Introduction

Within the context of western countries that have witnessed (in some cases, dramatic) increases in imprisonment rates over the last half century, Canada constitutes an anomaly. Restraint in the use of the criminal justice system generally and in the recourse to incarceration in particular has been part of Canadian criminal justice culture for many decades (Reference Webster and DoobWebster and Doob 2007, Reference Webster, Doob and Reitz2018). The result has been relatively stable adult incarceration rates since the 1950s. This long-term stability contrasts dramatically with Canada's closest comparators. Since 1980, Canada's overall imprisonment rate vacillated between a low of 92 (1980) and a high of 116 (1994) per 100,000 residents. In contrast, levels of incarceration in England/Wales nearly doubled from 85 (1980) to 153 (2011) and those in the United States more than tripled from 222 (1980) to 756 (2008) (Reference Webster, Doob and ReitzWebster and Doob 2018).

While Canada's stability in imprisonment has garnered international attention (e.g., Reference TonryTonry 2013), Canada's real “criminological fame” (at least from a national perspective) may ironically lie in the dramatic, yet virtually unnoticed, decline in youth imprisonment between the mid-1990s and 2015. An average of 3825 youths (aged 12–17) was serving custodial sentences in 1997/8.Footnote ¹ In 2015/16, it had decreased to 527—an 86 percent reduction. The rate of (sentenced) youth imprisonment dropped from 157 youths per 100,000 youths in the population to 23.Footnote ² Notably, a 2003 legislative change only partially explains this decline.

A comparison with Canada's adult system makes the point more vividly. Looking at total imprisonment (sentenced + pretrial), the youth imprisonment rate went from 192 per 100,000 youths in 1997 to 51 in 2015—a 73 percent reduction. In contrast, an average of 109.4 adults per 100,000 residents was in Canada's prisons in 1997. In 2015, this rate was essentially unchanged. Figure 1 displays these divergent trends.

Figure 1. Imprisonment Rates (per 100,000) of Youths and Adults in Custody in Canada (Remand + Sentenced Counts) (1997–2015).Footnote *

Given concerns about higher-than-optimal imprisonment rates, Canadian youth justice takes on wider relevance. Paralleling the proliferation of theories offered to explain America's “imprisonment binge” (e.g., Reference Ruth and ReitzRuth and Reitz 2003; Reference TonryTonry 2004), American scholars have proposed various mechanisms to reduce incarceration (Reference TonryTonry 2014, Reference Tonry2017). Notably though, few analyses exist of successful decarceration attempts.

Arguably the best understood case is Reference Lappi-SeppäläLappi-Seppälä's (2000, Reference Lappi-Seppälä2007, Reference Lappi-Seppälä2012) work on Finland's reduction in adult imprisonment from 150 per 100,000 residents in 1960 to 60 in 2000. This change occurred largely because Finland—a Nordic country—found in the 1950s that its incarceration rate was dramatically higher than those of its neighbors. This perceived disgrace (Reference Lappi-SeppäläLappi-Seppälä 2000) initiated a four-decade effort to reduce imprisonment that ensued as a result of conscious, long-term, and systematic criminal justice policy with judicial support.

Similarly, the Dutch adult prison population decreased from approximately 70 to 20–25 inmates per 100,000 residents between 1945 and 1973. Reference Downes and van SwaaningenDownes and Van Swaaningen (2007) identified several immediate interacting factors contributing to this decarceration. More broadly, the suffering of members of the Dutch underground in German prisons during World War II encouraged the recognition that prisons should be governed by minimalist and humane policies.

Both the American state of California and the Canadian province of Alberta also experienced substantial decreases in incarceration rates. Between 1968 and 1972, Governor Ronald Reagan proudly presided over a 34 percent drop in adult imprisonment rates (Reference Gartner, Doob and ZimringGartner et al. 2011) while Ralph Klein (Alberta's Premier) enabled a similar reduction (32 percent between 1993 and 1997) in Alberta's provincial adult incarceration rate (Reference Webster and DoobWebster and Doob 2014). In contrast to European counterparts though, neither case involved a principled reduction in imprisonment. Rather, the catalyst in both examples was financial—a desire for a balanced budget.

Nor is successful decarceration restricted to adults. England/Wales as well as the United States experienced substantial reductions in youth imprisonment—albeit at different times and rates. In England/Wales, youth imprisonment (for those under 18) decreased from roughly 2700 to 3000 youths in custody during the period 2000/1–2008/9 to about 1000 in 2015/16 (Reference BatemanBateman 2017: 48, figure 15). Reference BatemanBateman (2014) suggests that this success is largely attributable to explicit government goals for reducing the number of first-time entrants into the youth justice system. This decline was achieved primarily through decreased reliance on police arrest and increased use of informal or diversionary measures/initiatives. As such, entry into the court system was delayed, extending “the number of options…before custody appears inevitable, while simultaneously reducing the appearance of persistence, which frequently triggers deprivation of liberty” (Reference BatemanBateman 2014: 420). Notably, Scotland also reduced (by 64 percent between 2006 and 2016) the number of 16-/17-year olds in custody through various policy initiatives (Scotland 2017).

Youth imprisonment also dropped in the United States (albeit not as dramatically). While the number of juvenile offenders in residential placement facilities increased during the 1990s, it subsequently decreased from roughly 100,000 in 1999 to approximately 50,000 in 2015 (OJJDP 2018). Although a “national” explanation for American trends would almost certainly fail to explain substantial state variation, Reference Taylor, Decker and MarteacheTaylor (2017) suggests that the U.S. juvenile commitment rates generally have declined sharply due largely to “declines in juvenile offending and changing national sentiments toward confinement” (p. 144). This latter change may reflect both the “recent shift of the juvenile justice system back to a treatment orientation” (p. 135) and state needs to reconsider juvenile confinement practices because of their high costs (especially following the 2007 Great Recession). States have “increasingly turned to alternative options that allowed juveniles to remain in their communities, while still being supervised and provided with rehabilitative services” (Butts & Evans, cited by Reference Taylor, Decker and MarteacheTaylor 2017: 147).

