Introduction
Combating violence against women has been a key policy issue for Canadian women since the early 1970s when grassroots groups established the first transition houses and shelters to meet local demand. Since then, advocates have lobbied the state for funding for adequate service delivery, better and more effective laws to protect women victims and more attention for the root causes of violence, including women's structural societal inequality. The majority of these demands require heightened state involvement and increases in public spending. However, governments in Canada and in other Western democracies have increasingly adopted neoliberal approaches to the welfare state since the late 1970s and early 1980s. This neoliberal approach encourages less state involvement and reduced public spending in order “to put an end to a perceived culture of welfare dependence, and to reinvigorate the nation by giving free rein to individuals' own entrepreneurial proclivities” (Kendall, 2003: 6).
Even though many feminist political scientists have argued that this welfare state retrenchment has been disproportionately devastating to women's lives as women are more likely employed by the state and more dependent on welfare state programs than men (Brodie, 1996; Bashevkin, 1998), some have questioned whether neoliberalism has negatively impacted state willingness to address the issue of violence against women. According to S. Laurel Weldon, some see violence against women policies as “mainly symbolic measures that involve little redistribution. For this reason, they present an opportunity for right-wing or neoliberal governments to mollify women's organizations without spending any money” (2002: 58). Yet Weldon herself questions this position noting examples of neoliberal regimes that have refused to address the anti-violence issue because they saw it as a private instead of a public matter. Still other right-wing administrations, particularly the Republican-controlled Congress in the United States, authorized nearly a billion dollars in anti-violence expenditures in 1998 in an era of budget cutting and deficit reduction (2002:59).
Since anti-violence measures generally cost less than other welfare state programs,1
This is particularly true with respect to state child care expenditures. See Collier (2006).
Lakeman argues that both of these approaches individualize the issue and remove the “women” from violence against women policy and therefore solutions are not aimed at more societal structural causes of inequality. She also argues that deference to law enforcement agencies perpetuates “the patriarchal order” in society—a main structural cause of violence against women (Lakeman, 1999: 28).
This article seeks to answer these questions in the Canadian context, focusing specifically at the provincial level. I have chosen to focus on the substate level because fiscal responsibility for social program delivery has been steadily downloaded from the federal to the provincial level—a reflection of neoliberal trends. This was evident in the 1990 cap on Canada Assistance Plan (CAP) payments to the three “have” provinces (Alberta, Ontario and BC) and the reduction of transfer payments through the introduction of the amalgamated Canada Health and Social Transfer, which replaced CAP in 1996–1997.3
Note the CHST was split into the Canada Health Transfer and the Canada Social Transfer in 2004.
The phrase “violence against women” is often used to cover a wide variety of violent acts perpetrated against both women and children, including child abuse. This article does not use the phrase to describe children's experiences of violence even though the author recognizes that often the violent situation in a home affects both the mother and the children. For more on this, see Gotell (1998).
Even though anti-violence services and advocacy began at the local grassroots level, less scholarly attention has been paid to provincial violence against women policies. This article aims to shed light on this important level of women's movement activity and to add to our wider understanding of the impact of welfare state retrenchment in the provinces and how this directly impacts women's lives. To do this, I have chosen to compare provincial anti-violence policy expenditures and program changes between 1985 and 2005 in Ontario and British Columbia. I will argue that, although aggregate expenditure statistics show that most governments have, over time, increased funding for anti-violence programs, there has been much more variation in how responsive this spending has been to feminist anti-violence advocates during this 20-year period. Policy variation can best be explained by the partisan theory of public policy, which argues that ideological differences between different party governments explain diversity in public policy directions (Castles, 1982; Hicks and Swank, 1992; Schmidt, 1996). The article further argues that, although party differences have been somewhat muted during neoliberal times, significant differences in left- and right-wing approaches to violence against women were still present. These differences suggest that the impact of neoliberalism on anti-violence policy is often more pronounced under right-wing regimes than left-wing ones. Thus, neoliberalism's policy impact cannot be fully determined outside of the partisan framework, as different party governments had the ability to diverge in their responses to anti-violence movements and were not necessarily beholden to negative pressure associated with welfare state retrenchment convergence.
