue of multiple SIU stays where prisoners are frequently moved from one location to another requires close attention. Second, more generally, the involuntary movement of any prisoner from one region to another raises its own set of issues. This latter issue, however, needs to be examined across CSC, and not just in the context of the SIUs.

Mental Health and SIUs

It is not surprising that there is a relationship between mental health and time spent in the SIU. It seems likely that this relationship flows in both directions; specifically, those prisoners facing mental health challenges may be more likely to be transferred into an SIU, and those experiencing confinement in an SIU may be more likely to develop or increase the severity of their mental health problems.

The CSC website, with respect to prisoners in SIUs, claims that:

“Each individual is referred to health services for a health assessment, including mental health, within 24 hours of being transferred to an SIU. Individuals in SIUs are seen daily by a health care professional to monitor their health care needs. Individuals in SIUs are offered health and mental health assessments throughout their stay in the SIU” (emphasis in the original).Footnote 19

Though it is unlikely that a prolonged stay in an SIU would be recommended as treatment by a competent mental health professional, it is noteworthy that those who are identified by CSC as experiencing deteriorating mental health status are considerably more likely than others to be held in SIUs for very long periods of time (Table 10). Importantly, CSC specifically states that SIUs are not a treatment mechanism.

Table 10: Mental health status (as recorded by CSC) and the likelihood of a prolonged stay of 62 days or more in the SIU (2020, 2021, and 2022)
Mental health status Proportion of those with these mental health needs who have SIU stays of 62 days or more
Low or some mental health needs, but not getting worse 21.7%
High mental health needs, but not getting worse 17.5%
Various mental health needs and getting worse 40.5%

The Panel was unable to determine whether the prolonged length of time for those with deteriorating mental health was a result of CSC not finding interventions more appropriate for a prisoner with deteriorating mental health, or if there is a belief that extra time in an SIU is the only option available for people with deteriorating mental health.

In any case, Table 10 demonstrates that 40.5% of those with mental health problems that were getting worse (as assessed by CSC itself) were kept in the SIU for 62 or more days. When one looks at those with low or some mental health problems but who were not assessed as getting worse, we see that “only” 21.7% are held in the SIU for 62 days or more.

If deteriorating mental health status is in part related to the experience of isolation in an SIU cell, then it is unacceptable that this group is especially likely to have long stays (62 days or more) in an SIU.

We have no data on why CSC keeps this particularly vulnerable group in SIUs for very long periods of time. This data is particularly troubling considering the courts’ determination that the transfer of mentally disordered prisoners into administrative segregation was highly problematic.

In our data sample, 56% of those whose mental health status was deteriorating and who were kept in an SIU for 62 or more days were Indigenous prisoners. This can be compared to the fact that “only” 42% of all of person-stays in the SIUs involved Indigenous prisoners. This proportion is higher than the proportion of Indigenous people in Canada’s penitentiaries - 32% in 2020/2021). Said differently, those with deteriorating mental health status are disproportionately Indigenous and are disproportionately likely to have prolonged stays of 62 days or more in SIUs. Detailed data (released in our early 2023 mental health update) show that this effect is evident across years and can be seen in each of CSC’s 5 regions.

In the context of questions about mental health and SIU prisoners, the concern about the treatment of prisoners is especially acute if they experienced the “double whammy” consisting of stays with the following characteristics:

39% of those with mental health needs (as identified by CSC) experienced these circumstances, compared to “only” 30% of those without mental health needs (as identified by CSC).

Those who, during the three years that SIUs have been operating, have experienced multiple stays in SIUs are considerably more likely to have been identified by CSC as facing mental health problems compared to those with only one stay. As shown in Table 11, of those with one SIU stay, 25.2% had an indicator that there were mental health problems compared to 59.3% of those with five or more SIU stays. These findings become even more troubling when considered in the context of CSC’s very deliberate statements that SIUs are not mental health units, and those prisoners that require mental health care and treatment will receive appropriate care and treatment in their parent institutions or the Regional Treatment Centres (RTCs).

Table 11: Percentage of prisoners identified as having a mental health problem with one or more SIU stay (2020, 2021, and 2022)
Number of SIU stays Was there a flag during any SIU stay indicating that the person had mental health problems? Total (row) percent Number of people with this many person-stays in SIUs
No Yes
One 74.8% 25.2% 100% 1,073
Two 67.4% 32.6% 100% 399
Three 60.5% 39.5% 100% 243
Four 57.6% 42.4% 100% 118
Five or more 40.7% 59.3% 100% 182
All prisoners who entered an SIU at least once 67.5% 32.5% 100% 2,015

In Table 12, we see that those with “considerable or higher” mental health needs are most likely to experience a “treatment” that we discussed earlier - being placed in multiple SIUs in different regions. As we suggested earlier, it may well be that transfers to a different SIU or region are seen as a way of removing a challenging prisoner from a particular institution. It is hardly surprising, therefore, that those with serious mental health challenges are most likely to receive this “treatment”.

Table 12: Mental health needs and the prevalence of multiple stays in SIUs (2020, 2021, and 2022)
New measure of mental health issues. At some point during SIU stay, mental health needs were assessed as: Description of SIU stays Total % Total stays
Only 1 SIU stay multiple stays in one SIU in one region multiple SIU stays in different SIUs, but same region multiple SIU stays in different SIUs in different regions
None 58.3% 19.8% 8.3% 13.7% 100% 1,509
Some 44.1% 24.4% 8.2% 23.2% 100% 487
Considerable or higher 31.4% 21.5% 13.8% 33.3% 100% 261
Total 52.1% 21.0% 8.9% 18.0% 100% 2,257

Section 116 of Commissioner’s Directive 711Footnote 21 states that:

Everyone working with inmates will:

  1. work collaboratively with health care to ensure health needs of the inmate are appropriately accommodated based on the assessments of registered health care professionals.
  2. refer, in accordance with relevant policy and institutional Standing Orders, the case of an inmate to a registered health care professional when the staff member believes that confinement in an SIU is having detrimental impacts on the inmate's health (…).
  3. inform the Institutional Head of the referral to the registered health care professional.