Our study adds to this literature on decarceration by presenting a detailed, comprehensive account of the dramatic drop in Canadian youth imprisonment rates since the late 1990s. We take our cue from prior case studies documenting reductions in prison populations that underline the need to address two interrelated—yet distinct—issues. First, operations matter—both individually and collectively. Hence we examine not only the operational mechanisms (e.g., police charging practices, prosecutorial policies) contributing to the reduction in the number of youth in prison as separate and discrete factors, but we also bring all of them together to demonstrate their combined effects. Second, context matters. Thus, we situate Canadian youth decarceration within its broader historical, sociocultural, legal, and political contexts, describing how these wider events/factors initiated and ultimately accomplished (as well as perpetuated) the dramatic drop.

In brief, we address both “how” youth imprisonment was reduced and “why” these operational changes occurred. To this end, we examine the wider contextual factors at play and how they might have special effects in the youth system. Indeed, our analysis builds on Reference Zimring and RosenheimZimring's (2002, Reference Zimring2005) suggestion that the existence of separate youth justice systems makes it possible to develop policies and practices for youths that might be politically impossible for adults. We seek to understand how and why the treatment of youths does not necessarily mirror what happens with adults within the same jurisdiction.

The paper proceeds as follows. Part I provides an overview of Canadian youth justice policy until the early 1990s. It highlights a gradual change in culture, manifest in a shift in approach to young offenders from a social welfare model in which prison was seen as a rehabilitative tool to one rooted in the belief of restraint in the use of incarceration. Part II describes the broader historical, sociocultural and political context of the 1990s, which set in motion the dramatic decline in youth imprisonment. In particular, we discuss the politics of change—that is, the (bumpy) path leading to broad political support for youth decarceration.

Part III describes the current youth justice legislation introduced in 1999 and promulgated in 2003. This law represents more than the culmination of decarceration processes already underway. We demonstrate that it simultaneously constitutes the most obvious “cause” of youth decarceration. Within this context, we highlight its lengthy “gestation period” (between its initial introduction in Parliament and its ultimate enactment) as well as the shift from a mere aspirational framework to one of an operational nature. Part IV presents empirical data on the location within the youth justice system accounting for the reduction in the use of custody, describing the cumulative/interactional nature of the various decision-making points in bringing about decarceration.

In Part V, we employ several “counter factual” arguments to support our general thesis. First, we contrast what happened with the rates of youth sentenced custody to those of youth pretrial detention. Indeed, this latter custodial population constitutes an issue, relatively speaking, ignored within the decarceration movement. Second, we attempt to explain why youth incarceration declined while adult incarceration did not. Part VI concludes by discussing the lessons that Canada's youth criminal justice system might teach us about decarceration. We highlight the importance of values consistent with decarceration as a necessary—yet insufficient—condition for bringing about change. Rather, these values must be coupled—we argue—with broad political will to change. Indeed, it was likely fundamental that the new youth justice legislation promoting decarceration was both a result of the will to change as well as a cause of it. However, it was equally important that this legislation was prescriptive—providing explicit legislative direction—and not simply aspirational in orientation.

Similar to Reference CampbellCampbell (2014), our work draws from more than 25 broad criminal justice policy statements of the Government of Canada published since 1914.Footnote ³ These reports/documents describe how governments and government-appointed bodies saw criminal law and the role of the justice system. They are complemented by an examination of federal bills (enacted and not) related to youth justice since the 1960s. Quantitative data are drawn from Statistics Canada's CANSIM (see Footnote 2) which includes data from police (incidents recorded across Canada and adults/youth charged), corrections (admissions and counts), and courts (cases to court, decisions, sentences). In cases of incomplete court data for some provinces/territories, we used conservative estimates (relative to the question being addressed). Examining interprovincial differences was beyond the scope of this paper.

Part I—Gradual Cultural Change: Canadian Youth Justice (1908–1990)

Under Canada's constitution, the federal government is responsible for criminal law, but the administration of justice is a provincial responsibility. In 1908, the federal government took responsibility for youth justice by creating “criminal” youth justice legislation. Entitled the Juvenile Delinquents Act (JDA), it was welfare based in orientation and broadly similar to contemporaneous U.S. youth justice practices. It created a single “offense” of delinquency, allowing intervention into the lives of youth age 7 or older (to age 16–18, depending on provincial choices) who broke any federal, provincial, or municipal law. The Act emphasized “welfare” principles throughout and permitted youths to be placed in secure facilities indefinitely (until age 21).

The JDA survived with few modifications until 1984 (Reference Doob and SprottDoob and Sprott 2004). However, a process of change in youth justice culture began in 1961 when the federal justice department established the Committee on Juvenile Delinquency. Just as “rights oriented” issues (e.g., In re Gault, 387 U.S. 1 [1967]) were influencing American youth justice, similar concerns were debated in Canada.

The committee's 1965 report recommended a shift from the welfare approach and, importantly, endorsed restraint in the use of custody. The committee agreed

The report included an explicit recommendation to strengthen the goal of restraint in the use of custody (Recommendation #76: 295). While constituting a break with the welfare approach underlying the JDA in which prison was often seen as the most appropriate (and benevolent) response to offending, the cultural values underlying these recommendations were familiar to Canada. In the adult system, numerous formal statements of criminal justice policy throughout the twentieth century repeatedly endorsed an official culture of restraint in the use of incarceration (Reference Doob and WebsterDoob and Webster 2006, Reference Doob and Webster2016).

