post-materialism versus the welfare state?

03 Erickson (JB/D)

21/3/02

9:20 am

Page 301

PA RT Y P O L I T I C S

V O L 8 . N o . 3 pp. 301–325

Copyright © 2002 SAGE Publications

London

Thousand Oaks

New Delhi

POST-MATERIALISM VERSUS THE WELFARE STATE? Opinion Among English Canadian Social Democrats Lynda Erickson and David Laycock ABSTRACT

Theories of contemporary social democracy suggest parties of the left may have to alter their views on the welfare state and shift their emphases in policy priorities to remain electorally competitive. In this article, we examine the structure of opinion within the New Democratic Party with regard to conventional welfare state policies, ideas on equity and post-materialist issues. We find no evidence of a muted egalitarianism or increasingly selective support for the redistributive agenda of the welfare state, and little evidence of internal party divisions rooted in a middle-class defection from the traditional programs of the welfare state. Nor is there a materialist–post-materialist division within the party. Our data show that post-materialist inclinations increase with support for a class redistributionist, materialist policy agenda. Our findings cast doubt on the ‘new politics’ thesis that post-materialist agendas will crowd out materialist policy issues and add to skepticism concerning an imminent post-materialist redefinition of social democratic strategic and policy choices.

KEY WORDS Canada political parties post-materialism social democracy welfare state

Introduction Like social democratic parties across the Western world, Canada’s federal New Democratic Party (NDP) faces a major challenge responding to a political climate and public discourse increasingly hostile to many elements of conventional social democratic ideology. The political right has successfully promoted reduced public spending and lower taxes, fueled antipathy to public ownership and government regulation and placed welfare reform high 1354-0688(200205)8:3;301–325;022817

Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 302

PA RT Y P O L I T I C S 8 ( 3 )

on the political agenda. Given the NDP’s traditional support for a strong welfare state to address social and economic inequality, and the popularity of the new right reform agenda, substantial dilemmas confront the party. Complicating its strategic options is the increased salience of ‘new politics’ issues among many contemporary left-leaning groups. ‘Struggles for recognition’ (Fraser, 1995) and issues involving more equal access to social, economic and political opportunity structures for historically disadvantaged, non-class groups, especially women and visible minorities, have become an important part of the cultural milieu of social democracy (Kitschelt, 1994). These issues are part of what is characterized as a ‘post-materialist’ agenda (Inglehart, 1990) that has gained particular support among younger generations. But by placing an emphasis on post-materialist issues, rather than on issues of economic class equality, social democratic parties may risk alienating their more traditional constituencies, especially among the working class. The party’s disastrous experiences in recent federal elections make facing programmatic issues seemingly unavoidable. In the 1993 election, it lost close to two-thirds of its 1988 popular vote, plus official party standing and any real voice in Parliament. Although the NDP increased its seat total from 9 in 1993 to 21 in 1997, its share of the vote only increased from 7 to 11 percent, as it continued to have difficulties making its appeal resonate with the public (Nevitte et al., 1999: 24–44). In the autumn 2000 federal election, the NDP dropped back to 13 seats, netting only 8.5 percent of the national vote, one seat in Ontario, and exceeding one quarter of the popular vote only in Saskatchewan (26 percent). These electoral shocks have given the question of intra-party differences on values and policies increased significance. Policy renewal and strategic planning initiatives may be compromised if leaders and activists are at odds on these matters. Opinion differences on such questions among distinct categories of party activists may reduce members’ willingness to work for the party. Such differences may seem fairly unproblematic during periods of relative electoral success. But they attain greater significance in times of electoral failure. This article examines the structure of opinion on values and policy choices among English Canadian New Democrats, using data from the 1997 Canadian Social Democracy in Transition (CSDT) survey.1 Our analysis examines opinions relevant to support for the welfare state, views about state economic strategies, and ideas on equity and post-materialist issues, and undertakes comparisons among party members and activists across several, mainly demographic, variables.

The Setting Since the 1993 election, the federal NDP has been through two timeconsuming exercises in self-evaluation and is currently mid-way through a 302 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 303

E R I C K S O N & L AY C O C K : P O S T- M AT E R I A L I S M V S . W E L F A R I S M

third. Yet, given the magnitude of their electoral rejection since 1993 and obvious incentives to reconsider seriously their programs and strategies, the most striking thing about these NDP reviews has been their status-quoreinforcing character. Delegates at the party’s 1999 policy convention did not substantially alter party policies. They endorsed a balanced budget ‘within the business cycle’ and a modest tax cut for low- and middle-income Canadians as necessary strategic concessions to the media, government and opposition consensus on such fiscal options over the past decade. However, for convention delegates and party strategists alike, these policy positions had little effect on their stance with respect to other, logically related and expensive social policy expenditure items. Convention delegates since 1993 have voted to enhance existing universal programs, supported creation of a universally accessible, state-financed child-care system and endorsed expansion of other high-cost, state-financed social programs (NDP, 1999). The party view, reflected in its 2000 federal election campaign, seemed to be that the NDP should bide its time waiting for its former voters to return to their earlier, more social democratic dispositions. Was this view shared by its membership, and, if so, by some groups within the party more than others? The answers matter because members’ community-level contributions have always been crucial to the party’s impact on the federal scene, and because a party leadership that becomes disconnected from activists and traditional supporters courts serious disaster. Was the federal leadership disconnected? Or were NDP members solidly behind the party’s non-transformation on basic policy questions?

Theoretical Bearings To help frame our analysis, we have chosen three perspectives on the ideological context of social democratic politics and party competition in contemporary European welfare states. They suggest, for different reasons, that social democratic parties may have good reasons to alter their views on the welfare state and shift the emphases in their policy priorities. They also imply some potential bases of internal party difference and divisiveness. In the mid-1980s, Gosta Esping-Andersen (1985) attempted to predict the challenges Scandinavian social democratic parties would face over the following generation. Assuming that fiscal pressures would produce increased taxation and a decline in public services, he argued that state involvement in the distribution of either opportunities or resources would lead to new cleavages and equity conflicts. When faced with the choice of supporting inter-group equity or paying reduced taxes, even many equalityoriented citizens will choose reduced taxes and a less intrusive state presence in social and economic opportunity structures. Then and since, EspingAndersen has argued that emerging social cleavages will pit poor against 303 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 304


