Trends in the imprisonment of women in Canada.

• Author: Gartner, Rosemary

Increasing punitiveness is seen as a defining feature of late-modern liberal democracies by scholars who cite the growth in prison populations in the United States, England and Wales, the Netherlands, and New Zealand as evidence of this expansion in state punishment (see, e.g., Pratt 2007; Roberts, Stalans, Indemaur, and Hough 2003). However, detailed case studies of imprisonment in other countries demonstrate that this growth is much less general than is often claimed. As Tonry (2007: 1) notes, in his introduction to a volume featuring such studies, "Imprisonment rates have not risen substantially everywhere in the last 15 years." For example, Finland, Japan, Germany, Belgium, France, and Canada have experienced neither sustained nor dramatic escalations in their prison populations.

Indeed, in Canada, the imprisonment rate at both the federal and provincial levels has been relatively stable over the past 45 years (Doob and Webster 2006; Webster and Doob 2007) (see Figure 1). However, as Webster and Doob (2007: 312-313) note, their conclusion that "imprisonment rates have not changed dramatically since 1960" reflects the pattern for all offenders. Because males account for about 95% of Canada's prison population, (2) the trends shown in Figure 1 are driven largely by changes in the male prison population and, therefore, may tell us little about whether or how the female prison population has changed over time.

Many would argue that the stability described by these overall numbers masks a different and disturbing pattern for females. For example, Kim Pate, the executive director of the Canadian Association of Elizabeth Fry Societies recently stated to a Parliamentary Committee that "[w]omen are the fastest growing prison population [in Canada]" (Canada, House of Commons 2007; see also Balfour 2006a; Elizabeth Fry Society 2008). Boritch (2008: 406) agrees that "Canada has seen a significant increase in the punishment and incarceration of women over the past few decades," an increase that, according to Robert, Frignon, and Belzile (2007: 177), mirrors the growth "in many other western countries." While some see a different pattern--for example, Kruttschnitt and Gartner (2003: 16) conclude that "female imprisonment in Canada in the 1990s [did] not follow the upward trend observed in the United States and in England and Wales"--this appears to be a minority view. This paper is an effort to adjudicate between these competing claims by providing a detailed description of the scope of women's imprisonment in Canada since the early 1980s. Our goal is to determine whether the imprisonment of women in Canada has, as many argue, increased in scale, even as the overall imprisonment rate has not.

[FIGURE 1 OMITTED]

We begin by considering possible reasons for the view that women's imprisonment has increased in Canada. We then describe the different ways in which imprisonment trends can be measured and argue that the common practice of measuring trends in women's imprisonment relative to men's imprisonment can lead to misleading conclusions. In the analyses that follow, we present a variety of data on women's imprisonment to determine whether there has been an increase in the scale of women's imprisonment in Canada, as some have claimed. We begin by examining, separately, trends in the sentencing of women to federal and provincial imprisonment. Here, we see some evidence of growth in women's imprisonment in federal institutions. However, those sent to federal institutions account for only about 2.9% of all sentenced women admitted to custody in Canada. We then analyse combined data on the sentencing of women to federal and provincial imprisonment, because our aim is to provide a picture of the total population of women in penal confinement in Canada, regardless of where and for what reasons they are in custody. In combining the federal and provincial data, we also follow the practice used by the Canadian Centre for Justice Statistics to calculate Canada's total imprisonment rate and to describe adult correctional services in Canada (see, e.g., Beattie 2006). The claim that women's imprisonment in Canada has increased in the last few decades receives no support when we analyse combined provincial and federal sentenced-admissions data. In the final section of our analysis, we identify-for one province (Ontario)--an important shift in the composition of the incarcerated female population in very recent years: an expansion in the number of unsentenced women in custody. We conclude by urging greater attention to the troubling trend toward imprisoning growing numbers of people who have not yet been tried or convicted.