CSC’s Regional Treatment Centers (RTCs) are relevant in the context of a discussion of the mental health of prisoners. For the 167 prisoners in SIUs on 23 March 2023, we were given data on their “institutional history” relating to the previous 2-3 institutions that they had been housed in.Footnote 22 Thirty of these 167 prisoners had spent at least some time in one (or more) RTCs at one of their previous 2-3 institutions.

Below, we review two randomly chosen records taken from these 30 prisoners. Note that without examining medical records, it is difficult to know exactly what the relationship is between the SIUs and the RTCs and that we did not review the full institutional history of all 30 prisoners.Footnote 23

The first is Pierre (not his real name), a Black male in his 20s, admitted to CSC in March 2016 with an indeterminate sentence. His institutional movement is shown in Table 13.

Table 13: Pierre’s experience in the SIU
Step Number Institution Start Date End date Mental health challenges SIU experiences (minimum standards)
1 SIU 28 June 2021 20 Sept 2021 None Prolonged stay, failed to get minimum standards on 2 & 4 hours
2 SIU 29 Nov 2021 3 Dec 2021 None Relatively short stay
3 SIU 14 Jan 2022 14 Feb 2022 None Prolonged stay, didn’t get 4 hours, but got the 2 hours
4 SIU 14 June 2022 10 August 2022 Some Prolonged stay, didn’t get 4 hours, but got the 2 hours
5 SIU (East-1) 9 January 2022 12 January 2023 considerable Relatively short stay
6 Regional Treatment Centre, East-1 12 January 2023 30 January 2023
7 SIU East-1 30 January 2023 27 March 2023 considerable Prolonged stay, failed to get minimum standards on 2 & 4 hours

In terms of understanding the relationship between the SIUs and the RTCs, it is perhaps easiest to start from Step #7 and look at what happened before then. At Step #7, Pierre spent about 56 days in an SIU. CSC determined that he suffered from considerable mental health challenges. During that stay, he did not receive anything close to the minimum of 4 hours out of cell and two hours of meaningful human contact. Indeed, most of his time spent in SIUs would meet the international definition of solitary confinement (two hours or less out of cell).

Pierre had, however, been transferred directly from a RTC (where he had spent 18 days) to the SIU. Prior to that, he had 5 stays in SIUs, though his mental health problems weren’t noted until the fourth stay. It is difficult to speculate why the East-1 SIU (at Step 7) was the best place to put Pierre.

In another case, George (not his real name), a First Nations male in his early 20s serving a sentence of 10 years or more, started his sentence with CSC on 12 December 2016.

We find it difficult to evaluate George’s treatment by CSC except to say that his short stay in the West-1 RTC seems unimportant in this trajectory since he had three subsequent SIU stays (across three of the five regions of Canada).

There needs to be a thorough review of the Service’s practice of transferring SIU prisoners to RTCs and the returning of these prisoners to SIUs following their discharge from the RTC. The review must consider the reasons for transfer to the RTCs, the treatment received, the reasons for return to a mainstream institution, and the reason for transfer back into an SIU.

Independent External Oversight

Public Safety Canada tells visitors to its website that:

“Independent External Decision Makers (IEDMs) ensure that decisions related to the confinement of an inmate in a Structured Intervention Unit (SIU) are subject to scrutiny and ongoing assessment throughout this placement. IEDMs are appointed by the Minister of Public Safety and operate at arms-length from the Correctional Service of Canada (CSC).”Footnote 24

As will become clear, an important issue raised by this quotation is this: How long is the “arm”?

CSC notes that:

“IEDMs review inmate cases on an ongoing basis, in real-time, and provide recommendations related to an inmate's conditions and duration of confinement. Their decisions are binding” (emphasis in the original).Footnote 25

It is relevant to note that the original version of Bill C-83 that was first tabled by the government establishing the SIUs, the sections creating IEDMs did not exist. Nor was there any reference to any “independent” body that would assess CSC’s decisions to keep prisoners in the SIUs, or monitor/review their treatment while there. It was only once this gap was noted, and after the House of Commons committee examining the legislation subsequent to second reading recommended changes, that amendments creating IEDMs were added to the bill and became part of the implementation of these newly established CSC units.

As the MP who introduced the amendments pointed out:

The most significant amendment is the one I have introduced today at report stage, which would provide independent oversight of the new structured intervention units….

My amendment would create an independent external decision-maker who will monitor a number of factors for inmates in SIUs, including whether inmates avail themselves of the time out of their cells or if there is a disagreement with a health care provider's recommendation to transfer an inmate out of an SIU.

With this amendment, if an inmate does not receive the required minimum hours outside of the cell or the required minimum hours of human contact for five straight days or 15 days out of 30, the independent external decision-maker can investigate whether the Correctional Service has taken reasonable steps to provide opportunities for those hours, make recommendations to the Correctional Service to remedy the situation, and if the Correctional Service has not acted accordingly after seven days, the decision-maker can direct it to remove the inmate from the SIU and give notice to the Correctional Investigator.

In addition, the independent external decision-maker will also have the power to review cases and provide direction in the event that the senior Correctional Service health care committee disagrees with the recommendation of a health care provider to transfer an inmate out of an SIU or alter conditions of confinement.

Finally, the independent external decision-maker will conduct a review of each offender's case after 90 days spent in an SIU and every 60 days thereafter.

The creation of an external oversight mechanism was supported by the majority of witnesses we heard at committee. I am so pleased that we were able to respond to their input and move forward with this vital independent oversight mechanism.Footnote 26

There are a set of important reviews that involve the IEDMs in the SIU process. First of all, they review cases in which:

“for five consecutive days or for a total of 15 days during any 30 day period, an inmate confined in a structured intervention unit has not spent a minimum of 4 hours a day outside the inmate’s cell or has not interacted, for a minimum of 2 hours a day, with others, an IEDM shall, a soon as practicable, determine whether the Service has taken all reasonable steps to provide the inmate with the opportunities [for time out cell] and to encourage the inmate to avail themselves of those opportunities” (s. 37.83 (1) of the CCRA).