middle class, private against public sector employees, men against women and culturally dominant but subjectively threatened white majorities against culturally distinct immigrants and visible minorities. Fiscal pressures will push social democratic parties to accept a ‘recommodification’ of aspects of welfare-state-mediated social life that social democratic governments had partially ‘decommodified’ (Esping-Andersen, 1996). Social programs like unemployment insurance and pensions will become more income dependent and less generous. The prospect of social democratic party policy shifts is also addressed in Herbert Kitschelt’s analysis of the transformation of European social democracy. Kitschelt argues that the axis of political competition has moved to the right in advanced capitalist welfare states in response to the demands of contemporary market capitalism and global competition. Electorally competitive social democratic parties are limited, at best, to a ‘moderate defense of the accomplishments of the welfare state’ (1994: 21). Given the issue space Kitschelt sketches for contemporary politics in Western industrialized societies, social democratic party competitiveness may require a post-materialist transformation of social democratic appeals. It would incorporate ‘libertarian’ and ‘communitarian’ orientations. He portrays the libertarian tendency as essentially anti-authoritarian and autonomy-seeking, opposed to unaccountable state bureaucracies and corporate power. The communitarian dimension focuses on participation and citizen’s self-governance, and complements left-libertarian political orientations. Both of these orientations, he says, fit in with what Inglehart and his collaborators (Abramson and Inglehart, 1995; Inglehart, 1990; Inglehart et al., 1996) argue is the post-1970s tendency for well-educated middle-class people to be attracted more to ‘self-realization’ than material consumption goals in new political agendas. Kitschelt acknowledges a substantial component of the traditional social democratic constituency remains conventionally ‘materialist’ in orientation. In the short term a social democratic party risks alienating industrial working-class supporters by emphasizing post-materialist concerns. But as the industrially employed proportion of the working class declines, it becomes an increasingly weak core vote for parties of the left. Consequently, most Western social democratic parties are best advised to advance a postmaterialist agenda while selectively promoting conventional materialist and redistributive objectives, such as labor market protections and minimum income programs. The ‘new politics theory’ posits an increasing willingness in the growing ranks of post-materialists to see a reduced redistributive role for the state. Among social democrats, the question is, will libertarian and communitarian attitudes lead post-materialists to support a post-materialist agenda on top of, or only at the expense of, some ‘materialist’ redistributive programs? Kitschelt’s perspective suggests that the optimal scenario for NDP electoral 304 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 305


viability would involve the party’s gradual detachment from the full Keynesian agenda of the industrial working class with increasing commitment to decentralized ‘left-libertarian’ and ‘communitarian’ approaches to politics. This could include prominent support for participation by groups that have been historically marginalized in public life, attention to selfgoverning practices within workplaces and other social sites of power, and opposition to the forces and policy choices shaped by ‘globalization’. Taking this approach might hasten the party’s encounter with a strategic dilemma. In the short run, this approach could easily lead to loss of working-class votes to right-wing populist parties, along with an ‘identity crisis’ and bitter in-fighting when the party can least afford it. On the other hand, consider Kitschelt’s argument about the reconstitution of issue space in Western polities (1994) applied to Canada. The NDP’s failure to trade off some materialist priorities for a new agenda could alienate major sources of potential support, and precipitate a further major vote loss in provinces where it has been strong in the past. To make matters worse, the party could be largely abandoned by its trade union core, worried about wasting votes that could otherwise keep a new right-wing force (the former Reform Party, now called the Canadian Alliance) from gaining more welfare-state-threatening leverage in Canadian legislatures (Gidengil et al., 1999). Claus Offe provides a third perspective from which to approach our data. In 1987, Offe argued that as bonds of social solidarity once furnished by working life and smaller-scale community life rapidly decline, working-class and other middle-income citizens will be more inclined to ‘defect’ from support for an increasing number of welfare state programs. Released from the moral constraints of social solidarity, and increasingly insecure about their own prospects, these citizens will be more attracted to parties promising reduced taxes and decreased social service regimes (Offe, 1987). Offe may be mistaken regarding a mass transformation of previously welfare-state-supportive citizens. Still, he appears to be right about middleclass willingness to withdraw support from some social goods of the welfare state. So it makes sense to ask whether the major programmatic components of Canada’s social welfare policy regime are being reassessed among NDP members and activists. Are these NDPers contemplating a replacement of their traditional ‘materialist’ agenda by a ‘post-materialist’ agenda? Or do they see these components as a unified and inviolable package that defines Canadian citizenship, despite recent changes to Canada’s ‘citizenship regimes’ (Jenson, 1997)? Are the major programmatic components of Canada’s social welfare policy regime being reassessed among NDP members and activists generally, or among particular groups within the party? The recent experience of the Labour Party in Britain also helps to put the NDP case into profile. Under Tony Blair’s leadership, ‘New Labour’ regained public office with a substantially reformed party and programme. Having dramatically curtailed the power of party activists and organized 305 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 306


labor within the Labour Party, Blair’s centre–left circle fashioned a lower tax, anti-re-nationalization, middle-class appeal that is strong on educational reform, neighborhood security and revitalization of British industry through incentives to private-sector players (Seyd, 1999). While the Saskatchewan NDP government openly endorsed a ‘Third Way’ similar to the British model, the federal party has offered only cautious support for some elements of Blair’s policy package. What lies behind the federal NDP’s reluctance to adopt a strategy so electorally attractive in Britain and arguably responsible for retention of government power in Saskatchewan? Do party members and activists support the federal party leadership’s promotion of a combined materialist and post-materialist agenda? Or is there evidence of internal cleavages within the party that reflect the patterns of social conflict depicted by Offe or Esping-Anderson, or the materialist/post-materialist conflict posited by Kitschelt?

The Data Our analysis is based on data from the 1997 CSDT study of NDP member/activists from British Columbia, Saskatchewan, Manitoba and Ontario. These provinces are areas of historically concentrated federal NDP strength, and have elected at least one NDP provincial government since 1968. In early 1997, 3879 surveys were mailed to: (i) 250 randomly selected party members from each of the four provinces; (ii) all provincial constituency association executive members and provincial council members in the four provinces; and (iii) all incumbent NDP MLAs and MPs from the four provinces as of January 1997, as well as all surviving NDP MPs and MLAs from 1968 to 1997. 1440 mail questionnaires were returned, yielding a response rate of 37 percent. In regional terms, the sample distribution is a reasonable reflection of the party size in each province: 35 percent are from British Columbia, 15 percent from Saskatchewan, 15 percent from Manitoba and 36 percent from Ontario.2

Values and Attitudes in the NDP Defense of the Welfare State A good place to begin our consideration of NDP activists’ attitudes to the welfare state is to assess NDP members’ commitment to equality. Respondents were asked to choose between equality, defined as ‘nobody is underprivileged and social class differences are not so strong’, and freedom, defined as ‘everyone can live in freedom and develop without hindrance’. By this measure, the NDP is unarguably a party of the egalitarian left: just 15 percent preferred the freedom option, whereas 80 percent preferred equality. 306 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 307