In order to illustrate these arguments, the article will begin by briefly outlining the theoretical debate between welfare state retrenchment convergence and the partisan theory of public policy. It then will describe the methodological choices for the comparative study, compare anti-violence policy in the two provincial cases and conclude by directly addressing the questions posed above.
Welfare State Policy Convergence and the Partisan Theory of Public Policy
There is solid evidence of general welfare state convergence in the comparative public policy literature. Sylvia Bashevkin's comparative study of women's policy in the US, UK and Canada concluded that welfare state retrenchment continued past its origins under neoconservative regimes of the 1980s well into the 1990s and 2000, even though supposedly more “moderate” governments held power during these years (2002: 14).5
Bashevkin's study recognizes diversity in government approaches, but still affirms the existence of tangible welfare state convergence despite these differences.
However, Colin Bennett has argued that sometimes policy convergence conclusions are drawn because studies fail to look closely at the salient public policy details. Thus, “aggregate cross-sectional studies … in some ways resemble photographs taken from a high-flying aircraft; the main features stand out, but much [important] detail is lost” (1991: 219). If we look closely enough at this relevant detail and see more policy divergence instead of convergence, this can best be explained by the partisan theory of public policy. The theory that party ideology or ideas explain why governments make different policy decisions is supported by authors such as D.A. Hibbs, Jr., who argues that party composition is often the main cause of variation in policy outputs and choice in constitutional democracies (Schmidt, 1996: 155).6
When researching women's policy, I argue that it is essential to look at the finer details of state policy responses not just to establish whether spending increased or decreased, but to determine how well the policy output addressed long-standing women's movement demands in the particular policy arena.7
Elsewhere I argue that a measurement of steady welfare state expenditure increases over a significant period of time will not provide a nuanced enough picture of what policy decisions are being made: a government may increase its welfare spending overall while decreasing spending on unemployment insurance and simultaneously increasing its health care budget (Collier, 2006). Similarly for anti-violence policy, a focus on aggregate expenditure statistics alone will not illuminate whether or not any of the responses are feminist or anti-feminist in nature or how well they respond to advocacy demands. In order to avoid false assumptions that all increases in expenditure on anti-violence policy are positive and similarly that all decreases in expenditure levels are negative, I use both quantitative expenditure statistics alongside qualitative program and policy analysis to provide the full picture of anti-violence decision making over time.
By comparing anti-violence policy development between 1985 and 2005 in Ontario and BC, this article can test whether significant provincial policy convergence was evident or whether different ideological governments responded in unique ways to feminist anti-violence demands. The latter would reaffirm claims made by the partisan theory of public policy and could lead us to question the existence of widespread welfare state retrenchment convergence.8
Convergence is defined as the “tendency of societies to grow more alike, to develop similarities in structures, processes and performances” (Kerr, quoted in Bennett, 1991: 219). Therefore it has been used to describe increasingly similar policy directions among states over time, not simply a softening of ideological positions at particular points in time. Thus if diversity is sustained over time as opposed to a more singular policy direction, this would help dispute a theory of convergence.
Comparing Ontario and BC
To uncover the impact of neoliberalism on anti-violence policy, this paper will comparatively measure aggregate changes in provincial government anti-violence expenditures and the qualitative progression of anti-violence policies and programs9
For the purposes of the qualitative policy and program analysis, only provincial government initiatives directed specifically toward the issue of violence against women have been included and rated in this article. Although I recognize and briefly mention the impact that broader welfare state cuts can have on victims of violence, these broader policy initiatives are not included in the more detailed analysis due to time and space limitations. These other initiatives, which purport to be gender neutral, are indeed important to the anti-violence issue (as are more gendered policy areas, including availability of child care, for example), but they do not necessarily speak directly to the major focus of this particular study, state decisions regarding the problem of violence against women.
Ontario Governments

Anti-violence policy is measured in two ways to avoid the pitfall of missed details noted by Bennett above. First, the paper measures anti-violence expenditures as a percentage of overall program spending for each province, where possible,10
For both provinces, anti-violence expenditures were only identified in the Provincial Public Accounts beginning in the 1990s. In BC, these statistics were available only until 2001.