The Panel has not yet looked at time-out-of-cell reviews; however, we know that prisoners are not routinely getting the minimum hours out of cell indicated in the legislation.

The second set of reviews around the length of stay has, in fact, received a fair amount of attention from this Panel and from research carried out on the SIUs prior to the establishment of this Panel.Footnote 27 This Panel has addressed the operation of the IEDM model in dealing with the length of stay in the SIU in two separate reports, including its 2021 to 2022 Annual report and a thematic update focused on the IEDM model which was completed in late 2022 and released by the government roughly two months later.

As noted in the 2022 annual report, IEDMs are in part trained by CSC. Furthermore, they are largely dependent on CSC for the information on which they base their decisions. CSC must provide the IEDM with whatever information under its control that the IEDM requests (CCRA s. 37.7).Footnote 28 Further, nobody is specifically acting on behalf of the prisoner to ensure full and timely disclosure or to challenge the accuracy of the file material that is shared. Anecdotally, the Panel has been told that SIU prisoners have difficulty accessing Counsel when facing hearings or reviews. This introduces a new concern related to the fairness of the oversight process.

Our understanding is that the IEDM is typically given a package of “standard” information for each case that they are to review. Prisoners in an SIU are to be given an opportunity to provide written information to the IEDM (CCRA s. 37.72) and IEDMs may, if they wish, communicate directly with prisoners (CCRA 37.73). Before a decision is made, the IEDMs are required to provide the prisoner with the information that they are considering (or a summary of that information), though they don’t have to indicate to the prisoner what information they received that is not explicitly being considered. Furthermore, certain information can be withheld from the prisoner for the safety/security of any person or the institution or for “public interest” reasons (CCRA s. 37.71).

Although IEDMs are required to make their decisions about a specific prisoner available to that prisoner, decisions, more generally, are not public. Indeed, although IEDMs may (CCRA s. 37.77) publish or otherwise disseminate information, there is no requirement that they do so. As far as we can determine, in an information age in which it would be easy to post or provide anonymized IEDM decisions, a representative sample – or indeed the full set – of IEDM decisions on “time in the SIU” or any other issue are simply not made available. Hence this Panel has not, thus far, been able to examine the manner in which important decisions about prisoners are made by the IEDMs.Footnote 29

Having access to the IEDM decisions would help but would not be sufficient. We are unable to reach any conclusions about what information the IEDMs find persuasive or how much weight is given to that information. We do know that there is significant variation across IEDM decisions. Looking at three years of decisions (2020, 2021, and 2022), Table 14 shows that two IEDMs have never once rendered a decision removing the prisoner from the SIU. That stands in contrast to another two IEDMs who have rendered that decision in 43.9% of their cases they reviewed, which stands in contrast to the remaining ten IEDMs who have rendered that decision in 12.4% of their cases. This variation appears not to have changed over time.

Table 14: Decisions made by individual IEDMs (2020, 2021, and 2022) on whether the prisoner should be moved from the SIU cell
IEDMs IEDM Decision Total decisions (%) Total number of decisions
Prisoner should not remain in the SIU Prisoner should remain in the SIU
IEDMs #2 and 13 0% 100% 100% 113
IEDMs #3 and 8 43.9% 56.1% 100% 82
All other IEDMs decisions 12.4% 87.6% 100% 461
Total 14.2% 85.8% 100% 656

The variation seen here may well relate to the training that IEDMs receive, the information that they get, the professional background of the IEDM or variation across IEDMs in their tendency to accept recommendations from authorities. It should be remembered, of course, that there is known to be substantial variation in decision making across other criminal justice officials (e.g., judges at sentencing). Hence variation across IEDMs dealing with similar cases is not surprising.

Notwithstanding the above, there is general consistency in the decisions over time. In 2020, for example, 13.3% of decisions were that the prisoner should be removed; this increased to 15.5% in 2021 and declined to 14% in 2022.

The timing of these IEDM reviews also needs to be considered. As we have pointed out, defining what a prolonged stay is in an SIU is arbitrary (and for that reason we have started with the international, Mandela rules, standard). The CCRA does not define it explicitly except to say that stays in SIUs should be as short as possible. But short, as in the length of the arm in an “arm’s length” decision, is not defined.

It is worth considering the list of ‘reviews’ that take place before an IEDM sees the case.Footnote 30 In reading this list (for prisoners who were in an institution with an SIU), it is also worth keeping in mind that 15 days is the internationally accepted point after which prison authorities should be especially concerned about the imposition of the most restrictive practices such as isolating conditions of confinement. Given that prolonged isolation (beyond 15 continuous days) attracted the attention of Canadian courts, not having a review of time spent in an SIU cell until after 60 days cannot be considered an acceptable safeguard. The timing for reviews does not support the goal of short stays:

Day 1
Prisoner is transferred to an SIU
Day 30
By this date, the Institution Head must make a determination that the prisoner should stay in the SIU (CCRA s. 37.3(1)(b))
Day 60
By this date (and every 60 days after that), the Commissioner (or a person designated by her) must review the case to determine if the prisoner should be in the SIU. (CCRA s. 37.4).  If the “Commissioner” decides that the prisoner is in need of a longer stay in the SIU, the case is referred to an IEDM
Day 90
An IEDM decision on whether the prisoner should stay in the SIU (CCRA s. 37.8)

This may appear straightforward; however, CSC’s own data suggests that this is not the case. In Table 15, we examine whether cases were referred to IEDMs by Day 75 of the prisoner’s stay in the SIU, which is roughly midway between the Commissioner’s review and the day that the IEDM decision is due. Table 15 shows that roughly 30% of cases that involve stays over 75 days are not sent to IEDMs. This changes little over time. In terms of actual numbers, this means that during this three-year period, there were 821 cases where the stay was over 75 days, and 261 of them were never sent to an IEDM for review. Even among those who stayed over four months (N=336), 30% (N=102) were never sent to an IEDM despite the legislated requirement to have an IEDM decision within 90 days of entering the SIU.