This new approach to youth justice led to the introduction of a Parliamentary bill in 1970 (Bill C-192) which shifted toward definite sentences and other restrictions on state intervention. Although abandoned by the government, this bill was followed by proposals which focused, in part, on diversionary mechanisms to keep youths from being formally charged as well as calls for restraint in the use of youth incarceration (Solicitor General Canada 1975). These recommendations suggest that, by 1975, consensus had developed around the importance of restricting youth imprisonment. “Highlights” for proposed legislation from the Liberal (1977) and Conservative (1979) governments argued for restraint in the use of youth court and custody (Solicitor General Canada 1977, 1979).

This gradual shift in youth justice culture culminated in a 1981 bill proposing completely new legislation. The new act passed with all party approval. Becoming law in 1984, the Young Offenders Act (YOA) limited youth justice intervention to what were criminal offenses for adults and restricted the permissible length of custodial sentences to 2/3 years (depending on the offense). Restraint in the use of the criminal law generally and custody in particular was also explicit, though only through vaguely aspirational language within the broad Declaration of Principles:

As if to emphasize the law's ambivalence on restraint though, this section also made it clear that society needed protection and youths might need treatment:

Nevertheless, the YOA clearly reflected a cultural change. Restraint in the use of custody became a guiding principle. However, this new philosophy was still in its fledging years. No explicit direction—only vague/weak aspirations—requiring restraint in imprisonment existed. Perhaps the most significant shift was to fixed-length sentences (3-year maximumsFootnote ⁴) and the restriction of the Act to criminal matters. Consistent with this approach, the Conservative government amended the new law in 1985–1986 to include another aspirational section to further limit the use of custody:

Part II—The Politics of Change: Canadian Youth Justice (1993–1999)

Despite indications of increases in the use of youth court and custody into the early 1990s (Reference Bala and AnandBala and Anand 2012), the story of the remainder of this decade is one of the creation of the political will to decarcerate youths. Restraint in the use of prison—as a core Canadian value—had been affirmed in the YOA but had not garnered unambiguous political support, despite unanimous support for the bill in the House of Commons. Furthermore, the call for restraint was codified in vague or only aspirational terms. More notably, the legislatively sanctioned use of custody for either treatment purposes or the protection of society reflected continuing ambivalence vis-à-vis the goal of restraint.

Part III—Accomplishing Decarceration: The Youth Criminal Justice Act

Acting upon the political momentum to decarcerate, the government introduced completely new youth justice legislation in March 1999. Importantly, this legislation was carefully presented to Parliament and the Canadian public. In the week preceding its introduction, newspaper headlines described it as “tough-on-crime”. Whether or not these stories resulted from planned government “leaks,” the public clearly was primed to expect a tough bill. Similarly, the government's press release was framed to suggest that the act was “harsh” (Reference Doob and SprottDoob and Sprott 2004: 230). Specifically, the first 11 of 13 bullet points highlighted harsh elements (e.g., presumptive adult sentences—subsequently ruled unconstitutional—for certain offenses). Only the final two points were “softer” in tone. By appearing to respond to public calls for harsher treatment of young offenders, the government might have tactically garnered public support.Footnote ¹¹ Indeed, the emphasis on tough-sounding provisions was consistent with the belief that it was politically advantageous to be seen as “tough-on-youth-crime” (Reference Barnhorst and IsmailiBarnhorst 2012).

But appearances can be deceiving. The bill itself tells a different story. The preamble acknowledges explicitly the over use of youth custody and the need to reduce it:

More importantly, the “aspirational” tone quickly becomes “operational.” The biggest change introduced by the Youth Criminal Justice Act (YCJA) is the directness of “guidance” on administering the law. The new legislation largely structures the decision-making process. Its predecessor (YOA) contained “admonitions” not to use custody when other sanctions were plausible. It also allowed police and prosecutors to keep youths out of the formal court system. However, these prior “admonitions” became—in the YCJA—much more explicit directives or formal hurdles that needed to be overcome. The government had seemingly decided that, having legal responsibility for young offenders, it should make explicit policy on how they are handled.

Illustratively, the bill contained clear operational principles designed to screen cases away from court and to restrict the use of custody. While the YOA suggested that alternatives to court “should be considered”, the YCJA (s.6(1)) directs police officers to consider noncourt responses in all cases before charging any youth. Although the failure to do so does not invalidate any charge, the statutory intent is clear. Similarly, s.4 states that noncourt responses not only are “often the most appropriate and effective” and presumed to be adequate for first-time nonviolent offenses, but also can be used repeatedly and for those offenders previously taken to court. Furthermore, the explicit focus of these directives on extrajudicial measures is reminiscent of the English approach in the early 2000s. Reference BatemanBateman (2017) notes that the dramatic decarceration of youths in England/Wales was achieved primarily by delaying entry of first-time entrants into the court system through decreased reliance on arrest and increased use of diversionary initiatives.