This suggests that a move away from emphasis on the party’s class equality agenda would elicit activists’ spirited opposition. To place our respondents’ preferences in context, we consider data on the same question from the Canadian sub-sample of the 1990 World Values Survey (WVS).3 While aware of the need for caution in comparing responses to surveys conducted in different temporal contexts and economic circumstances, we believe the extent of difference between the CSDT and the WVS samples is instructive.4 Among the WVS respondents, freedom is the clear choice, at 58 percent. This proportion preferring freedom over equality was almost four times the comparable number for the CSDT sample.5 Moreover, just 32 percent in the WVS chose the equality option. Even among NDP supporters in the WVS, 48 percent said they preferred freedom, while just 42 percent preferred equality. These figures support the hypothesis that party activists tend to take more extreme positions than party voters on political issues (May, 1973). Indeed, the gap on this issue between our activists/members and NDP supporters is even greater than the distance between NDP voters, on the one hand, and Liberal, Conservative and even Reform voters, on the other. In that same WVS survey, 62 percent of Liberal supporters, 61 percent of Conservatives and 67 percent of Reform Party supporters preferred freedom to equality. Turning to the degree to which our respondents’ were opposed to state retrenchment in social welfare provision we find that far from following Kitschelt’s advice, our NDP activists prefer increased state support for citizens (see panel 1, Table 1). They were asked to rank their views on a scale with ‘1’ meaning they agreed that individuals should take more responsibility for providing for themselves, and ‘7’ meaning the state should take more responsibility to ensure that everyone is provided for. Just 20 percent placed themselves on the individual responsibility side (i.e. scores of 1, 2 or 3), while almost three times that number favored the state taking more responsibility (scores of 5, 6 or 7). Consensus in support of universal public provision of health care and pension programs was overwhelming. Most respondents rejected the removal of any services from Medicare and opposed health care user fees, while as many as 90 percent disapproved of making the Canada Pension Plan a voluntary one (panel 2). On each of these questions half or more of the respondents indicated they strongly disagreed with removing universal state provision. Respondents also broadly endorsed more public spending on higher education and rejected increased university fees (panel 3). It may not be surprising that our respondents strongly support programs, such as those above, which benefit the middle class as well as the less well off. However, contrary to Offe’s expectation of ‘defection’ on public goods that less clearly advantage the middle class, they also endorsed relatively generous social provisions that tend to favor the working class or the poor. When questioned about employment insurance policies, few respondents supported further cutbacks in eligibility. And when asked whether the 307 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 308


Table 1. Support for welfare state issues General Individual versus state responsibility Mean score: 4.73

% on individual responsibility side

% on state responsibility side

(N = 1390)

20

57

Universalisma

Agree or strongly agree

Disagree or strongly disagree

Remove some services from Medicare for those who can afford (N = 1412)

14

80

Health care user fees (N = 1412)

9

85

Canada Pension Plan should be voluntary (N = 1420)

7

91

Educationa More public money on higher education (N = 1413) Universities should raise tuition fees (N = 1407) Employment insurancea Unemployment insurance should be harder to collect (N = 1411)



77 9

10 78



8

80

Unemployed take any job versus unemployed right to refuse a job

% on take any job side

% on right to refuse job side

Mean score: 4.88

19

62



Much talk of welfare abuse exaggerated (N = 1398)

79

15

People on social assistance become dependent versus work hard to get off it Mean score: 4.26

% on dependency side 31

% on work hard side 46

(N = 1420) Welfare issuesa

(N = 1393) Source: 1997 CSDT survey. a Neutral responses included in percentage calculations.

unemployed should have to take any job in order to keep their benefits (scored 1), or they should have the right to refuse a job (scored 7), more than 60 percent scored in the latter direction (panel 4). Similarly, a strong majority of the respondents said talk of welfare abuse was exaggerated. The only question eliciting any ambivalence about programs for the disadvantaged addressed the effects of social assistance. Respondents were asked whether ‘people receiving social assistance tend to become dependent on it’ (scored 1) or whether those ‘people on social assistance tend to work 308 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 309


Table 2. Social policy priorities Item

% ranking item as most important Mean score

Reducing poverty (N = 1113)

64

1.82

Increasing public support for health care (N = 1084)

15

2.85

Increasing resources for retraining unemployed people (N = 1111)

10

3.26

Spending more on education (N = 1117)

10

3.28

Increasing public support for child care (N = 1117)

3

3.66

Source: 1997 CSDT survey. Scoring: 1 = highest; 5 = lowest.

hard to get off of it’ (scored 7). Here, although fewer than one-third agreed that people become dependent on assistance, still less than half said that people on social assistance work hard to get off it (panel 5). In ranking different social programs, the respondents’ redistributive concerns clearly prevailed. Asked to order the importance of five different social policy goals, a class redistributive agenda (reducing poverty) clearly took first place (see Table 2). The mean ranking of the second ranked goal, increasing support for health care, also shows distinctly the greater priority respondents give health care compared to more productivity-oriented programs of retraining and education. Overall, the social democrats in our sample are strongly supportive of the various material provisions of the welfare state. There is little evidence that they are willing to support the ‘recommodification’ of universal social programs that Esping-Andersen and Kitschelt suggest may be a condition of renewed social democratic party success. Our data do not allow us to say much about why this is the case, beyond suggesting that our respondents’ strong commitment to equality prevents them from considering this as a viable option for a party whose raison d’être has always been reduction of class inequality. Economic and Fiscal Issues Redistributive preferences also prevailed in answers to questions about economic and fiscal issues. In assessing government options for responding to difficult fiscal situations, our respondents were almost unanimous in their support for raising taxes for wealthy Canadians and corporations (see Table 3). Similarly, they strongly opposed across-the-board spending and service cuts by governments. On the other hand, almost half of the sample accepted cuts to the size and cost of government when governments face difficult fiscal situations. 309 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 310


Table 3. Opinions on economic and fiscal policy items When faced with a difficult fiscal situation governments should:a



Increase revenues by raising the income taxes of wealthy Canadians (N = 1426)

91

4

Generate more revenue by raising corporate taxes (N = 1424)

91

4

Cut spending and reduce services across the board so that everybody pays equally (N = 1411)

16

76

Cut its own expenditures by reducing the size and cost of government (N = 1424)

49

31

Increase private ownership of business and industry

% on private ownership side

% on government ownership side

24

42

% on more equal incomes side

% on individual effort side

68

18

versus Increase government ownership of business and industry Mean score: 4.75 (N = 1402) Incomes should be more equal versus Greater incentives for individual effort Mean score: 2.91 (N = 1382) Source: 1997 CSDT survey. a Neutral responses are not included in the table but have been included in the percentage calculations.