Table 15: For person-stays in an SIU that were 75+ days, was the case sent to an IEDM for a ‘binding’ length-of-stay review?
When the person’s SIU stay started Was the case ever sent to an IEDM for a review of the length of stay? Total (%) Total cases
Case was never sent to an IEDM Case sent one or more times
2020 33.8% 66.2% 100% 328
2021 31.5% 68.5% 100% 279
2022 29.0% 71.0% 100% 214
Total 31.8% 68.2% 100% 821

These findings were disconcerting. Prisoners’ files were not sent for review even though the review was required by the legislation. The only explanation we have is that the IEDMs “lose jurisdiction” once the Institution Head or the Commissioner (or delegate) determine a prisoner should be transferred out of the SIU. The rationale is that an external independent review is not necessary if the decision to transfer has already been made. The problem is that the decision to transfer a prisoner out of the SIU, and the implementation of that decision are two different things. Many prisoners have remained for weeks or months in an SIU cell despite the existence of an order to move the prisoner. This requires a legislative fix.

When we reported on regional variation at the end of 2022 in the SIU Implementation Advisory Panel UpdateFootnote 31, we noted that the variation was substantial. This variation persists. It appears, for example, that it is less likely to have decisions that the prisoner should be removed from the SIU in the Prairie and Pacific regions (roughly 3%) compared to Ontario (19.8%; see Table 16). We see similar variation in each of the three years for which we have data. In this same December 2022 report, we first noted that prisoners who had received IEDM decisions that they should be transferred out of the SIU were, in many cases, kept in the SIU at least 61 days after the case had been submitted to the IEDM. Over the last three years, looking only at those who had one IEDM review and were then released (N=237), 36.4% of those whose IEDM decision was that they should not remain in the SIU were still in the SIU over 60 days after their case was referred to an IEDM. Only 10.5% of those whom the IEDM said should remain in the SIU ended up staying that long (see Table 17). This problem was true in each of the three years. Indeed, as we pointed out in our December 2022 update, if a prisoner wanted to get out of the SIU without a long stay after the decision is rendered, they are best off if the IEDM does not order them to be transferred out.

To conclude this section, we want to be clear that we support independent external oversight.  The concerns we have noted flow not from the concept of the IEDM oversight model, but its execution. We find the legislative framework to be inadequate. The deployment of IEDM reviews is overly complex and confusing. The timeframes for reviews are not well aligned with the goal of ensuring that SIU stays are as short as possible, and they do not reflect emerging internationally accepted norms of prohibiting isolating conditions of confinement for more than 15 consecutive days.

Table 16:  IEDM Decisions Across Regions and Over Time
Year case was sent to IEDM Region where prisoner was located when case was sent to IEDM IEDM Decision Total % Total Decisions
Inmate should NOT remain in SIU Inmate should remain in SIU Transferred out of SIU before decision rendered
2020 Atlantic 11.3% 54.8% 33.9% 100% 62
Quebec 13.5% 56.3% 30.2% 100% 126
Ontario 22.0% 53.7% 24.4% 100% 41
Prairies 2.4% 67.7% 29.9% 100% 167
Pacific 7.7% 51.3% 41.0% 100% 39
Total 9.2% 59.8% 31.0% 100% 435
2021 Atlantic 25.0% 34.4% 40.6% 100% 32
Quebec 16.4% 35.6% 47.9% 100% 73
Ontario 21.9% 37.5% 40.6% 100% 32
Prairies 3.3% 56.0% 40.7% 100% 182
Pacific 0.0% 52.7% 47.3% 100% 55
Total 8.8% 48.1% 43.0% 100% 374
2022 Atlantic 16.7% 54.2% 29.2% 100% 24
Quebec 9.5% 23.8% 66.7% 100% 42
Ontario 13.0% 26.1% 60.9% 100% 23
Prairies 4.3% 43.8% 51.9% 100% 162
Pacific 3.8% 43.4% 52.8% 100% 53
Total 6.6% 40.5% 53.0% 100% 304
Total Atlantic 16.1% 49.2% 34.7% 100% 118
Quebec 13.7% 44.4% 41.9% 100% 241
Ontario 19.8% 41.7% 38.5% 100% 96
Prairies 3.3% 56.0% 40.7% 100% 511
Pacific 3.4% 49.0% 47.6% 100% 147
Total 8.4% 50.6% 41.1% 100% 1,113
Table 17: Those with only one IEDM review and then were released: Time in SIU until Release by IEDM decision
Year when case was referred to IEDM IEDM Decision IEDM Decision Total % Total Decisions
Stay ended within 30 days of referral to IEDM Stay ended within 31-60 days of referral to IEDM Stay ended 61 days or longer after referral to the IEDM
2020 Prisoner should not remain in 31.3% 40.6% 28.1% 100% 32
Prisoner should remain in 20.3% 71.9% 7.8% 100% 64
Total 24.0% 61.5% 14.6% 100% 96
2021 Prisoner should not remain in 5.0% 45.0% 50.0% 100% 20
Prisoner should remain in 14.0% 74.0% 12.0% 100% 50
Total 11.4% 65.7% 22.9% 100% 70
2022 Prisoner should not remain in 35.7% 28.6% 35.7% 100% 14
Prisoner should remain in 15.8% 71.9% 12.3% 100% 57
Total 19.7% 63.4% 16.9% 100% 71
Total Prisoner should not remain in 24.2% 39.4% 36.4% 100% 66
Prisoner should remain in 17.0% 72.5% 10.5% 100% 171
Total 19.0% 63.3% 17.7% 100% 237


This Panel has three clear conclusions that flow directly from our findings.