As noted above, we argue that trends in women's imprisonment deserve analysis in their own right and not simply in comparison to trends in men's imprisonment. Nevertheless, assessing the meaning of trends in women's imprisonment depends to some extent on trends in men's imprisonment. For example, stability in women's imprisonment rates may be cause for concern if men's imprisonment rates have been on the decline. We return to this issue in our conclusion.

Reasons for the view that women's imprisonment has increased

The presumption that women's imprisonment in Canada has increased would be easy to understand were there clear empirical evidence showing a growth in the size of the female custodial population. However, comprehensive time-series data on trends in women's imprisonment in Canada have not been compiled and analysed up until now. We discuss reasons for this below.

In the absence of such evidence, there are at least three factors that could account for the belief that women's imprisonment has grown. The first relates to the claims, noted in our opening paragraph, about trends in imprisonment in liberal democracies more generally. Simply put, the increase in imprisonment in the United States, England and Wales, New Zealand, and elsewhere could understandably result in an assumption that the punitive turn in these countries is also reflected in imprisonment trends in Canada (see, e.g., Pratt 2002: 177). This assumption, in turn, might lead to a belief in an increase in women's imprisonment.

The second and third factors that could account for this belief are specific to women's imprisonment. Since the late 1980s, most of the attention on women's imprisonment in Canada has focused on federally sentenced women. For example, the 1989 Task Force on Federally Sentenced Women (TFFSW) and Creating Choices, the 1990 report it produced (TFFSW 1990), were widely covered by the media and have been the subject of a number of books and articles (e.g., Hannah-Moffat 2000; 2001; Hayman 2001; Balfour 2006b; Boritch 2008). In addition, the events at the Kingston Prison for Women in 1994--which included the strip-searching and shackling of female prisoners by an emergency response team composed solely of males--were met with national attention and a commission of inquiry headed by Madam Justice Louise Arbour. In the same year that Madam Justice Arbour released her report (Arbour 1996), a series of slashings, escapes, and suicide attempts at some of the new regional federal prisons for women further heightened concerns about federally sentenced women.

In this context, and given that relatively detailed data on federal imprisonment are easily available, it is not surprising that federally sentenced women have been the subject of more academic analysis and popular attention than provincial sentenced women. Because--as we will see shortly--the number of women in federal prisons has increased since the early 1980s, some may have concluded that there has been an overall growth in women's imprisonment in Canada.

A third factor encouraging the belief in an increase in women's imprisonment in Canada has to do with how changes in women's imprisonment have been measured. Often, these are expressed in terms of changes in men's imprisonment. In the next section, we give examples of how the use of such comparative measures can and has lead to misleading conclusions about trends in women's imprisonment. Before doing so, we need to discuss the various ways in which trends in women's imprisonment can be measured.

Measuring trends in women's imprisonment

Two types of data can be used to measure trends in women's imprisonment: (1) admissions of women to prison in a year; and (2) a count of women in prison on a designated "census" day or the average number of women in prison each day in a given year. The former are referred to as admissions (or flow) data. The latter are referred to as count (or stock) data. Some argue that admissions data are an appropriate measure of a nation's propensity to incarcerate its citizens (e.g., Lynch 1995; Sutton 2004). Others view count data--on which imprisonment rates are usually based--as a good indicator of the overall punitiveness of a nation (e.g., Webster and Doob 2007). (3) Ideally, both measures should be used to describe trends in imprisonment. (4)

Admissions and count data are a function of somewhat different criminal justice processes. For example, policies that lead to a large number of very short prison stays can affect prison admissions quite dramatically but have little impact on prison population counts. Because count data are determined not only by the number of people who enter prison but also by how long they stay there, policies that increase imprisonment length can have large effects on count data but no effect on admissions data. Compared to count data, admissions data will include a higher proportion of those with short sentences. As a consequence of these differences, admissions and count data are not comparable and using them interchangeably can lead to inaccurate conclusions. An example of this is the claim made by some scholars that the female imprisonment rate in Canada is higher than the rate in the United States (see, e.g., Boritch 1997; Linden 2000: 166). The basis for...