While an important subset of New Democrats might support some shrinking of the state in times of fiscal pressure, this did not translate into support for more private ownership. Using a 7-point scale again, where ‘private ownership of business and industry should be increased’ was scored ‘1’ and ‘government ownership of business and industry should be increased’ was scored ‘7’, fewer than a quarter scored in the private ownership direction. Still, the proportion who scored in the direction of greater public ownership (42%) represents a substantial decrease in support for government ownership over the past decade. Seventy-nine percent of a 1989 sample of NDP federal convention delegates preferred more public ownership (Archer and Whitehorn, 1997: 134).6 This decrease notwithstanding, clearly, many NDP activists and members continue to prefer state enterprise to private entrepreneurship, even if they may like to see the former undertaken by fewer government employees. In their commitment to an economically activist state, New Democrats are certainly closer to the historical norm for social democrats than most of their 310 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 311


European counterparts, and certainly closer to this norm than most European social democratic governments. Finally, when faced with a trade-off between income equality and individual incentives, our sample decisively chose greater income equality. Our 7-point scale placed ‘incomes should be made more equal’ at ‘1’ and ‘there should be greater incentives for individual effort’ at ‘7’. More than twothirds of the respondents supported more income equality, and fewer than one-fifth opted for greater incentives for individual effort. The Equality Rights Agenda To both legislators and ordinary citizens, one of the most sensitive extensions of the ‘new politics’ agenda has occurred in the labor market, where programs to expand opportunities for historically disadvantaged groups come up against notions that non-designated groups face an unfair playing field. Although the equality-seeking dimensions of such programs may have a special ideological appeal for social democratic parties, such programs may be internally divisive for them in the context of a shrinking workingclass labor market. We begin examining such equity issues by looking at respondents’ general views of an equality-seeking agenda. As Table 4 demonstrates, there is little support for the idea that ‘we worry about equality too much in Canada already’. Fully four-fifths disagreed, and almost half of these disagreed strongly. By comparison, a similar question asked in the 1997 Canadian Election Study (CES)7 elicited agreement from 48 percent of the national Table 4. Opinion on equity issues

Item


We worry about equality too much in Canada already (N = 1413)

11

8

82

The government should extend its employment equity policies beyond the public sector to enterprises which do business with the public sector (N = 1373)

73

14

14

We should have laws that make all employers hire a certain proportion of women and minorities (N = 1391)

42

19

38

Hiring people because they are women or members of a visible minority undermines the principle of merit (N = 1396)

31

13

55

Employment equity programs create reverse discrimination (N = 1351)

27

15

58

Neutral

Source: 1997 CSDT survey.

311 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015


03 Erickson (JB/D)

21/3/02

9:20 am

Page 312


sample.8 On this question, NDP voters in the CES sample were closer to the NDP members in our sample than they were to the national sample or the other parties. 79 percent of the NDP voters disagreed with this claim, compared with 55 percent of Liberal voters, 52 percent of Conservative voters and just 30 percent of Reform voters.9 Our other equity questions also elicited a majority or plurality of respondents in support of the ‘equity options’ (see Table 4). A substantial minority even supported requiring all employers to hire a certain proportion of women and visible minorities. Such a quota policy would implement employment equity more thoroughly than is done in any Western democracy. Still, there was some ambivalence about equity programs. When faced with the language of liberal opposition to affirmative action in the labor market, the majorities rejecting these claims were not large: just 55 and 58 percent, respectively, disagreed with the claims that equity programs ‘undermine the principle of merit’, and that they create ‘reverse discrimination’. This is not just a semantic quibble. When a significant proportion of social democratic party activists fail to reject the liberal discourse and accompanying attitude with respect to non-class group inequalities in the labor market, we must acknowledge two things. First, the liberal discourse is powerful indeed, since it throws them off their otherwise consistently redistributionist egalitarian track. Second, the distributionist track they appear to be most easily thrown off involves the equality agenda of ‘new social citizenship’, in which not class, but other group bases of inequality, are the main cultural and policy targets. Post-materialism What about other post-materialist dispositions beyond equity issues? Do New Democrats support elements of a post-materialist agenda relevant to quality of life, participation and self-government, and self-realization? Our respondents were asked questions based upon two post-materialism scales, one partially modeled on Inglehart’s original measure (1971), and a second one taken directly from the World Values Survey. Each scale asked respondents to rank ‘in order of importance to them personally’ four different goals. The goals in the first, modified, scale were: (i) maintaining a high rate of economic growth; (ii) fighting rising prices; (iii) seeing that people have more say in how things get done at work and in their communities, and (iv) giving people more say in government decisions.10 The goals in the second scale included: (i) maintaining a stable economy; (ii) the fight against crime; (iii) progress to a less impersonal, more humane society; (iv) progress toward a society where ideas count for more than money. For each scale respondents were classified as materialists if they chose only items (i) and (ii) (in either order) for their first two preferences. Respondents who chose only items (iii) and (iv) (in either order) for their first two preferences were classified as post-materialists. The rest were deemed to have mixed value priorities. 312 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 313


On these scales, our New Democrats clearly tilt toward the post-materialist value dimension. 44 and 48 percent of our respondents ranked as postmaterialists on the first and second scales respectively, and just 10 and 9 percent ranked as materialists. The comparison with the Canadian sample from the WVS is again instructive. There, just 26 percent scored as postmaterialists using Inglehart’s original measure. On the measure that was directly comparable to our second scale, only 12 percent of the national sample scored as post-materialists, whereas 36 percent scored as materialists. The comparatively high level of post-materialists among respondents so strongly committed to redistributive policies and a program of state intervention leads us to wonder whether a communitarian agenda is seen, at least within the party, to be at odds with traditional social democracy. It may even be that communitarian and class egalitarian goals are mutually reinforcing values within parties of the left. This is not ruled out by Kitschelt. But we should recall that he expects that a left party’s strong communitarian traditions will increase the attractiveness of the left libertarian agenda over the class equality agenda for its leaders and activists.