Conclusion 1 – Unfulfilled Promises

The operation of the Structured Intervention Units is not sufficiently consistent with the understood goals of the Bill C-83 legislative amendments (42nd Parliament, 1st session). The implicit promises made when Parliament changed the law in 2019 are not being fulfilled.

Details of the inconsistency between the intent of the legislation and current practice are contained throughout this report. One can point to the length of the stays in the SIUs, the absence of timely independent review, and the inability to achieve the minimum time out of cell and meaningful human activity as obvious problems.

The response by CSC that prisoners do not sufficiently avail themselves of opportunities to leave the SIU is not sufficient. The experience of isolating conditions of confinement has been demonstrated to be harmful. To attribute the failure to provide minimum time out of cell to prisoners, many with identified mental health concerns, who decline the opportunities offered by CSC is akin to suggesting that CSC bears no responsibility for preventing self-harm if a prisoner chooses to hurt themselves.

There are other problems that demonstrate that the operation of the SIUs is inconsistent with the goals of the legislation. As we have documented and as noted by the Office of the Correctional Investigator, the system of review by the IEDMs – Independent External Decision Makers – is inadequate. We are not faulting the IEDMs as individuals; rather the system itself. It is our view that these “External” decision makers are not sufficiently structurally independent of CSC since their initial training comes from CSC, they depend, primarily, on information compiled by CSC, and increasingly depend upon CSC to provide administrative support.

Although the IEDMs are decision makers, we have seen that in a significant number of cases, they are not asked to make decisions due to gaps in the law, and when they do make their decisions, CSC doesn’t necessarily implement these supposedly “binding” decisions. What does it mean to have an independent external decision maker if CSC fails to act on the decision in a timely way, if at all?

Notwithstanding the problems we have identified with the SIU scheme in our analysis, it is worth remembering that many prisoners prefer isolation in an SIU to the alternative, which is being placed in the mainstream penitentiary population. We see this as an indication more of the inadequacy of CSC’s administration of the rest of the penitentiary than a finding of adequacy of the SIUs. Said differently, when long stays in isolating conditions of confinement are preferable to an alternative, what does this say about the alternative?

Conclusion 2 – The Operation of the SIUs is not Improving

The operation of the SIUs is not improving in any material way. This Panel acknowledges that the implementation of the SIU regime was never expected to be easy, and that in the last year, some very small operational improvements have been detected. We have, however, looked at three full years of data on the operation of the SIUs and are very concerned that there is little evidence of meaningful improvement in the way these units are being operated. Said differently, improvement in the SIU system is unlikely if CSC continues operating the SIUs as they have since 2019.

Table 2 shows that there is little reduction in the rate of prolonged stays in SIUs. And these data do not correct for split stays, thus turning a single problematic long stay into two or more less problematic relatively short stays. Nor does the data consider stays in hidden (i.e., non-SIU) cells, or “segregation by another name” elsewhere in CSC’s institutions.

In this context, CSC is not adapting its administration of the SIUs in a manner that is most consistent with the change in the law that took place in 2019. We see little evidence that CSC has succeeded in learning much from its “best” or “better” practices. Among other things, we see important (and disturbing) variation in the management of certain groups, such as racialized or Indigenous prisoners, or prisoners with mental health problems. We have found in this review – as we have found in our previous five reports – that there are significant changes in the operation of the SIUs that need to be made as soon as possible.

One approach that CSC employs in responding to difficult prisoners is particularly disturbing. This is the practice whereby difficult SIU prisoners are moved from one SIU to another SIU, often in a different region. This practice suggests that CSC, as a national institution, does not want to take a cohesive national approach to the coherent and safe management of prisoners with complex and challenging needs.

In the context of unexplained variation in the operation of the SIUs, the special needs of Black and Indigenous prisoners, as well as those with mental health problems, need to be addressed.

Conclusion 3 – Reform of the SIUs Must be Carried Out Within the Context of CSC’s Overall Operations

A complete operational review of CSC is required. SIUs must be understood within the context of broader CSC institutional operations. Changes to SIUs necessarily affect the whole institution and vice versa. There is little point in fixing the SIUs if the result is simply that the problems are shifted to other and less visible parts of CSC’s institutions (e.g., the so-called ‘hidden cells’). Every indication is that the problems with the SIU reflect larger organizational problems.

We have been told that there are many prisoners who prefer to stay in SIUs because of fears about violence in the general population. This problem cannot be fixed by looking at SIUs in isolation.

Similarly, mental health issues cannot be examined in isolation from the rest of CSC’s operations. Once again, we can see the intersectionality of problems. As noted in this report, those with multiple SIU stays are especially likely to have mental health needs. And those with considerable mental health needs are especially vulnerable to being sent to multiple SIUs in different regions. It would appear to us that the examination of the treatment of those with mental health problems needs to be expanded and carried out in the broad context of CSC institutions. A policy or practice that implicitly endorses “Treatment by Travel” (moving difficult prisoners to another institution or region) should not be tolerated.

Work on the reform of the SIUs – by Public Safety Canada and CSC – must start immediately so that by the end of 2024, when this Panel submits its final report and the required Parliamentary review begins, plans for meaningful change in the operation of the SIUs will be underway and available for Parliamentary consideration. Work needs to begin on the legislative changes that might be necessary to have a more humane system for dealing with especially difficult prisoners. For Public Safety Canada and CSC, the message should be clear: change is needed now. The immediate changes we are calling for in our recommendations will not fix all that is not working in the SIUs, but that should not be used as an excuse for waiting an additional year to begin work when prisoners are, today, not being treated in a manner consistent with the goals of the legislation.

The urgent need for change has been clear since October 2020 when the first set of findings on the operation of the SIUs were released. Many of these same problems were highlighted in our three “updates” released in 2023.