For youth found guilty, the YCJA (s.38) also provides clear directives to judges. Explicitly, “[t]he purpose of sentencing… is to contribute to the protection of society by holding a youth accountable for an offence….”. More importantly, “the sentence must be proportionate to the seriousness of the offence and the degree of responsibility of the young person for that offence.” Furthermore, subject to the proportionality requirement, “the sentence must be the least restrictive sentence that is capable of [holding the youth accountable]” and must be “the one that is most likely to rehabilitate the young person and reintegrate him or her into society.”Footnote ¹²

While the proportionality principle determines the relative severity of sentences, the actual sentence (within limits set by proportionality) must be that most likely to rehabilitate/reintegrate. However, proportionality alone does not define the actual level of punishments to be imposed. Thus some benchmarks were necessary. The YCJA focused on the decisions regarding imprisonment. Specifically, custodial sentences can only be imposed if one or more of four conditions are met:

• • it is a violent offense;

• • the youth has previously failed to comply with noncustodial sentences;

• • the youth was found guilty of a moderately serious offense and has previous findings of guiltFootnote ¹³; or

• • exceptional circumstances exist requiring a custodial sentence in order to be consistent with sentencing principles (including proportionality) (s.39(1)).

Furthermore, judges are explicitly required to explain why noncustodial sentences are not appropriate. Moreover, courts have generally taken a narrow interpretation of these provisions (Reference Bala and AnandBala and Anand 2012: 573–591).Footnote ¹⁴

Part IV—Decarceration in Action: The Impact of the YCJA

Figure 2 gives the punch line—at least in abbreviated form. The drop in the rate of sentencing youths to custody immediately following the implementation of the YCJA in 2003/4 is evident. It is the largest 1-year decline between 1991/2–2015/6.

Figure 2. Rate (per 100,000 12–17-Year Olds) of Sentencing Youths to Custody (Youth Court Data) (1991/2–2015/16).

These sentencing data are corroborated by the average counts of youths serving custodial sentences. Corresponding with the YCJA's implementation, Figure 3 Footnote ¹⁵ shows an equally impressive drop in 2003.

Figure 3. Rate (per 100,000 12–17-Year Olds) of Youths in Sentenced Custody (Correctional Counts) (1997/8–2015/16).

The striking drop in both the sentencing and correctional data corresponds with the coming into force of the YCJA and is consistent with the law's goals. Unsurprisingly, this finding has been discussed in detail by others (Reference Bala, Carrington and RobertsBala et al. 2009). Our point lies elsewhere.

First, we suggest that there are actually three periods of decarceration which deserve attention: one occurring between 1998 and 2003—several years before the YCJA came into force; one immediately following the law's enactment in 2003 and likely a reflection of the law's explicit directives; and one which extends well beyond the direct or proximate effects of the YCJA. We suggest that they tell their own (equally interesting) stories. Second, various factors/contributors are likely at play in producing this decarceration. We propose to disentangle these diverse mechanisms—or, more aptly, decision-making processes—to better understand their individual as well as cumulative effects.

(2) Disentangling the Overall Process

Given that our “story of decarceration” covers 1998–2015, the missing chapter is an examination of the criminal justice mechanisms through which this overall reduction in the use of court and custody was accomplished. While the drop in overall youth imprisonment rates is clear, it is less obvious which decision-making points are responsible.

We begin with the police. From the early 1990s to 2015, there were two quite different changes in the criminal justice process. First, a drop in the number of youth cases coming to police attention occurred. Second, there was a reduction in the proportion of cases handled by the police that resulted in a charge being laid.

We think that it is useful to consider the fact that the reduction in youth imprisonment started around 1998 (Figure 2). Figure 4 shows that the number of youths charged by the police began declining in 1991.

Figure 4. Rate (per 100,000 12–17-Year Olds) of Youths Charged by Police (1977–2016).

Data on youths “not charged” were unavailable before 1998. As such, we combined the rate of “youth charged” and “youth not charged” from 1998 onwards to create an index of “youth crime coming to police attention”. This constructed rate was relatively steady between 1998 and 2007 (Figure 5). It subsequently declined substantially. Hence, the decrease in youth incarceration (Figures 2 and 3) occurring before 2007 cannot be attributed to changes in the availability of offending youths.

Figure 5. Youth Crime Coming to Police Attention (1998–2016).

Figure 5 shows that the proportion of youths charged by the police declined somewhat before 2003. From Figure 4, we know that the rate of youths being charged had been dropping since 1991. Unfortunately, since data are unavailable, it is unknown whether this decline was a result of changes in police decision making.

But there is a clear impact of the 2003 implementation of the YCJA on police charging. As discussed elsewhere (Reference Bala, Carrington and RobertsBala et al. 2009), the YCJA resulted in the greater use of diversion by police. Hence although charging by police was decreasing prior to 2003, the YCJA had its own unique effect. In fact, the rate at which youth were not charged exceeded the rate of those being charged for the first time in 2003. Furthermore, despite subsequent declines in both rates, the former has consistently remained higher than the latter. In sum, the overall decline in youths being brought into the court system has seemingly changed because of a decline in both the number of youths coming to the attention of the police and the proportion whom the police charged.Footnote ¹⁸ These results are summarized in Table 1.

Table 1. Breakdown of Changes in the Operation of the Youth Justice System (1998–2015)

During the period central to our argument (1998–2015), crime associated with youth declined substantially (Row 1). Violent youth crime coming to police attention also declined (Row 2). Notwithstanding these two declines, the percentage of all incidents (Row 3) and violent incidents (Row 4) that resulted in a youth being charged also declined. Clearly police decisions not to charge youths are implicated directly in the reduction of the use of court (and presumably custody), as is the reduction in youths coming to police attention. Said differently, the police charged a declining portion of youths (Rows 3 and 4) drawn from a declining pool of youths who could be charged (Rows 1 and 2).

The next key decision-making point occurs in the court process and directly involves the prosecutor. Formal charge screening by prosecutors exists in three Canadian provinces (British Columbia, Quebec, and New Brunswick). But in addition, the YCJA encourages all decision makers to divert youth cases into “extrajudicial measures” before and after the court process has begun. Row 5 shows that the number of youth cases per 100 youth charged who actually entered the court process declined somewhat between 1998 and 2015.