Intra-Party Divisions? Notwithstanding the relatively high level of agreement among our respondents on many of the issues described above, some respondents are comparatively receptive to a rightward swing for the party. In this section, we address two questions about internal party differences. First, do the patterns of social conflict Esping-Andersen and Offe envisioned as undermining traditional welfare state programs underpin internal party cleavage in the NDP? Second, is post-materialism, which Kitschelt sees as a party-divisive phenomenon, a central axis of intra-party division that undermines the classic state interventionist and class equality agendas of the party? The Group Basis of Cleavage To examine the social bases of internal party divisions we used factor analysis to construct scales that test the impact of our independent variables along different ideological dimensions. These scales tap five dimensions: two measure attitudes relevant to defense of the welfare state and its programs, two measure attitudes to the size of government and other economic/fiscal issues and a final scale measures attitudes to gender and minority equity policies.11 On questions relevant to defense of the welfare state, our factor analysis suggested an underlying structure that differentiates attitudes towards what might be described as more middle-class programs – Medicare, Canadian Pension Plan and university education – from other issues and programs, especially those assisting the unemployed and recipients of social 313 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 314


assistance.12 Scales based on the first factor we call ‘social programs’ and, following Esping-Andersen, scales based on the second factor we call ‘decommodification’. For the fiscal/economic issues, respondents’ views concerning three solutions to governments’ difficult fiscal situations – raising corporate taxes, raising income taxes of the wealthy and putting more Canadians to work – formed one scale (fiscal progressivism). The remainder of the fiscal/economic items in Table 3 formed a socialism scale. The equity scale includes all five equity items in Table 4.13 The scales range from 0 to 1. Given the high levels of support for most of the classically left-wing alternatives in each scale, mean scores for all five scales were well above the mid-point. They ranged from 0.61 for the socialism scale to 0.87 for the social programs scale. In order to test for the relevance of more traditional social divisions, for our independent variables we introduced two class-related measures: income and education, and added two other economic variables: union membership and private sector employment.14 In addition, we included an age variable – over 55 – to measure the potential traditionalism of older members. For ‘new politics’ variables we incorporated youth, gender and visible minority measures. Finally, given the salience of regionalism in Canadian political life, and the history of clear regional conflict within the national NDP, we included regional variables representing the four provinces the CSDT study surveyed. Table 5 presents a series of correlations between the five scales and these various characteristics. Two general features warrant comment. First, Table 5. Patterns of internal difference – bivariate correlationsa Social programs Union Private sector

0.068* 0.007

Fiscal Decommodification progressivism Socialism 0.076*

-0.063*

0.075**

0.082**

0.001

0.172***

0.174***

0.062*

0.048

0.127***

0.134***

Female

0.137*** 0.170** 0.016

0.092***

0.076**

0.102***

0.010 0.003 0.064*

0.074**

0.064*

0.064*

University degree Youth

0.125***

0.031

High income Over 55

0.091**

Equity

0.033

0.142*** 0.165*** 0.084** 0.001

0.048 0.183***

Visible minority

0.142*** 0.070*

0.008

0.114***

Saskatchewan

0.076**

0.214**

0.116***

0.169*** 0.228***

B.C.

0.062*

0.117**

0.094***

0.035

Ontario

0.108***

0.252***

0.104***

0.102***

Manitoba

0.018

0.031

0.098***

0.087**

0.037 0.092*** 0.273*** 0.017

Source: 1997 CSDT survey. a sTo simplify the presentation, the number of respondents for each correlation has been omitted. The numbers range from 1040 to 1415. *sig. ≤ 0.05; **sig. ≤ 0.01; ***sig. ≤ 0.001.


03 Erickson (JB/D)

21/3/02

9:20 am

Page 315


although many of the correlations are statistically significant, few are substantial. Group-based differences within the party are not very sizeable. Second, while there are a number of significant correlations, few are as large and none as consistent as those found for regional groups, in particular Saskatchewan and Ontario. We return to the regional findings later. With respect to our class-related/economic variables, the findings do not suggest that internal party divisions accord very closely to the patterns of social conflict our theoretical accounts depicted. Thus, while union members tended to show greater support for traditional social democratic policies (columns 1 to 4), so too did those with university degrees. Even high-income earners had higher scores on the social programs scale than did those with lower incomes. Only respondents employed in the private sector showed any particular tendency to more centrist politics, and even here the correlations are small and only two are statistically significant. In addition to these patterns on more traditional policy issues, there is little evidence that support for equity issues, a new politics issue, is any weaker among unionized party members. Indeed, we found a small but statistically significant positive correlation between equity scores and union membership. Nor is the difference in equity scale scores between public sector and private sector unionists either substantial or statistically significant (data not shown). At least among party members and activists, equity conflicts are unlikely to pit unionists against non-unionists or private against public sector unionists. As might be expected, members over 55 were more distinctly conservative in their attitudes to the new politics of equity. But they were also consistently and distinctively more conservative in their attitudes to three of the other four issue areas, including traditional social programs, amongst which are pensions, where we might have expected them to be the most resistant to state cutbacks. Like the traditional group factors, the new politics variables do not demonstrate a consistent pattern. Younger respondents (under 35) appear to be somewhat more left wing, but they are not very distinctive, as the positive correlations are weak. Moreover, on the equity variable, where we might expect an expression of generational difference to be most evident, the correlation is not significant. For the visible minorities in our sample, their more conservative tendencies with respect to most of the traditional issues of social democracy are suggested in the pattern of mainly significant negative correlations. To be consistent with a notion of conflict between new and old politics, we would have expected to find a fairly strong positive correlation for visible minorities on the equity issue. This was not evident. On the other hand, for the gender variable, the equity correlation is both significant and comparatively large. Indeed, in a pattern consistent with some findings with respect to gender differences in the population (Everitt, 1998) female respondents are generally more left wing on issues than males. Two features stand out with respect to region. First, the kinds of policy positions that differentiate the Saskatchewan NDP government from the 315 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 316


federal party are displayed in our sample as well. The Saskatchewan party’s more conservative stance is echoed in the pattern of correlations in Table 5, where on all dimensions, except attitudes to traditional social programs, Saskatchewan respondents recorded higher negative correlations than any other category. Second, in a virtual mirror image of the Saskatchewan variable, the Ontario group similarly shows consistent and comparatively high positive correlations on all five dimensions. British Columbia respondents are, on the other hand, positioned more closely to their Saskatchewan counterparts than to Ontario members/activists. While these correlations imply that there are social factions in the party where moves closer to the center of the partisan spectrum would be most strongly resisted, there are overlaps and possibly cross-cutting cleavages among these groups. To isolate more clearly the primary sources of ideological difference among our respondents, we subjected our data to a series of multiple regression analyses. The independent variables, including region, were a series of dummy variables.15 The results are set out in Table 6, showing only the statistically significant coefficients. For all scales, the r squares are relatively low, indicating that group-related differences account for a small proportion of the variance on any of the scales. We reiterate, then, that group differences, while measurable, are not substantial. Moreover, the mixed pattern with respect to old and new politics persists. Now, however, university education is the most consistent predictor variable positively related to both traditional welfare state and new politics agendas. Youth disappears as a predictor variable but gender and now visible minority designation are both positively related to the equity scale. But visible minority remains as a negative predictor on support for the more traditional social democratic agenda. This is the only evidence that equity issues may alter the usual dynamics of intra-party relationships. More generally, ideological groupings, as limited as they are, remain similar whether the issue is one of non-class equity or class equality. The robust character of the regional variables suggests that regionalism is not a cover for differences based on other socio-demographic characteristics, but may reflect long-standing regional variation in intra-party culture reinforced by the differential success of the provincial wings of the party.16 In particular, Saskatchewan and British Columbian members appear to be the most consistent anchors for what may count as the political right in the party on all but equity issues, where our British Columbia variable is no longer significant. On the other hand, Ontario is only distinctive as a source of left-wing opinion on equity issues and those related to decommodification concerns. On the whole, we find little evidence that internal party divisions are based on a kind of middle-class defection from the traditional programs of the welfare state. Nor do we see the sort of ‘invidious cleavages’ that EspingAndersen predicted among traditional social democratic party supporters over equity issues. There are some group differences on equity issues but 316 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 317