Drawing on anti-violence advocacy interview data and internal anti-violence movement documents collected between 2000 and 2005, I classify government anti-violence policy and program initiatives as significant or as secondary announcements.11
Significant policies could be classified as positive, negative or mixed but needed to include one or more of the following to be included in this study. The policy or program needed 1) to have a significant amount of new or reduced spending, depending on the specific program in question; 2) to be a major move in a new area that had not been acted on previously by the provincial government; 3) to be identified as significant by anti-violence advocates interviewed for this study or in historical advocacy documents as a long-standing movement demand. Therefore, this study chooses to focus more on policy action by government than on policy inaction (which, depending on the circumstances, could also be considered “significant”).
Anti-Violence Expenditure and Policy in Ontario
Figure 1 illustrates the changes in Ontario government expenditure levels for anti-violence programs as a percentage of overall program expenditures between 1992 and 2004. The graph begins in 1992 because this is the first year that the province separated out anti-violence expenditures in the provincial public accounts.12
I include the Ontario Public Accounts listing for violence against women expenditures made within the Ministry of Community and Social Services. Between 1995 and 1999, I add in Sexual Assault Initiatives expenditure within the Attorney General's Office, as well as women's anti-violence services and centre expenditures in the Ontario Women's Directorate.
I also used historical advocacy and government data to compare demands over time, recognizing that advocates can be overly critical of state responses in order to maintain a strategic lobbying position vis-a-vis the state.

Ontario VAW Expenditure as Percentage of Total Programme Spending
Qualitative policy data (summarized in Table 2) show considerable variation in significant anti-violence program and policy announcements over time.14
A listing of the specific policies evaluated is found in appendix A. For more detailed discussion see Collier (2006).
Summary of Significant Ontario Government Anti-Violence Policy Responses to Movements 1985–2005

Between 1985 and 1987 under the Liberal/New Democratic Party (NDP) Accord,15
The Liberal and NDP parties signed a formal accord after defeating the minority Conservatives shortly after the 1985 election. The accord allowed the Liberals to govern for two years and the Liberals agreed to follow a written policy agreement with the NDP during this time.
Attention to anti-violence issues increased under the Liberal majority government (1987–1990), but the only significant response during these three years was the mixed $7 million increase in family violence spending, including new funding for wife assault initiatives and the first announcements of second-stage funding in the province (Walker, 1990:203). Groups such as the Ontario Association of Interval and Transition Houses (OAITH) and the Ontario Advisory Council on the Status of Women had campaigned strongly for affordable longer term or second-stage housing for battered women and were very pleased with this part of the funding announcement (OACWI, 1992; Women's Movement Archives). But according to Gillian Walker, OAITH was critical of the wide range of “traditional agencies and institutions” that would receive provincial funding, which left only a small portion for feminist organizations (1990: 15). Although specific expenditure statistics are unavailable for the years the Liberals held office, the 1987 increase in family violence spending was likely one of the largest, if not the largest, increase to date.
Responses to activists improved dramatically once the centre-left NDP assumed power in 1990. Five significant positive policy and program announcements were made during the Bob Rae NDP years, including those that advocates consistently praised as being quite responsive to the movement. Significantly, no negative policy announcements were made under this regime.16
Confidential activist interviews 2000–2002. A dozen semi-structured interviews were conducted in person and over the phone (in Ontario and BC) around the anti-violence issue and governments' responses to it. Interviews lasted one hour on average, with more open-ended follow-up questions rounding out the discussion. Of the fifteen initial questions posed, five specifically addressed policy demands and reaction to policy decision making. Transcripts are on file with the author. See Collier (2006).
Like previous governments, the NDP tended to adopt a gender-neutral law-and order-approach to fighting the problem of violence against women, which was generally viewed as a mixed result by advocates. However, the NDP government tempered this approach which represented a departure from other provincial regimes. For example, in 1991 the NDP attorney general directed crown attorneys to make every effort to fight attempts to bring up a victim's sexual history during sexual assault trials; this move was consistently praised by activists (Women's Movement Archives).