Recommendations to the Minister of Public Safety Canada

The following three recommendations address the need to ensure that any further change in CSC operations is evidence-based and consistent with the central mission of CSC, which is the rehabilitation of prisoners and their reintegration into Canadian society (see CCRA s.3(b)).

These recommendations are for immediate action in 2024 and directed to the Minister of Public Safety Canada because of his responsibility for Correctional Service Canada. The Panel will incorporate the response to these recommendations into its work plan during the final year of its mandate.

1. CSC must be directed immediately to examine the relationship of the operation of the SIUs to its overall operations.
CSC staff have frequently suggested that the SIU model has had negative impacts on core correctional operations and have also suggested that SIUs are responsible for an erosion of the conditions of confinement for general population prisoners. An examination of these issues must start immediately. It must be supported by systematic data on the operation of the SIUs and CSC institutions more generally. And it must be completed before the end of 2024, be made public, and contribute to the pending 5-year review of Bill C-83.
2. CSC must be directed immediately to develop and implement a plan to reduce the high number of Black prisoners, Indigenous prisoners, and prisoners with mental health problems who are being transferred into SIUs, as well as reducing their lengths of stay while in the SIUs.
The action plan needs to be developed and made public within 90 days of the release of this report. Updates should be made public every 60 days thereafter documenting both the changes that have been made and the impact of these changes. The actions taken must also prevent the prolonged isolation of these prisoners in any other forms of restrictive custody.
3. CSC must be directed to immediately ensure that all prisoners held in an SIU for more than 15 days will have their case referred to an IEDM, regardless of whether or not the prisoner had been previously ordered released.
Enhanced reporting on the number of prisoners and their relevant characteristics (e.g., whether the prisoner is Indigenous or from a racialized group, the presence of mental health problems, length of stay) must be implemented. These reports will present data on prisoners in any SIU anywhere in Canada and who have been in an SIU for a total of 30 days or more. These reports must be released to the public monthly, starting immediately.

Each of these recommendations is grounded in the urgent need to align the operationalization of the SIU Model with the legislative goals of Bill C-83.

Actions must be undertaken by CSC to comply with the spirit, intent, and letter of the law. Changes need to be made quickly and monitored closely.

Our three recommendations should be seen as interim measures. As noted in our three conclusions, there is a need for a detailed and thorough examination of CSC’s operations as a whole, not just the operation of the SIUs. It is clear to this Panel that the aspirations of transformative change in federal corrections cannot be achieved within the current operational context. SIUs alone will not drive this change.

Appendix A - Responses to the 2021 to 2022 Annual Report Recommendations by Public Safety Canada and Correctional Service Canada

The SIU IAP’s 2021 to 2022 Annual Report, dated September 2, 2022, was publicly released and posted on the Public Safety Canada website on October 28, 2022Footnote 32. The SIU IAP’s 2021 to 2022 Annual Report contained forty-one recommendations addressing the following fourteen areas of concern:

Thirty-five of the recommendations were directed to the Correctional Service of Canada, which touched on all topics except the future of the SIU IAP. The six recommendations to the Minister addressed Independent External Decision-Makers, the future of the IAP, and enhanced accountability.

Responses for the Minister of Public Safety and the Correctional Service of Canada were posted at the same time the Report became public. The SIU IAP’s 2020 to 2021 Annual Report and the responses to the recommendations can be found at the same location as can be found all of this Panel’s previous reports.

The SIU IAP prepared an initial review of the responses and determined that more information was required in order to reach any conclusions regarding the degree to which the responses addressed the identified issues.

Between November 2022 and June 2023, several meetings and discussions were held with Correctional Service of Canada Executive Committee members, senior Public Safety officials, and staff in the Minister’s office. On May 10, 2023, Correctional Service of Canada provided an update on actions taken. However, neither the department of Public Safety nor the Minister’s office updated their response.

The initial response, the formal Correctional Service of Canada update, and information gathered during informal discussions were reviewed and incorporated into the SIU IAP analysis of the responses. The SIU IAP developed a four-point scale to assess the response to the recommendations (i.e. Responsive, Partially Responsive, Non-Responsive, and Recommendation Rejected). Of the responses, 20% (n=8) were found to be Responsive, 44% (n=18) were Partially Responsive, 29% (n=12) were Non-Responsive, and 7% (n=3) of the recommendations were Rejected.

The Panel is disappointed in this outcome. The Correctional Service of Canada and the Minister were not fully responsive to the fourteen areas addressed in the Annual Report. Notably, Correctional Service of Canada was non-responsive to all the IAP’s recommendations about alternatives to SIUs and length of stay in SIUs. Correctional Service of Canada also rejected the IAP’s recommendations about the Senior Deputy Commissioner approving inter-regional transfers to or from an SIU. Aside from making a commitment to “enhance the timeliness and effectiveness of IEDM decisions” the Minister has not clearly committed to amending the Corrections and Conditional Release Act to establish more timely and enforceable reviews and orders regarding the length of SIU placements. Likewise, there has been no commitment to implement the recommendation to put into place Ministerial monitoring and public reporting of progress against the Commissioner of the Correctional Service of Canada’s Mandate Letter Priorities.

Appendix B - Some methodological and presentational notes

Measures and the grouping of data.

1. For most of our presentations, we have used, as our unit of analysis, the “person stay” – a single stay in an SIU (not a restricted movement cell) by a prisoner. Two separate stays by a single prisoner are therefore counted as two independent stays. Obviously when we are looking at multiple stays by a single prisoner our ‘unit of analysis’ shifts to being the prisoner, not the ‘person stay.’

The distinction between analyses using a “person stay” and analyses that look at the number of stays in an SIU that people who had at least one stay experienced is important. For example, in Table 2, we note that there were 4969 person-stays across the three-year period.  Many of these involved people who were transferred to an SIU more than once. Table 11, on the other hand, describes the number of separate transfers a given individual had who had been transferred at least once. These are obviously very different measures. The first describes each stay as being separate. The second aggregates the data for individuals who have experienced SIUs only once or many times.