However, a substantial number of cases in Canada are also withdrawn during the process when it becomes clear that it is not in the public interest to pursue the matter. Row 6 reveals that the percentage of cases with a guilty finding decreased during this period (detailed in Figure 6). The proportion of youth found guilty began to decline in 1998. While the YCJA clearly had an immediate impact, the percentage of cases with a guilty finding remained stable for several years following the enactment of the YCJA before dropping until roughly 2010.

Figure 6. Percent of Youth Cases with a Guilty Finding and Given a Custodial Sentence (1991/2–2015/16).

In sum, it would appear that those deciding on whether to take a case through the court process to the point of a guilty finding (typically the prosecutor) also contributed—independently—to the reduction in cases going to sentencing. The denominators used in this illustration are the number of cases resulting in charges by the police (Row 5) and the number of cases in which some kind of decision was made (Row 6). Consequently, these declines represent reductions occurring above and beyond the changing police practices.

Predictably, cases involving violence were less likely to benefit from these screening processes. The logic of reducing the number of cases going to court is typically that minor—largely nonviolent—cases will be screened out. As such, one would naturally find that the proportion of cases involving violence would increase. Row 7 shows that the proportion of court cases involving violence did rise, as did the proportion of violent cases with a guilty finding (Row 8). The less serious (nonviolent) cases are seemingly being screened out, leaving a higher proportion of violence cases. It follows that the average seriousness of the cases left for sentencing increased between 1998 and 2015—an important consideration when examining sentencing decisions.

The final major decision-making point is sentencing. As previously seen (Figure 2), judges reduced their use of custodial sentences throughout the period in question. In order to disentangle the impact of changes in sentencing from changes earlier in the process, Rows 9 and 10 present the proportion of cases with a guilty finding leading to a custodial sentence. Row 9 shows that the percentage of all cases with a guilty finding receiving a custodial sentence decreased from 29 percent in 1998 to 16 percent in 2015. Similar declines in the use of custody occur with those cases involving violence (Row 10).

Figure 6 shows that the proportion of cases with a guilty finding leading to a custodial sentence did not begin declining in any obvious way until roughly 2001—a few years after the police charging of youths started to decline. Apparently, the decarceration processes underway by 1998 did not affect judicial decisions to send youth to prison until several years later. Notwithstanding this slower impact, the YCJA brought in the single largest decrease in the percent guilty sentenced to custody, dropping from 26.9 percent in 2002 to 21.9 percent in 2003. This decline is consistent with the “hurdles” to the use of custody contained in s.39 of the YCJA. Equally notable though, one must remember that the police and prosecutors were screening out the less serious cases. Thus from the mid-1990s until 2015, the proportion of cases entering court involving serious violence (violence other than common assault) increased from 10.4 percent in 1995 to 22.1 percent in 2015. The proportion of cases being sentenced that involved serious violence rose during this same period from 9.7 to 23 percent. Hence the cases before the courts and at sentencing became, on average, more serious over time. Nevertheless, the proportion sentenced to custody declined until roughly 2009, at which point it stabilized.

Looking more closely at these custodial cases, youths sentenced to custody are presumed to be released automatically at the 2/3 point in their sentences. Under the YOA, youths were expected to serve their full sentences. Nevertheless, custodial sentence lengths did seemingly increase. In 1991,Footnote ¹⁹ roughly 44 percent of cases sentenced to custody received a sentence of 1 month or less. This value increased to 55 percent in 2002 and then returned to 44 percent in 2014/15.Footnote ²⁰ Across the whole period, fewer than 1 percent of custodial cases received a sentence of 24 months or more.

The use of probation or other community-based sanctions also did not increase. The rate of sentencing youths to probation (as the most serious sanction within the sentence) declined by 71 percent (from a rate of 1199.3 per 100,000 12–17-year olds in 1991/2 to a rate of 351.8 in 2015) and the rate of sentencing youths to community service orders (as the most serious sanction within the sentence) declined by 83 percent (from a rate of 370.5 per 100,000 12–17-year olds in 1991/2 to a rate of 64.2 in 2015). However, a slight increase in the length of probation sentences may have occurred. Excluding Manitoba, 14 percent of cases in which probation was imposed (as the most serious sanction or not) received a 1–2-year probation term in 1991/2. By 2015, 23 percent were receiving probation terms of this length. This increase may reflect the removal of minor offenses from youth court.

Clearly, each of the principal decision makers made their own independent contributions to youth decarceration in Canada. We would suggest that the decline in youth custody was the result of a general reduction in youth crime coming to the attention of the police, as well as changes in decision making at three stages in the youth justice system: the police decision to charge, the decisions at court to proceed to a finding of guilt, and judicial decisions at sentencing. However, their combined/cumulative effect is arguably even more impressive. Rows 11 and 12 (Table 1) show that the rate of youths sentenced to custody per 100,000 youths in the community declined by 83 percent and the rate of youths in sentenced custody (correctional counts) per 100,000 youths declined by a similar amount (85 percent).

Part V—The “Counterfactuals” to Decarceration: Youth PreTrial Detention and Adult Incarceration

Few would argue that Canada's decarceration project for its young offenders was unsuccessful. However, this achievement would appear specific to sentenced custody. Trends in pretrial detention tell a different story (Figure 7).

Figure 7. Rate of Youths in Sentenced and Remand Custody (Counts) and Proportion of the Custodial Population that is Pretrial Detention (2003/4 to 2015–16).