Eventually, NDP attention to anti-violence issues cooled somewhat during the party's final years in office as economic concerns took precedence. There were no new anti-violence policy announcements in 1994 and only one announcement in 1993; neither was significant. Even so, an overall look at policy evidence clearly demonstrates that the NDP government was the most committed to a pro-feminist agenda against violence in Ontario to date, despite broader neoliberal pressures. This was illustrated by the fact that no cuts to anti-violence programs occurred during a tough provincial recession.
When the right-wing Mike Harris Conservatives took power in 1995, government responses, not surprisingly, turned negative. Five negative responses were recorded during this period, including the first recorded provincial cuts to anti-violence program funding (Collier, 2006). These included significant cuts in 1995 to the entire Ministry of Community and Social Services budget for second-stage housing, education and prevention services, male batterer counselling programs and culturally specific anti-violence services. The Conservatives cut interval and women's shelter funding by 2.5 per cent in 1995 and another 5 per cent in 1996. Eileen Morrow of OAITH predicted that the cuts would have “deadly consequences.”17
Quoted in Lightman and Baines (1996: 150).
Anti-violence expenditure percentages decreased slightly between 1995 and 1997, in all likelihood for the first time. Even though the economy was nearing the end of a tough recession, the targeting of anti-violence programs did not in fact save the province much money because the programs were not worth much in the first place. Cuts to anti-violence expenditures only accounted for .002 per cent of total government program spending between 1994 and 1997 (see Figure 1). This suggests that the cuts were more ideologically aimed against the anti-violence movement, particularly since they were targeted toward feminist anti-violence service delivery (OAITH, 1997).
In 1996, the Conservatives decided to re-examine anti-violence policy and drew up terms of reference for what they called the Framework for Action on the Prevention of Violence against Women in Ontario (the McGuire Report). Increased publicity surrounding the issue of violence against women put it into sharper focus for the Conservative government, but the movement was not invited to comment on the government's new anti-violence agenda. Advocates were particularly worried about the report's view of women's shelters and rape crisis centres as helping to create a victim-centred dependency on services instead of empowering women (OAITH, 1997: 2–3) and about its implied support for further cuts to services (OAITH, 1997: 3).
In 1998, the chief coroner of Ontario launched an inquest into the high profile murder of Arlene May by her male partner. Advocates were pleased with the 213 recommendations released by the inquest later that year, many of which echoed their demands, but the Conservatives largely ignored the report in the belief that the government was already implementing most of the coroner's recommendations. This statement was made even when a prominent judge on the inquest publicly stated that this was not the case (Canadian Press NewsWire, April 4, 2002). The 1998 May inquest report and three high-profile murders of women in the Toronto area in just over a week's time in 2000 heightened media attention and public pressure between 1998 and 2000.18
This prompted the chief coroner of Ontario to announce yet another inquest into one of the spousal murders—that of Gillian Hadley in 2000. The movement was baffled by the announcement arguing that the government still had not addressed the issues raised in the May inquest (OWJN, 2000a).
However, while the restoration of funding and increases, particularly in 2000, were significant, the corresponding policy responses were not. Money was not directed to feminist anti-violence services, but instead to gender-neutral law-and-order and victims-rights programs. Thus, feminist activists were still struggling to provide front-line anti-violence services, despite increased government anti-violence spending commitments. Bill 117, The Domestic Violence Protection Act introduced late in 2000, continued a law-and-order focus. However, the movement was initially pleased with the fact that the act appeared to address some of the recommendations made by the May inquest. Yet, despite a speedy passage through the legislature, the Conservatives dragged their heels on proclaiming and implementing the law well into 2002 (OWJN, 2002). At the same time, the Conservatives cut funding to five women's centres because they were offering second-stage services to women who had already escaped violent homes (Canadian Press NewsWire, Oct. 2, 2000). Ostensibly, the Conservatives did not believe that government should be supporting these types of services.19
Three Conservative cabinet ministers met with members of the Cross-Sectoral Violence against Women Strategy Group in 2000 and made it clear that second-stage housing was off the table and “would not come back onto the table for as long as this government is in power” (OWJN, 2000b: 2).