Table 6. Multiple regression analysis of scales by respondent characteristicsa Social programs

Fiscal Decommodification progressivism Socialism

Union

0.025** (0.009) 0.023 (0.012)

Private sector

0.038** (0.012) 0.037** (0.013)

0.018* (0.009

High income University degree

Equity

0.035** (0.011)

0.048*** (0.012)

Over 55

0.034** (0.013)

0.046*** (0.013)

Female

0.025* (0.011)

0.035* (0.011)

0.052*** (0.012)

0.073*** (0.014)

0.044*** 0.047*** (0.014) (0.015) 0.018* (0.009)

0.083*** (0.013)

Visible minority

0.104*** (0.022)

0.045* (0.022)

Saskatchewan

0.047* (0.020)

0.102*** (0.020)

0.055*** (0.016)

0.120*** 0.087*** (0.021) (0.024)

B.C.

0.044** (0.017)

0.055*** (0.017)

0.059*** (0.014)

0.069*** (0.018))

Ontario Constant Adjusted r2 N

0.110*** (0.023)

0.039* (0.016) 0.824*** (0.020) 0.06 1017

0.069** (0.026)

0.079* (0.020)

0.670*** (0.020) 0.13

0.884*** (0.016) 0.04

1003

1053

0.665*** (0.021) 0.11 1012

0.570*** (0.024) 0.16 963

Source: 1997 CSDT survey. a Entries are unstandardized regression coefficients, figures in parentheses are standard errors. *sig. ≤ 0.05; **sig. ≤ 0.01; ***sig. ≤ 0.001.

they are modest and there is no pattern that suggests they will combine with a party-debilitating debate over reduced support for universalistic programs and public goods, as Kitschelt’s and Offe’s analyses each might lead us to expect. The Post-materialist Dimension Earlier, in looking at the level of support for post-materialist values among our New Democrats, we wondered whether communitarian and class egalitarian goals, rather than being at odds with each other, might indeed be mutually reinforcing. This would seem to challenge the view articulated by Kitschelt and others (Inglehart and Abramson, 1999), that post-materialist goals are not related to redistributive priorities. Their assumption seems to be that, among post-materialists, classic left–right issues are eclipsed by concerns for ‘new’, quality-of-life issues. To quote Inglehart and Abramson, 317 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 318


Table 7. Relationships between post-materialism and other dimensions of socialist opiniona Post-materialism Social programs (N = 1067) Decommodification (N = 1323) Fiscal progressivism (N = 1215) Socialism (N = 1346) Equity (N = 1259)

0.23* 0.36* 0.20* 0.40* 0.30*

Source: 1997 CSDT survey. a Entries are Pearson correlations between scale scores. *sig. ≤ 0.001.

‘such classic Left–Right issues as state ownership of business and industry and income redistribution . . . the theory explicitly identifies as issues not salient to postmaterialists’ (ibid.: p. 669). An analysis and comparison of the relationships between post-materialism and the dimensions represented in our other scales allow us to examine whether this might be contradicted within the NDP. Given the tendency for greater predictability with a larger scale our two measures were combined to create an index that ranged in values from 0 to 4. As Table 7 demonstrates, using this index, the other five scales are all positively related to post-materialism. Moreover, although the equity scale, concerned with issues connected to the ‘new politics’ agenda, is arguably the most post-materialist of the five scales, the socialism and decommodification scales showed the strongest correlations. It would therefore seem that, among our New Democrats, post-materialist values do not eclipse classic left–right concerns. Further, multiple regression analysis indicated that differences on post-materialism are only weakly related to social groupings within the party. There were comparatively few significant coefficients and the r square for the equation was a mere 0.05.17 Overall, it appears that English Canadian social democrats support a post-materialist agenda on top of their class equality goals, not at their expense. This finding adds to previous doubts concerning the broader thesis of mutual exclusion between ‘old class equality’ commitments and ‘new post-materialist values’ (Seyd and Whitely, 1992). That this mutual exclusion is so strongly contradicted among NDP party activists suggests that more research is needed to determine whether post-materialist/libertarian/communitarian values are at odds with ‘materialist’ and class equality commitments among potential social democratic voters.

Conclusions Social democratic parties have been under pressure to shift their ideological ground in response to the changing issue space of contemporary politics. 318 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 319


With this in mind, we went looking for evidence of a muted egalitarianism and increasingly selective support for the redistributive agenda of the welfare state among members and activists in Canada’s New Democratic Party. However, our New Democrats resolutely supported egalitarian values and redistributive policies, and indicated little interest in reducing the state’s role in social change. NDP members/activists have not selectively reassessed the major components of the social welfare policy regime, which they see as a largely inviolable package. And support for the ‘new politics’ agendas involving equality rights for historically disadvantaged groups is, by our measures, also robust. In looking for sources of opposition to a generous welfare state or to equity policies, we found few pockets of more conservative opinion. The most consistent predictors of such sentiment in the party were region, age and, with the exception of equity issues, visible minorities. But even the conservatism of these groups was not far removed from the party norm. Nor did we find evidence that these social democrats feel compelled to choose between (redistributive) materialist and post-materialist values, as Kitschelt and Esping-Andersen predicted of besieged social democratic parties. Among our respondents, post-materialist values were positively related to support for all elements of traditional social democratic and newer, nonclass equity policies. These NDP members agree with the thrust of their party leadership’s refusal to follow many other Western social democratic parties along the road prescribed by Kitschelt and expected by Offe and Esping-Andersen. NDP member/activists’ enthusiasm for social equality via an extensive welfare state has not been dampened by the electoral successes of parties of the right, by national governments’ fiscal difficulties, or by their own support for a range of ‘new social rights’. Despite major policy shifts by some of their provincial counterparts, the federal party is not for turning, and appears to be under no significant pressure from its members/activists to do so. Unlike most social democratic parties in Western democracies, the federal NDP has never been close to governing. But from the 1960s through the 1980s it was a serious player in national policy debates, and succeeded in pushing federal governments to expand the Canadian welfare state well beyond the US model. To make contributions to national debates on welfare state restructuring, equity or other policies, the NDP needs to recover at least its pre-1993 level of national popular vote and parliamentary strength. After three disappointing elections, this is not an imminent prospect. This is so, even though Canadians have been through more than a decade of state retrenchment, social program cuts, the challenges associated with governments attempting to do more with less, assorted side effects of ‘globalization’, and a slide in the direction of more inequalities. What might turn the NDP’s electoral fortunes around? One possibility is the decline of the 319 Downloaded from ppq.sagepub.com at SIMON FRASER LIBRARY on January 2, 2015