While sentenced custody counts fell dramatically from at least 1998 (Figure 3), rates of youth remand custody displayed a different pattern (at least from 2003 onward).Footnote ²¹ Although lower than youth sentenced custody rates when the YCJA was enacted, pretrial detention rates showed relative stability for the first several years. Peaking in 2007/8, the number of youths in pretrial detention exceeded that in sentenced custody for the first time—an inversion which continued through 2015/16. As a general description since the YCJA implementation, remand custody rates were steady or declining slowly at a time when sentenced custody counts declined more rapidly. The proportion of the custodial population in pretrial detention increased relatively steadily over this period.

Within the decarceration movement, pretrial detention seemingly presents an intriguing incongruity. No obvious reason exists to suggest that the political will to reduce youth imprisonment would be reserved exclusively for sentenced offenders. Rather, we would suggest that the explanation for the failure to reduce pretrial detention lies elsewhere.

Notably, the issue of remand was, relatively speaking, ignored in the YCJA. At the time of its drafting, some concern existed about pretrial detention. However, it may not have been seen as problematic in Canada until such time as sentenced imprisonment rates were controlled in a manner consistent with Canadian values.

Perhaps for this reason, the framework for pretrial detention was lacking in various important ways. In particular, the original YCJA sections on this criminal procedure were weak in their orientation. Although remand custody for social welfare reasons was prohibited (s.29(1)), little else was prescriptive. Until 2012, the law stated simply that detention was presumed unnecessary in the case that a youth could not normally be sentenced to custody. The mere aspirational nature of this section contrasts dramatically with the general structuring of other decisions in the YCJA through relatively strong and directive wording/provisions. Additionally, justices of the peace—who, in many provinces, handle youth pretrial issues but do not sentence youths—would be ill-equipped to determine whether a youth could be sentenced to custody if convicted.

In fact, the provisions concerning remand custody easily enabled the detention of youths. Few new provisions were contained in this part of the Act and they were applied within the general and vague Criminal Code framework for adults (Reference Barnhorst and IsmailiBarnhorst 2012). Sentencing was governed by precise prescriptive provisions. Pretrial detention was not. As Reference Barnhorst and IsmailiBarnhorst (2012: 130) notes, “The most dramatic impacts [of the YCJA] have been achieved in the two areas in which the Act is most detailed in terms of decision-making rules—extrajudicial measures and custody sentences”.

Amendments introduced in 2012 to the provisions governing pretrial detention attempted to provide stronger guidance by reducing judicial discretion. Hurdles were introduced to the use of remand custody. While some (Reference Bala and AnandBala and Anand 2012) believed that they should have reduced pretrial detention, Figure 7 suggests that the impact was, at best, small and short lived. Furthermore, it does not indicate an effect similar to that of the 2003 law (Figure 3). This lack of success may reflect the government's failure to promote the change. There was no publicity or even real acknowledgment that pretrial detention was a problem. Furthermore, no education/training was provided to those implementing the new provisions. Given the “natural” attractiveness of adopting the “safest” decision by criminal justice professionals (in this case, denying or delaying a bail decision),Footnote ²² education/training would seem crucial in combatting risk aversive decision making at bail (Reference WebsterWebster 2015).

An equally intriguing counter to the success of youth decarceration lies in its comparison with adult imprisonment (Figure 8).

Figure 8. Adult Imprisonment Rates (per 100,000 Residents) (1977–2015).

Overall stability in adult imprisonment is the result of an increase in the use of remandFootnote ²³ and a decrease in sentenced incarceration.Footnote ²⁴ While this relative stability may be noteworthy on the world stage, its contrast with trends in youth imprisonment may shed valuable light on those factors necessary—at least in Canada—to bring about a genuine reduction in the use of incarceration.

The difference does not appear to be rooted in sociocultural factors. Decarceration of both adults and youth is consistent with long-standing core Canadian values. Official documents written since the early 1900s typically suggest that Canada's incarceration rate is too high (Reference Webster and DoobWebster and Doob 2007, Reference Webster and Doob2015) and that Canadians are not well served by high imprisonment. Yet adult imprisonment has not decreased. This lack of decarceration—particularly contrasted with youth rates—can be explained (at least partially) by two factors. First, Canada's attempts to reduce adult incarceration have been largely aspirational rather than prescriptive. Second, Canada has seemingly lacked clear, unambiguous political will to decarcerate adults.

Largely paralleling youth criminal justice policy during the same era, various attempts were made from the 1990s onwards to deal with adult incarceration. Illustratively, a Federal–Provincial–Territorial working group was established in 1995 whose explicit and sole responsibility was to develop a plan to address the problem of correctional population growth (Federal–Provincial–Territorial Ministers Responsible for Justice 1996). Their views were obvious. The report contained a full-page bar graph of imprisonment rates comparing Canada's rate in 1992/3 to those of Australia, New Zealand, and 10 European nations. Canada had the highest rate. The incarceration rates of three high-imprisonment countries—South Africa, United States, and Russia—were reported in a box at the bottom of the bar graph. The message was clearly that Canada's incarceration rate was high compared to other reasonable nations.

The working group developed recommendations, apparently supported by all members. Their proposals included making use of diversion programs, employing community sanctions for low-risk offenders, increasing the use of restorative justice approaches and giving prison administrators more flexibility to release prisoners. Notably though, all recommendations were merely aspirational. Furthermore, though the working group's report was published, it received almost no attention.