Although the Conservatives continued to spend new money on anti-violence programs, expenditure percentages actually shrunk between 2001 and 2003. New money was welcomed by advocates but the lack of attention to feminist demands continued. The only significant positive response during these years was a much-needed increase of $4.5 million to expand a province-wide 24-hour crisis helpline (Leslie, 2004).
When the Liberals replaced the Conservatives in 2003, there was room for improvement in state responsiveness levels to anti-violence movements, particularly in recognizing feminist expertise on the issue. However, even though the Liberals continued to increase spending and maintain the percentage level of expenditure, they failed to dramatically improve state openness to advocate demands. After a small increase in sexual assault and shelter funding in 2004, the Liberals announced a major $60 million Domestic Violence Action Plan in 2005. Although the funding increase was welcomed, a continued focus on the gender-neutral term “domestic” violence instead of “violence against women” was a disappointment to advocates who had higher expectations from the centrist government. More disturbing were comments made by Premier Dalton McGuinty and Sandra Pupatello, the minister responsible for women's issues, that shelters would receive one-time funding and then were expected to become “financially independent”—an untenable situation for shelters that were already struggling to raise funds to offset inadequate government support (OWJN, 2005).
In the end, we can see much diversity in Ontario government responses to anti-violence advocates between 1985 and 2005. While there was evidence of cuts and retrenchment of anti-violence programs during the Harris Conservative years, all governments appeared willing to increase anti-violence expenditures which left the percentage of expenditure consistently between 0.13 and 0.16 per cent of overall program expenditure between 1992 and 2004. Yet the most positive significant announcements (five) occurred under the left-of-centre NDP, which proved to be the most open to feminist anti-violence approaches. By contrast the Conservatives had five negative and three mixed significant policy announcements and only one that was rated positive. The Liberals were mixed in both of their significant responses to the movement at different time points during the study period. Thus, the evidence for Ontario did not show consistent retrenchment convergence across governments with respect to anti-violence policy. Variation in state responsiveness levels, for the most part, seemed to indicate that centre-left governments were more open to feminist demands than right-wing regimes, with centrist governments falling somewhere in between.
British Columbia Governments

Anti-Violence Expenditure and Policy in BC
Anti-violence policies and program expenditure levels in Figure 2 demonstrate some similarities to data recorded for Ontario in Figure 1. Particularly, the percentages of expenditure did not fluctuate dramatically over time, as was the case in Ontario. However, specific anti-violence expenditure data were left off of the provincial public accounts in BC after the Liberals took office in 2001.20
In BC, expenditure data from the Ministry of Women's Equality's “Stopping the Violence” initiative is included alongside any other identifiable anti-violence program-spending, including the Intervention and Prevention of Violence against Women expenditures listed in 1992–1993.

BC VAW Expenditure as Percentage of Total Programme Spending
Summary of Significant BC Government Anti-Violence Policy Responses to Movements 1985–2005

Qualitative policy data also show variation in anti-violence program and policy announcements as in Ontario. Of 12 significant policy responses, eight were positive (67 per cent), one was mixed (8 per cent) and three were negative (25 per cent).21
The higher percentage of positive announcements in BC versus Ontario can be attributed in part to the longer tenure of the BC NDP in power as opposed to the Ontario NDP and the fact that earlier Social Credit governments chose not to respond to the issue in significant terms until 1989–1990.
Just prior to the period of study, in 1984, the right-wing Social Credit Party under Bill Bennett introduced a provincial policy covering wife assault. This policy was first announced in the 1984 BC submission to the Federal/Provincial/Territorial Working Group on Wife Abuse and was largely inspired by the working group itself as well as the fact that other provinces (such as Ontario) had already enacted a similar policy. The wife assault policy encouraged police forces and the justice system to treat domestic violence, particularly wife assault, more seriously and to lay more charges. However, this policy statement was not backed up with government action. Thus, advocates charged that both the 1984 (and later the 1986) revisions to the Wife Assault Policy were largely ineffective (Kachuk, 1998: 4).