03 Erickson (JB/D)

21/3/02

9:20 am

Page 320


Canadian Alliance, a right-populist party that has had substantial internal conflict, including battles between social conservatives and those whose conservatism is primarily fiscal. Another possibility is that social democratic governments abroad may demonstrate that they can adapt to globalization without the combination of tax and program cuts prescribed by their political opposition (Garrett, 1998; Pierson, 2001). Over the next decade, such a demonstration could lend more legitimacy to social democratic policy options in Canada than they have recently enjoyed. Finally, the NDP might become a major player in the federal Canadian scene again by significantly altering its fiscal and social policy commitments. But our analysis suggests that the federal party is not prepared to undertake that sort of policy overhaul. New Democrats appear inclined to wait for the party’s former supporters to recover their social democratic sensibilities, or perhaps wait for new voters to discover a strong antipathy to elites’ commitments to globalizing free trade, rather than sacrificing either its materialist or its post-materialist commitments.

APPENDIX Question Wording of Scale Items Variables recoded to values from 0 to 1 then averaged.

Social Programs Scale Please indicate how much you agree or disagree with the following statements. Strongly Agree (1); Agree (2); Neutral (3); Disagree (4); Strongly Disagree (5). 1. To control costs, some services should be removed from Medicare for those who can afford to pay for such services themselves. 2. Participation in the Canada Pension Plan should be voluntary. 3. Health-care user fees should be instituted as a cost-control measure. 4. Universities should make up revenue short-falls by raising tuition fees.

Decommodification 1. Individuals should take more responsibility for providing for themselves 1 2 3 4 5 6 7 The state should take more responsibility to ensure that everyone is provided for. 2. People who are unemployed should have to take any job available or lose their benefits 1 2 3 4 5 6 7 People who are unemployed should have the right to refuse a job they do not want. 3. People receiving social assistance tend to become dependent on it 1 2 3 4 5 6 7 People on social assistance tend to work hard to get off it.


03 Erickson (JB/D)

21/3/02

9:20 am

Page 321


Please indicate how much you agree or disagree with the following statements. Strongly Agree (1); Agree (2); Neutral (3); Disagree (4); Strongly Disagree (5). 4. Unemployment insurance should be harder to collect than it is now. 5. Much of the talk about ‘welfare abuse’ is exaggerated.

Equity Please indicate how much you agree or disagree with the following statements. Strongly Agree (1); Agree (2); Neutral (3); Disagree (4); Strongly Disagree (5). 1. We worry about equality too much in Canada already. 2. We should have laws that make all employers hire a certain proportion of women and minorities. 3. Hiring people because they are women or members of a visible minority undermines the principle of merit. 4. Employment equity programs create reverse discrimination. 5. The government should extend its employment equity policies beyond the public sector to enterprises that do business with the public sector.

Socialism Please indicate how much you agree or disagree with each option. Strongly Agree (1); Agree (2); Neutral (3); Disagree (4); Strongly Disagree (5). 1. When faced with a difficult fiscal situation, governments should cut spending and reduce services across the board so that everybody pays equally. 2. When faced with a difficult fiscal situation, governments should cut its own expenditure by reducing the size and cost of government. 3. Incomes should be made more equal 1 2 3 4 5 6 7 There should be greater incentives for individual effort. 4. Private ownership of business and industry should be increased 1 2 3 4 5 6 7 Government ownership of business and industry should be increased.

Fiscal Progressivism Please indicate how much you agree or disagree with the following statements. Strongly Agree (1); Agree (2); Neutral (3); Disagree (4); Strongly Disagree (5). 1. When faced with a difficult fiscal situation, governments should generate more revenue by raising corporate taxes. 2. When faced with a difficult fiscal situation, governments should increase revenues by raising the income taxes of wealthy Canadians. 3. When faced with a difficult fiscal situation, governments should strengthen the economy by putting more Canadians to work.


03 Erickson (JB/D)

21/3/02

9:20 am

Page 322


Notes We thank Jim Bruton, who co-authored, with David Laycock, an earlier article on this topic, and Donald Blake, Ed Broadbent, Michael Howlett, David Farrell and two anonymous reviewers for their helpful comments. 1 The survey was conducted by David Laycock. Christopher Kam provided invaluable assistance in the survey design and Farrokh Heidari coded the responses. Financial support was provided by the Social Sciences and Humanities Research Council, the Boag Foundation, and a President’s Research Grant at Simon Fraser University. 2 The sample reflects the socio-demographic characteristics of people we would expect to be members and activists within a party. Respondents were more educated and older than the electorate. More than half the sample said they had a university degree, the modal age group of the sample was 45 to 55 years, with a further 39 percent over 55, and 59 percent of them were male. Also, consistent with what we might expect in a sample of New Democrats, almost half of those currently employed worked in the public or not for profit sector. 3 The WVS data were collected by the World Values Study Group (1994) and distributed by the ICPSR. Neither bear responsibility for our uses or interpretation of these data. 4 In 1990, when the WVS was administered in Canada, the recession of the 1990s had not yet hit, whereas in 1997, when the CSDT survey was conducted, the country was just beginning to pull out of it. While economic inequality had increased from the end of the 1980s until the latter part of the 1990s, the change was modest. Figures from Statistics Canada for 1989 and 1998 illustrate this. In 1989, the proportion of total-after-tax family income gained by the lowest 20 percent was 7.6 percent. In 1998 it was 7.1 percent. For the highest 20 percent, their portion of total-after-tax family income was 37 percent in 1989 and 38.8 percent in 1998 (Statistics Canada, 2000). 5 Two methodological issues deserve comment. First, although WVS respondents were given only the two response options for this question, interviewers coded ‘neither’ when it was volunteered. In the survey, 5.2 percent responded neither and 3.4 percent were coded ‘don’t know’. To make responses to the two surveys more comparable, we included ‘neither’ and ‘don’t know’ options in calculating the WVS percentage distributions and similarly included the ‘don’t know’ option in calculating the CSDT figures. Second, differences in survey populations should be noted. The CSDT sample was drawn from just four provinces while the WVS was drawn country-wide. However, analyses of the WVS results using only respondents from the four provinces sampled by the CSDT showed virtually no difference from the full survey in the overall distributions and very small differences in the party distributions. 6 Archer and Whitehorn (1997) report this figure as responses to the question ‘overall with regard to public ownership, how much should there be?’ 7 Data from the 1997 Canadian Election Survey were provided by the Institute for Social Research, York University. The study was funded by the Social Science and Humanities Research Council of Canada (SSHRC), grant number 412–96–0007, and was completed for the 1997 Canadian Election Team of André Blais (Université de Montréal), Elisabeth Gidengil (McGill University), Richard