However, it was not the case that adult imprisonment was being ignored during this (and the subsequent) era. During the same Parliamentary session (first session, 35th Parliament) in which the Justice Minister introduced the (aforementioned) YOA amendments, he tabled a comprehensive (adult) sentencing bill. Partially modeled on two earlier sentencing bills introduced (but not passed) in 1984 and 1990, Bill C-41 was designed largely to codify the law as it had developed. However, it also introduced several new provisions, some of which were explicitly designed to reduce—or at least cap—imprisonment. For example, in addition to formally legislating proportionality in sentencing, it stated that

Though this section might appear to endorse restraint, words like “should,” “unduly,” “may be appropriate,” “reasonable,” and “considered” would arguably allow almost anything. Notably, the adult sentencing provisions contain nothing resembling the YCJA “hurdles” to imprisonment, nor other directive provisions to structure criminal justice decision making.

The choice to adopt aspirational—not operational—language may reflect political ambivalence to decarceration which seemingly characterized this (and the subsequent) era. Symptomatically, the Minister reflected the political concern of being seen as “soft on crime” when introducing the Bill. Specifically, he noted early in his speech that

His prior strategic avoidance of appearing “soft” when introducing amendments to the youth justice law by targeting the few serious violent offenses (i.e., murder) is conspicuously absent. Instead, a much wider punitive net is cast for adults. Similarly, he subsequently mentioned “the will to protect society, to assist in rehabilitating offenders…” before mentioning proportionality and the bill's statement that “when appropriate, alternatives must be contemplated, especially in the case of Native offenders”.

Indeed, the first part of the speech focused largely on circumstances in which harsh sentences could be given. Furthermore, he noted that the bill tightened up a controversial provision on the parole ineligibility period for those convicted of murder. Yet, seemingly to emphasize his ambivalent position, he stated that although “[i]ncarceration must remain an option for offenders who need this form of punishment and must be separated from society to ensure the safety of the population, [i]t is worthwhile to remind the House that Canada's incarceration rate is extremely high compared with other industrialized countries.” And he continued:

Noting that only a small proportion of crime in Canada involves violence, he concluded that

Compared to his fairly explicit message that youth custody rates were too high, the statements concerning adults were less clear. We are not suggesting that the Minister's lack of clear direction in his speech accounts for the differing trends in youth and adult incarceration rates in the following years. Rather we are proposing that his more guarded speech about adults reflected the weaker political consensus about limiting the use of imprisonment for adults compared to youths.

In brief, the reduction of adult imprisonment—in contrast with youth—was not a key part of the legislative agenda. The sentencing bill reflected Canadian political leaders' ambivalence concerning adult incarceration. This uncertainty contrasts dramatically with the political will to reduce youth imprisonment. Importantly, the unambiguous political support for reducing the incarceration of youths likely led to the development of operational directives within the YCJA which brought about youth decarceration. For adults, the message—and, we suspect, the political resolve—are more mixed. By extension, the continued use of aspirational legislative language persists. Canada may have managed to “contain” adult imprisonment but has shown no clear political will to reduce it.

We have not addressed two other important “counterfactuals” to Canada's dramatic decarceration of youth—the substantial variation in the use of custody across provinces/territories and the overrepresentation of Indigenous youth. This decision reflects both our current focus on macro-level trends in Canada's youth population and especially our recognition of the (theoretical, methodological, and practical) complexities of these additional aspects of Canada's youth justice system which merit their own detailed analyses. Illustratively, Figures A1 and A2 present the trends in admissions to sentenced custody for two adjoining provinces (Ontario and Manitoba). Notably, the shapes of the curves are markedly different whereby the overall drop in Ontario is 79 percent while that of Manitoba is merely 22 percent. Clearly, changes in the law cannot explain this interjurisdictional divergence. More disturbing, while Indigenous people constitute 3 percent (Ontario) and 18 percent (Manitoba) of their populations, they accounted for 14 and 85 percent, respectively, of sentenced admissions in 2016/17. Although both groups may have benefitted from the changes that we have described (with the number of youth sentenced admissions having declined in each province between 2004/5–2016/17), this picture hides considerably more than it reveals. Clearly, more thorough study is needed.

Part VI—Concluding Remarks: Lessons from Youth about Decarceration

This paper is a story about change. It undertook to understand the “why” and “how” of the dramatic reduction in Canadian youth imprisonment rates. Unsurprisingly, the answers to these questions are intrinsically linked. The new youth legislation promoting decarceration constituted both a result of wider political, sociocultural and historical forces underway a decade before its enactment and a cause of the (immediate and long-term) decrease.

More technically, Canadian youth decarceration was the result of three conditions/factors. First, restraint in imprisonment was firmly established as part of Canadian core values. The deleterious effects of incarceration were widely recognized, as was the ineffectiveness of prison as a crime control strategy. While criminal justice policy and practices still flirted with the use of custody for deterrence/incapacitation and rehabilitative purposes, key policy makers were defining “good” social policy as being linked to low rates of youth imprisonment. Restraint in the use of incarceration was a fundamental value in the criminal and youth justice cultures.

Second, while values consistent with decarceration were necessary to bring about change, they were insufficient. Decarceration also required unambiguous political willingness to support change. This condition was facilitated by the general consensus—by the mid-1990s—that youth justice legislative reform was needed. Despite several amendments, the YOA continued to be deeply unpopular with both the public and most provinces/territories. However, it only became politically “safe” to promote decarceration because of two additional events/processes. On the one hand, the government adopted a bifurcated model which would be tough on violent and serious repeat offenders while being soft on minor offenders. On the other hand, it was widely asserted by the mid-1990s that Canada had one of the highest youth incarceration rates in the world—a “fame” generally seen as shameful and in urgent need of rectification.

Third, although the combination of these factors/events allowed for real change, that “change” still needed to be crafted. Hence, the new youth justice legislation operationalized decarceration in ways that promoted its realization. The YCJA made it “law” to reduce the use of youth court and custody by explicitly structuring decisions through prescriptive provisions. Decarceration did not just happen. It happened because explicit policies were implemented to ensure its realization. Further, a concern with education and training helped to ensure the effective application of the law while financial incentives to the provinces/territories enabled the system itself to change.