New Social Credit leader, Bill Vander Zalm had a weak track record with advocates earlier when he held the Human Resources portfolio under Bill Bennett, but his desire to bolster Social Credit support among females in the electorate after he became leader and premier led to two positive anti-violence announcements in the early 1990s. In 1990, the Social Credit government established the Advisory Council on Community-Based Programs for Women and announced a 25 per cent budget increase in shelter funding; in 1991 it established the Task Force on Family Violence (BC Ministry of Government Management Services and Ministry Responsible for Women's Programs 1991). Both were positive responses to the anti-violence movement after years of mainly inaction under the Bennett Social Credit government.
After the NDP took power in 1991, the report of the BC Task Force on Family Violence22
The task force report was well received by the movement except for a critique of the original terms of reference which meant it would look broadly at “family violence” instead of at the more obvious problem of violence against women (Report of the BC Task Force on Family Violence, 1992: 37).
As well, the report called for increased front-line services, better training for those offering the services, and better funding. See Report of the BC Task Force on Family Violence, 1992.
Confidential interview 2000.
The NDP program and policy agenda that followed the task force shared a willingness to respond well to the recommendations in the report. The movement particularly welcomed the large 1992 increases in government funding of $10 million per year over the next four years under the Stopping the Violence Initiative. The 1992 announcement of core funding for the BC/Yukon Society of Transition Houses as well as a reinstatement of core funding for the Vancouver Transition House25
Funding had been cut by the Bennett Social Credit government in the late 1970s.
The NDP revised the existing Wife Assault Policy in 1993, even though it continued the law-and-order approach that had prevailed under the Social Credit government. The new Violence against Women in Relationships (VAWIR) policy recognized that domestic violence is not exclusive to marriage and led to the change in name and focus from “wife assault” to “violence against women in relationships.”26
See Kachuk, 1998 for more on the VAWIR policy, including movement critiques.
The pace of anti-violence policy improvements slowed somewhat between 1995 and 2000. While the majority of the NDP government's program and policy responses were positive (nine), none of these was considered very significant by movement actors.27
Because the responses were rated as not significant, they are not included in Table 4.
Figure 2 shows that anti-violence percentage of expenditure marginally decreased between 1996 and 1999 under the Glen Clark NDP. However, in 2000, the Ujjal Dosanjh NDP government increased expenditures to their highest levels since 1995. The increased expenditure does not, however, coincide with significant new policies or programs, as noted above. It is important to note here as well that even at its height in 2000, anti-violence expenditure only made up slightly more than .2 per cent of overall program spending, which was only .03 per cent higher than levels in 1995. Likely fiscal pressures, including the fact that BC became a have-not province in 2000, had an impact on the NDP agenda and tempered progressive tendencies.28
Yet advocates saw the provincial government become much more negative in its approach to anti-violence policy after the election of the right-wing Gordon Campbell Liberals in 2001. All of the significant negative policy announcements in Table 4 occurred under this regime, including the first substantial cuts to anti-violence services in women's centres and a 50 per cent cut to male batterer counselling programs in 2002. The extent of the cuts29
These included cuts in legal aid in 2001 and cuts to welfare rates in 2002, which advocates argued would particularly hurt women victims of violence. See BCIFV, 2002.
This lack of commitment to women's anti-violence policy is evident in Figure 2 in the decreased expenditure percentages recorded between 2001 and 2004 for all women's services. It was not until 2005 that the Liberals began to reinvest in anti-violence policy. Even so, the reinvestment was a modest increase in expenditure of $12.5 million, including $5.1 million for transition houses, $2 million for the existing “Stopping the Violence” and the “Children Who Witness Abuse” counselling programs and $1.6 million to expand outreach and violence prevention initiatives (MCAWS, 2005). Although the expenditures were welcomed by provincial advocates, they were not enough to restore much of what had previously been cut, including core funding for women's centres.
As with the Ontario example above, we saw evidence of significant anti-violence policy diversity in BC between different party governments. Specifically, evidence showed negative responses under the right-wing Liberals who embraced neoliberal ideology and retrenchment of programs along with a gender-neutral approach to anti-violence issues. The latter position was typified by a 2001 suggestion by Liberal Attorney General Geoff Plant to rename the Violence against Women in Relationships policy the Violence against People in Relationships policy, ostensibly in order to recognize that “women, too, can initiate violence” (O'Neill, 2002: 2). All of the significant negative policy announcements (three) were made under the Campbell Liberals, with only one modest positive announcement at the end of the study period.
In contrast, the centre-left NDP was more responsive to feminist anti-violence advocates in the 1990s and resisted neoliberal retrenchment pressure into 2000, even during a tough provincial recession. Five significant positive policy announcements were made during the NDP years in office. This was the most positive period recorded between 1985 and 2005 for anti-violence policy and was recognized as such by provincial advocates. The Vander Zalm Social Credit party, perhaps surprisingly, also appeared to be more open to anti-violence interests in the early 1990s instead of following neoliberal retrenchment trends. The Social Credit party enacted two significant positive anti-violence programs during the study period. These decisions, however, were likely reflective of a more moderate right-wing approach precipitated by a desire to appeal to women voters in the run-up to a provincial election that the party was on the verge of losing.30
I discuss the impact of electoral factors on partisan differences in my larger dissertation. See Collier (2006).
Conclusion
In the end, the policy evidence presented above for both Ontario and BC raise questions about welfare state retrenchment trends and demonstrate that while governments have appeared to embrace retrenchment at certain time points, overall they do not seem to be converging or growing more alike in their approaches to anti-violence policy. Therefore it is difficult to pronounce definitively on the overall impact of neoliberalism on anti-violence policy in Ontario and BC because of this obvious variation in the extent to which each party government adopted or rejected neoliberal practices between 1985 and 2005. Diversity was even present when we compare similar party governments at particular time points in Ontario and BC. For example, between 2001 and 2002, the right-wing Harris Conservatives in Ontario were increasing anti-violence expenditures and enacting new programs to combat violence against women. While these initiatives were prompted in part by media attention surrounding a number of high-profile murders and were not very responsive to feminist demands, they stood in stark contrast to the significant cuts to anti-violence services that were enacted by the right-wing Campbell Liberals in BC at the same point in time. Clearly, retrenchment convergence was not in evidence in this provincial comparison and neither was support for the claim that anti-violence policy was somehow immune to neoliberal program cuts.
This comparative example of the Harris Conservatives and the Campbell Liberals mentioned above suggests that party differences are in fact more nuanced than indications earlier in this article—namely that the left is always positive in its anti-violence policies and that the right is always negative in its approaches to violence against women.31
I discuss this nuanced impact of the partisan variable in the larger dissertation (Collier, 2006)
While the comparative scope of this paper is limited, it lends more support to partisan theory of public policy explanations than to convergence theories. It also suggests that the impact of neoliberalism on women's anti-violence policy is not straightforward. While evidence above supports the notion raised by Teghtsoonian that neoliberalism has the potential to mute social democratic tendencies of left-wing governments, we also see that this is not always the case. Similarly, while neoliberalism can push right-wing governments to be less responsive to anti-violence advocacy demands, evidence demonstrates that there were times when right-wing governments, notably the Vander Zalm Social Credit party in BC, were not deaf to these voices. In the end, not all of the questions raised at the beginning of this paper can be answered definitively here. However, this study raises important reservations about some of the assumptions in the comparative women and politics literature regarding the impact of welfare state retrenchment and neoliberalism. These deserve further research to help us better understand how both neoliberalism and welfare state retrenchment impact women's policy and their lives in Canada and beyond.
Acknowledgments
An earlier version of this article was presented at the 2006 meeting of the Canadian Political Science Association. My thanks to the anonymous Journal reviewers for their helpful comments.
Appendix A—List of Significant Anti-Violence Policy Responses in Ontario and BC 1985–2005
Ontario's Significant Anti-Violence Policies 1985–2005

British Columbia's Significant Anti-Violence Policies 1985–2005