03 Erickson (JB/D)

21/3/02

9:20 am

Page 323


8 9

10

11

12 13 14

15 16

17

Nadeau (Université de Montréal) and Neil Nevitte (University of Toronto). The Institute for Social Research, the SSHRC and the Canadian Election Survey Team are not responsible for the analyses and interpretations presented here. The wording of the Canadian Election Study question was: ‘We have gone too far in pushing equal rights in this country’. In the 1997 CES, NDP voters comprised 10.2 percent of all voters. As with the earlier WVS comparisons, we used the national sample in our comparisons here. Again, differences between the national sample and one using only respondents from the four provinces sampled by the CSDT survey are small. Conservative voters were the only group for which the four-province sample was substantially different; 42 percent disagreed with this statement. This scale uses two responses from Inglehart’s original scale and two responses from a second scale used by the WVS. Changes were made in order to make choices more relevant to contemporary circumstances and thus make the materialist options more appealing to social democrats. Two groups of questions were subjected to factor analysis, using varimax rotation. The first group consisted of those concerned with traditional welfare state programs. The second group included various fiscal, economic and equity items. The item on the role of the state versus individuals in providing for themselves loaded on the latter factor as well. For the tabular results of the factor analysis, contact the authors. See the Appendix for the wording of the various scale items. The NDP has a policy by which affiliated unions are accorded special representation at party conventions and in party councils. The union members in our sample have joined the party as individuals. Their unions may or may not be affiliated to the party. For region, Manitoba is the reference category. The Saskatchewan NDP has held provincial power for two-thirds of the post1970 period, the BC NDP for one-third of this time and the Ontario NDP for only one-sixth of this period. The results of the multiple regression analysis are available from the authors. In order to check whether some of the internal party divisions we identified above, especially the regional ones, are, in effect, grounded in a materialist/postmaterialist division, we also introduced the post-materialism measure as an independent variable in our regression models for our five issue scales. In fact, there was surprisingly little change in the pattern of coefficients (data not shown). This was true especially for the regional variables, all of which remained as robust as in the original equations.

References Abramson, Paul R. and Ronald Inglehart (1995) Value Change in Global Perspective. Ann Arbor, MI: University of Michigan Press. Archer, Keith and Alan Whitehorn (1997) Political Activists: The NDP in Convention. Toronto: Oxford University Press. Esping-Andersen, Gosta (1985) Politics Against Markets: The Social Democratic Road to Power. Princeton, NJ: Princeton University Press. Esping-Andersen, Gosta (1996) ‘After the Golden Age? Welfare State Dilemmas in a


03 Erickson (JB/D)

21/3/02

9:20 am

Page 324


Global Economy’, in Gospa Esping-Andersen (ed.) Welfare States in Transition: National Adaptations in Global Economies. Thousand Oaks, CA: Sage Publications. Everitt, Joanna (1998) ‘The Gender Gap in Canada: Now You See It, Now You Don’t’, Canadian Review of Sociology and Anthropology 35: 191–219. Fraser, Nancy (1995) ‘From Redistribution to Recognition? Dilemmas of Justice in a “post Socialist” Age’, New Left Review 212: 68–93. Garrett, Geoffrey (1998) Partisan Politics in the Global Economy. New York: Cambridge University Press. Gidengil, Elisabeth, André Blais, Richard Nadeau and Neil Nevitte (1999) ‘Making Sense of Regional Voting in the 1997 Federal Election: Liberal and Reform Support Outside Quebec’, Canadian Journal of Political Science XXXII: 247–72. Inglehart, Ronald (1990) Culture Shift in Advanced Industrial Society. Princeton, NJ: Princeton University Press. Inglehart, Ronald and Paul R. Abramson (1999) ‘Measuring Postmaterialism’, American Political Science Review 93: 665–77. Inglehart, Ronald, Neil Nevitte and Miguel Basanez (1996) The North American Trajectory: Social Institutions and Social Change. New York/Berlin: Aldine de Gruyter. Jenson, Jane (1997) ‘Fated to Live in Interesting Times: Canada’s Changing Citizenship Regimes’, Canadian Journal of Political Science XXX: 627–44. Kitschelt, Herbert (1994) The Transformation of European Social Democracy. Cambridge: Cambridge University Press. May, John D. (1973) ‘Opinion Structure of Political Parties: The Special Law of Curvilinear Disparity’, Political Studies 21: 135–51. Nevitte, Neil, André Blais, Elisabeth Gidengil and Richard Nadeau (1997) Unsteady State: The 1997 Canadian Federal Election. Don Mills, ON: Oxford University Press. New Democratic Party of Canada (1999) Resolutions Adopted at the 1999 Biennial Convention, August 27–29, 1999. Ottawa: Federal New Democratic Party. Offe, Claus (1987) ‘Democracy Against the Welfare State? Structural Foundations of Neo-conservative Political Opportunities’, Political Theory 15: 501–37. Seyd, Patrick (1999) ‘New Parties/New Politics? A Case Study of the British Labour Party’, Party Politics 5: 383–405. Seyd, Patrick and Paul Whiteley (1992) Labour’s Grass Roots: The Politics of Party Membership. Oxford: Oxford University Press. Statistics Canada (2000) The Daily, Tuesday, June 12. [http://www.statcan.ca: 80/Daily/English/000612/d000612a.htm]

LYNDA ERICKSON is Associate Professor, Department of Political Science at Simon Fraser University. She has published articles on party recruitment, leadership selection, women’s political recruitment and representation and voting patterns in Canada. ADDRESS: Department of Political Science, Simon Fraser University, Burnaby, BC, Canada V5A 1S6. [email: [email protected]]

DAVID LAYCOCK is Associate Professor, Department of Political Science, Simon Fraser University. He is author of Populism and Democratic Thought in the


03 Erickson (JB/D)

21/3/02

9:20 am

Page 325


Canadian Prairies (1990), The New Right and Democracy in Canada (2001) and various articles on democratic political theory, business–government relations and Canadian party ideologies. He is co-editor of The Puzzles of Power: An Introduction to Political Science (1993 and 1998) and Studies in Comparative Political Economy and Public Policy, a monograph series with the University of Toronto Press. ADDRESS: Department of Political Science, Simon Fraser University, Burnaby, BC, Canada V5A 1S6. [email: [email protected]] Paper submitted 16 October 2000; accepted for publication 25 May 2001.