2. We normally “date” a person-stay in an SIU according to the date that it started. This creates a simple, and surmountable, problem: what should be done if the prisoner is still in the SIU at the end of the period being studied. For the most part, because we present categorical data (e.g., person-stay lengths of specific lengths), we simply follow those cases for a period of time that places them in the longest category (e.g., more than two months). We don’t know, therefore, whether, at the time of writing this report, there are still prisoners who entered an SIU in 2022 who were still there at the end of 2023. But we do know that they were there for more than 2 months. And in analyses of prolonged stays it means that we have at least two months of their experiences in an SIU.

3. There is a technical issue of how a “day” is defined. CSC’s standard way of defining a day in prison is such that any part of a day counts as a day. Technically, therefore, a stay in an SIU starting at 10 pm and ending the next day at 8 am would count, using this scheme, as two days. As a general rule, therefore, we subtracted one day from the simple “date-subtraction” calculation (which would have counted this example as one day). CSC created a more fine-grained definition of a day after the SIUs had been in existence for some time. We checked to see if this definition would make any substantive difference. Obviously, it would be most likely to affect the outcome of analyses involving very short stays in SIUs. We found no instances where any findings would differ in a material way. Intuitively, of course, the definition of a day would be expected to have little (if any) impact on any measure for those analyses involving prolonged stays in SIUs (16 or more days), especially since we used rather broad groupings of people.

The minor ambiguity of dates comes up in other areas as well (e.g., the timing of IEDM decisions). In general, we have tended to give CSC the benefit of the doubt in calculating these times.

4. Generally speaking, we have presented the findings in a manner where a measure is broken down into categories. In some previous reports, we used many categories for, as an example, the length of a stay. The number of categories that are used and the exact values of each of the categories should be determined by what one is interested in. We have been guided by our research questions. Sometimes we just looked at relatively short vs. prolonged stays where the dividing line was 15-16 days. In other instances, we were interested in much longer (prolonged) stays. Essentially, we decided these questions using two types of criteria: what we were interested in and what described the data most adequately, and efficiently.

5. For certain things, CSC seems to prefer one of the simplest measures of the “central tendency” of a distribution of a set of scores: the median. The problem in using the median – the score that represents the point where half the scores are higher, and half are lower – is that it could easily hide more than it reveals. Essentially it tells the user of this measure almost nothing about the variability in the scores.

Hence if a distribution of “time in the SIU” had only three values (26, 27, and 28 days, for example) and the three values had equal numbers of cases, the median would be 27 days.  But if one were to look at a distribution of scores where the lengths were equally divided into three different groups: 2, 27 and 127 days, the median for this second set of scores would also be 27 days. The fact that the first group had scores ranging from 26 to 28 days and the second group had scores ranging from 2 to 127 days would be obscured by the use of the median.

CSC told the Panel that “(t)he indicators for 2023-24 tell us that to date (December 2023) the median days in an SIU have decreased since 2021-22….” We have no idea how to interpret this statement (aside from the fact that we are talking about a finding that does not cover a full year). The hypothetical example below, involving 100 prisoners' hypothetical lengths of stay illustrates this problem. Imagine for each year there are only three “lengths of stays” for 100 people sent to SIUs during each period.

Period 1: Days in SIU for 100 prisoners

For all 100 prisoners, Median time in SIU is 27 days.

Period 2: Days in SIU

For all 100 prisoners, Median time in SIU is 25 days.

As in the statement that CSC made to the Panel in December 2023, the median decreased in these hypothetical data (from 27 days to 25 days).

Most people, we suspect, looking at the full distribution of these (hypothetical) data would not conclude that in Period 2, stays in the SIU had really decreased in length. For 60 prisoners, there were substantial increases; for 40 prisoners there were very slight decreases. We suspect that most reasonable people would doubt that the median does a good job of describing these data. 

We are not suggesting that our hypothetical example reflects the data that CSC reported to the Panel. All we are saying is that the usefulness of their simple measure (the median) is very limited. Most obviously, it treats a person who spent 1 day more than the median number of days exactly the same as it does a person who spent 500 days more than the median.

Presenting data in a manner that reflects the issues that the Panel was examining.

1. As already mentioned, we have tended to use categories representing ranges of lengths. As we have already pointed out, these vary a bit from table to table depending on what we are trying to describe.

The most common reason was that we were trying to capture, in as simple a way as possible, what the data were showing us. This would mean, for example, that if the pattern across groups with what might be thought of as “very long” stays was similar to that of “long stays”, we would tend to pool the data. If group differences were more dramatic or were simply different for “very long stays” we tended to keep that as a separate group.

In addition, the placement of the categorical breaks that we use in our tables were generally determined by what we were interested in. For example, in Tables 2 and 3 we used two categories in order to give CSC some leeway in achieving what some people might call a relaxed Mandela standard. But in Table 4, we looked at what might be considered to be very long stays in SIUs (91 days or more) because we noticed that certain groups were more highly represented in this grouping than others.

In other circumstances, where we felt it was important to examine stays that exceeded the time specified in the Mandela rules, we were inclined to use a break between 15 and 16 days as 16 days or more without meaningful human contact has special meaning under the Mandela rules.

In another instance (Table 1) we used slightly different categories for a very simple reason: we did not have access to raw data from “administrative segregation” and had to rely on data we found that had been categorized by others.

2. The question of how to describe whether a “difference” is real is not straightforward. Our preference, in general, is to present the numerical data and make our own judgements about whether a difference is important. The starting point, however, is the quantitative data.