And so decarceration came about in Canada. As the legislation and the political calls to reduce youth imprisonment became more explicit in the late 1990s and early 2000s, the size of the reduction in youth custody increased. While the largest decline occurred with the YCJA's enactment, use of youth court and imprisonment continued to decline after 2003. This sustained impact arguably reflects the creation of a new working culture. Criminal justice professionals hired more recently are trained within these new parameters, promulgating and perpetuating lower levels of incarceration as the normal way of “doing business”.

This achievement contrasts with the failure to reduce youth pretrial detention and adult imprisonment.Footnote ²⁵ It is not coincidental that the legislation governing these processes was merely aspirational in nature. Finland is a clear example that explicit directives are necessary. Reference Lappi-SeppäläLappi-Seppälä (2007: 234, table 1) lists 24 policy (and legislative) changes that transpired between 1996 and 2006 to reduce Finland's adult imprisonment rate. The ensuing decarceration efforts with respect to adults in California in the 1960s (Reference Gartner, Doob and ZimringGartner et al. 2011) or in Alberta, Canada in the 1990s (Reference Webster and DoobWebster and Doob 2014) constitute further examples of the importance of very specific legislation and/or policies whose explicit purpose was to reduce imprisonment.

However, the contrast of the dramatic drop in youth imprisonment with adult stability also highlights the importance of unambiguous political will in bringing about change. With youth, it became sufficiently safe to attempt decarceration (even though it might be important to impose a political spin on the legislation). These conditions were largely absent with adults. While having adopted a proportionality model at sentencing which places natural limits on the (excessive) use of imprisonment, the adult justice system has continued to flirt with utilitarian objectives encouraging incarceration. Furthermore, Canada's relative stability in imprisonment over the recent decades may have—ironically—limited its ability to be shamed into decarceration. At least in comparison with English-speaking nations whose incarceration rates increased, the “call to arms” for decarceration that the Canadian youth system experienced in the late 1990s would be almost impossible to replicate in the adult system.

But unambiguous political support for decarceration requires more than political safety. More important is what is at the root of this political will. Within this context, Reference BatemanBateman's (2017: 60) discussion of the criteria necessary to achieve long-term youth decarceration is compelling. He argues that sustained reduction in the use of court and custody cannot be driven by pragmatic accommodation to changing political and fiscal priorities. Rather, they must be rooted in “child-friendly, rights-compliant, philosophically coherent and evidence informed youth justice policy and practice that eschews short term punishment as a rationale for intervention.” In Canada, the principles of diversion and decarceration form the basis of current youth justice policy. Notably, they have been promoted on their own merits precisely because of their perceived ability to achieve the best long-term outcomes for youths rather than as pragmatic mechanisms that are politically or fiscally advantageous.

This principled approach may be a key to sustained success. Similar cases of long-term decarceration would support the importance of perceiving over-reliance on incarceration as morally wrong. The substantial reductions in the prison populations of Finland (1960–2000) and the Netherlands (1945–1973) seemingly coincide with changes in cultural attitudes whereby high imprisonment rates were seen as disgraceful (Finland) or inhumane (Netherlands). As Reference TonryTonry (2011: 647) reminds us, “[i]n countries in which imprisonment is widely believed to be a Bad Thing, policy makers will work to restrain its use”.

The counterfactuals are equally telling. The shorter-term reductions in imprisonment rates in California during the later 1970s and Alberta during the early 1990s were largely driven by economic imperatives. Similar concerns have been raised relative to the recent youth decarceration in England/Wales. Reference BatemanBateman (2017: 59) worries that the significant reduction in youth imprisonment may be stalled “without further systematic or philosophical change”. Although the recognition that recent declines in the incarceration of American youth are partially attributed to cultural change (rooted in the recent return to a treatment orientation) is promising, the parallel justification rooted in the high cost of youth imprisonment is more disconcerting.

More broadly, Reference ZimringZimring (2005) is almost certainly correct in recognizing that, as a society, we can treat youth differently from adults, at least in terms of imprisonment. In Canada, this license has found explicit translation in the YCJA (s.3) which lists diminished moral blameworthiness or culpability as an explicit justification for having a separate youth justice system. Furthermore, it has allowed this legislation—in contrast with the adult Criminal Code—to redefine what constitutes “good” criminal justice policy and practice within a philosophically-coherent, rights-based model. This principled approach (and its accompanying policies and practices) has, thus far, served Canadians well in terms of achieving—and sustaining—long-term decarceration.

While the technical or operational processes underlying reduction in the use of court and custody (e.g., less police charging; greater use of diversion; less Crown prosecution and less recourse to prison) are easily transferable to other nations, we would suggest—as have others before us (Reference TonryTonry 2017)—that the challenge for any nation hoping to decarcerate lies in creating the political, cultural and, most importantly, the moral will to implement change. To this end, reconsideration of the role of prison as an effective crime control strategy might be an initial step. Certainly the mounting criminological literature supplies little empirical support for this belief while also underlining the nontrivial (opportunity and financial) costs of the persistent recourse to incarceration.

But such instrumental concerns (already reasonably well entrenched in Canadian criminal justice culture) will likely need to be coupled with more normative arguments. As the failure to reduce Canadian levels of adult imprisonment may underscore, prison must also be seen as a “Bad Thing” whereby high rates reflect a society's moral failure. Templates or roadmaps for this type of change in heart are almost certainly country-specific and have proven—at least in many English-speaking nations—elusive. But it is not impossible. As Canada's story of decarceration underlines, change is possible as long as we have the will to make it happen.