In Table 6, for example, we suggest that the drop between 2020 and 2021 (from 52.0% to 43.8%) in the percent of people not getting their two hours of meaningful human contact was noteworthy but the drop from 2021 to 2022 (43.8% to 40.2%) was not. That clearly is a judgement call. That judgement, on our part, was based in part on the fact that the data show that there were 354 person stays (40.2% of the 880 stays) of 16 days or more that year where the prisoner did not get the treatment (meaningful human contact) that differentiates solitary confinement from what the SIUs were expected to be able to provide. But the most important point is that the data are in the table for readers to make their own judgement on whether a drop of a given magnitude is meaningful.

Interpreting the law

1. CSC has expressed its concern to us about what Parliament or the Government intended. They say, reading the letter of the law (the CCRA) that all CSC has an obligation to do is to provide prisoners with the opportunity to have 4 hours out of cell and 2 hours of meaningful human contact. We have suggested, instead, that Parliament, when it passed the law, actually intended that people would no longer experience solitary confinement (because they would spend time out of cell and would have time with other people). To support that observation, we reference the Hon. Ralph Goodale, Minister of Public Safety. The Minister, in introducing the bill at second reading might be understood as communicating the Government’s intent when he said:

The solution that we are proposing in Bill C-83 is to completely eliminate the existing practice of administrative segregation and replace it with a new approach, and that is the creation of structured intervention units, or SIUs.

These units will be separate from the general population so that the safety imperative will be met. But they will be designed and they will be staffed and resourced to ensure that the people who are placed there will receive the interventions, the programming and the treatment that is required.

Inmates in SIUs will be out of their cells for at least four hours daily, with a minimum of two hours of meaningful interaction with staff, volunteers, elders, visitors or other compatible inmates.

Additional mental health professionals will be hired and assigned specifically to the SIUs. The legislation will make it clear that inmates are not to be separated from the general population any longer than necessary.

This new approach will help to ensure the safety of correctional institutions and the public by strengthening the capacity of the Correctional Service of Canada to promote rehabilitation in a secure environment.

[Thursday 18  October 2018, emphasis added].

We suspect that most Canadians – including most Parliamentarians – interpret the words “Inmates in SIUs will be out of their cells for at least four hours daily….” as an intent.

2. There is a substantive and a methodological issue concerning prisoners not accepting CSC’s offers of time out of cell. CSC, of course, focuses on their obligation to “offer” whereas we have focused on what actually happens to prisoners.

But that is not the only problem. Our data from previous reports show quite clearly that CSC’s data demonstrates that CSC does not always even offer prisoners the time out of cells that they are promised by the legislation. As we pointed out in our first annual report:

“CSC has advised the Panel that despite daily offers of time out of cell, many prisoners refuse to avail themselves of the opportunity. When we looked in detail at prisoners who were in SIUs for 16 days or longer we found that for 1,335 of these long stay prisoners, or 64% of this group, refusals cannot explain the failure to achieve four hours out of cell. The results on achieving two hours a day of meaningful human contact were similar: Refusals do not explain all failures of prisoners to get two hours of meaningful human contact each day. We estimate that for 1,091 of the 2,071 long-stay prisoners (or 53% of these long stay SIU prisoners), refusals do not explain the failure to get the two hours of meaningful human contact promised in the legislation.”

In other words, in many instances, CSC does not offer prisoners time out of cell. In addition, in other cases, where time is offered, these offers are not always accepted.

From a policy perspective, the result is the same: prisoners are not getting what the Minister told Parliament they would indeed benefit from.

3. On a quite separate issue, there is a serious question as to what “mental health needs” mean in the context of measures related to mental health that we use in this report. In the December 21, 2023 Errors and Omissions Review, CSC told the Panel that:

“It is not clear what the working definition is of ‘mental health needs’ on the part of the IAP [Implementation Advisory Panel]. Therefore, it is not clear who is included in this group…”.

This appears to be a fair comment. However, we did find it strange that CSC asked the Panel to resolve this issue. CSC implies that it was the Panel that assessed the mental health needs of prisoners and created these measures. Given that the Panel relied on CSC’s measures and assessments, the Panel thinks it odd for CSC to ask the Panel to defend or explain CSC’s own working definitions. The Panel didn’t create the indicators or assessments - CSC did.

The Panel used two measures. The first was whether the prisoner had a “mental health flag” defined by CSC to us as a “YES/NO flag identifying if the offender had a mental health need at the time of the period start date, based on MHNS data from Health Services Sector.” At a later point, this scale was expanded somewhat (see Table 12).

The Panel also had a measure of whether the prisoner had (a) no or low needs, (b) some needs, or (c) considerable or higher needs. On a separate dimension, CSC indicated whether, in its view, these needs were decreasing, not changing, or increasing. Combining across levels of (three) needs and three types of changes gives us 9 groups (see Table 10). In order to communicate the importance of certain categories, we pooled these nine groups into 3 distinct groups. CSC seems to be suggesting that they would have preferred that we did something different, but it was not clear to us what the advantage of their approach was. The Panel was most interested in examining whether those whose mental health was, according to CSC, getting worse were especially likely to have long SIU stays. See Table 10.

There is a certain irony in the fact that CSC expressed concerns to us about what the working definition of these measures are since the measures came from them.

4. CSC would prefer that we not use the term “solitary confinement.” We sometimes use that term to describe what is happening to some prisoners for a simple reason: their treatment reflects what is broadly known and accepted (e.g., in the Mandela Rules) as solitary confinement. In other cases, we mention solitary confinement as a contrast to the intended operation of the SIUs.

A question about the history of oversight of the SIUs

1. For reasons we do not fully understand, as part of its Errors and Omissions feedback, CSC felt that this Panel should indicate in this report “why” CSC failed to provide the first IAP Panel with any data. That, of course, is something that CSC should be able to explain more authoritatively than could anyone outside of CSC. The best publicly available explanation that this Panel knows of can be found in an appendix to the first report on the operation of the SIUs. See Appendix D in the 26 October 2020 report entitled “Understanding the Operation of Correctional Service Canada’s Structured Intervention Units: Some Preliminary Findings” available at: Reports on Canada’s Structured Intervention Units.

Date modified: