Jump to ContentJump to Main Navigation
Political Parties in Advanced Industrial Democracies$

Paul Webb, David Farrell, and Ian Holliday

Print publication date: 2002

Print ISBN-13: 9780199240562

Published to Oxford Scholarship Online: November 2003

DOI: 10.1093/0199240566.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 17 September 2019

Canada's Nineteenth‐Century Cadre Parties at the Millennium

Canada's Nineteenth‐Century Cadre Parties at the Millennium

Chapter:
(p.345) 12 Canada's Nineteenth‐Century Cadre Parties at the Millennium
Source:
Political Parties in Advanced Industrial Democracies
Author(s):

R. Kenneth Carty (Contributor Webpage)

Publisher:
Oxford University Press
DOI:10.1093/0199240566.003.0012

Abstract and Keywords

The electoral explosion that overthrew the established patterns of Canada's national party system in 1993 marked the end of yet another cycle in Canadian party system development, for there had been similar reshapings in the early 1960s and in the 1920s. In each case, the party transformations were about more than shifting electoral alignments and party fortunes, they also involved radical changes in the organization and activities of the parties concerned. However, although the formal organizational structure of Canadian parties has varied considerably, they have all been essentially cadre‐like in their structure and character, and the core linkage problem has been one of tying an American society to European‐style governing institutions. Electoral realignments have also been cast in geographic rather than social structural terms, and recent decades have seen the disintegration of the party system in a different sense as federal (national) politics has become increasingly disentangled from provincial politics. The introduction discusses these topics; the next three sections cover the same topics as those in the other country case studies in the book, and examine the possible crisis in party legitimacy in Canada, the cadre organizations of the Canadian cadre parties, and the functions of the Canadian cadre parties in a modern polity (governance, political recruitment, interest articulation and aggregation, participatory organizations, and political communication and education).

Keywords:   cadre parties, Canada, case studies, federal politics, governance, interest aggregation, interest articulation, national party system, national politics, participatory organizations, party functionality, party legitimacy, party organization, party performance, party system, political communication, political education, political participation, political parties, political recruitment, political system, provincial politics

The natural form of the political party risks being corrupted into an unwholesome caricature, a machine for winning elections.

The pure and simple continuation of their own existence becomes their principal preoccupation and measure of their ideals . . .

Constituted on the British model, Canadian parties have not escaped the American contagion . . .

André Siegfried (1906)

An electoral earthquake smashed Canada's national party system in 1993. The governing Progressive Conservatives were reduced from one hundred and sixty‐nine to just two seats while, at the other end of the traditional spectrum, the country's left‐wing (social democratic) New Democrats also suffered a major electoral collapse, leaving the party with its smallest ever caucus: neither had enough seats to be even recognized as a party under House of Commons rules. The authors of this party system explosion were two new parties, the Bloc Québécois and Reform, both of which offered fundamental challenges to the existing parties, the existing patterns of doing politics, and indeed to the country. The Bloc, which won the fourth largest vote share but, thanks to the vagaries of the first‐past‐the‐post electoral system, ended up with the second largest parliamentary caucus and so formed the ‘Official’ opposition, was a child of French–Canadian nationalism and advocated the break‐up of Canada through the separation of Quebec, the second largest province. Reform combined a right‐wing economic agenda, a western regional appeal, and a commitment to a populist politics that eschewed the established norms of parliamentary discipline in favour of a constituency‐focused delegate style of representation. Ironically, the one party that emerged least shaken by the eruption of these new parties was the Liberal Party, long the country's dominant political force. In opposition as the forces of destruction were gathering steam, the Liberals suddenly found themselves with a new majority government, though one based on a precariously small 41 per cent of the vote and, for the first time, dependent on central, English‐speaking Canada.

(p.346) This electoral explosion, with its overthrow of established patterns of politics, marked the end of another cycle in Canadian party system development. The party system that it broke had emerged out of a similar fundamental reshaping of Canadian party life in the early 1960s; that party system had in its turn been created in the aftermath of an earlier party system collapse and rebuilding in the 1920s. Each of these party system transformations was about more than shifting electoral alignments and party fortunes, though each was certainly marked by them. In every case the very character of the parties, their organization and activities, was changed: new structures and organizational linkages were established, party–leader relationships were altered, different bases of party and election financing were created, and party–media relationships were restructured. These reinventions of what Canadian parties were and did flowed from the demands made upon them by changing patterns of Canadian governance. This latest collapse is but the beginning of a new cycle that is forcing the parties to evolve in ways that respond to a new set of latent functions.1 As the country turned the millennium, the parties were struggling to invent a new party system.

Through these cycles of party system transformation the same, apparently familiar, Liberal and Conservative parties persisted and dominated the electoral map. They did so as classic examples of Duverger's (1964: 64) ‘cadre’ parties: ‘groupings of notabilities for the preparation of elections, conducting campaigns, and maintaining contact with the candidates’. At the beginning of the twentieth century Siegfried (1906: 113) commented that they were little more than ‘machines for winning elections’ and little has changed in the century since. This preoccupation with electoral activity, and absence of any enduring concerns for principle or interest, leaves Canadian cadre‐style parties free to organize, advocate and campaign as suits both their needs and the moment. Nestled in the confines of an ancient single‐member plurality electoral system, Canadian parties are poised to enter the twenty‐first century much as they left the nineteenth.

While there has been considerable variation in the formal organizational structure of Canadian parties, at heart they have all been essentially cadre‐like in their structure and character. Though the New Democratic Party has something of the mass structure of other social democratic labour parties, including affiliated trade union members, it too engages its members in parochial local associations whose primary focus is basically electoral. The parties' constituency associations, rooted in the electoral districts, are the fundamental organizational units to which members are attached. It is in these local associations that members choose their electoral candidates and select delegates to national conventions that (nominally) decide policy and (more importantly) elect and remove party leaders. After a series of reforms to the electoral and party financing laws in the early 1970s, the parties (p.347) were, for the first time, able to develop permanent national headquarters with staff devoted to research and communication. Coupled with the emergence of national television news and advertising, this stimulated both a pan‐Canadian thrust to party policy and reinforced the focus on party leaders in a party system that had long made leadership a regular electoral issue and centred party organization and decision‐making on national leaders.

In Canada, the core linkage problem for political parties has been one of tying an American (mobile, plural, growing, changing) society to European‐style (disciplined, centralized, closed) governing institutions. The solution to this problem has been an organizational trade‐off in which local autonomy is exchanged for national (parliamentary) discipline. This provides local party associations in each electoral district the freedom to run their local affairs, especially in deciding upon their candidate and managing his or her campaign, while the elected politicians support the leadership's dominance of policy‐making and party decision‐making in the capital. Cadre‐style organization persists as the dominant Canadian party form because it provides the organizational form best suited to incorporate this fundamental local–national balance. Thus, successful parties have an organization that is akin to a modern franchise system with the national organization producing and advertising the product (leaders and policy) while the independent local outlets deliver it (candidates and local campaigns) to the voters at election time. This puts the parties in the hands of whatever local candidates, activists and volunteers they can attract and hold. Given that the autonomy for discipline bargain is inherently unsatisfactory, party organizations are inevitably unstable; given that politicians and their supporters live for electoral success, exit trumps voice and loyalty, and electoral volatility is endemic.

Table 12.1 summarizes the patterns of Canadian national party politics over the last four decades. Though three parties claimed to be national organizations, the New Democrats hardly existed in the eastern half of the country and the first‐past‐the‐post system ensured that they would remain a minor parliamentary player. The vagaries of the electoral system produced two decades of oscillating government as both Liberals and Conservatives formed minority and majority administrations. Indeed, the Conservatives' electoral victory in 1988 marked the first instance of a majority government being returned to a second term since the early 1950s (Table 12.2) which made their electoral melt‐down in 1993 all the more dramatic. This parliamentary and governmental turnover did not reflect competition between competing visions of Canada, for all three parties had adopted a pan‐Canadian, accommodative approach to representation, and a commitment to a style of government that was a bilingual, Ottawa‐centred approach to executive federalism, animated by the practice of ‘federal–provincial diplomacy’ (Simeon 1972). For most of the period big constitutional issues about the very state and shape of the federation drove the national agenda. And it was the three traditional parties' apparent uncritical consensus on these issues that eventually led to the new parties' 1993 attack on the system.

(p.348)

Table 12.1. Canadian National Election Results, 1962–2000

Year

L

PC

NDP

SC

R

BQ

O

ENEP

ENPP

Volatility

Turnout ‘A’

Turnout ‘B’

1962

37.2

37.3

13.5

11.7

0.4

3.2

2.9

16.6

79

73

(265)

99

116

19

30

1

1963

41.7

32.8

13.2

11.9

0.4

3.2

2.6

4.8

79

74

(265)

129

95

17

24

0

1965

40.2

32.4

17.9

8.4

1.2

3.3

2.6

5.4

75

70

(265)

131

97

21

14

2

1968

45.5

31.4

17

4.4

1.7

3.0

2.3

5.8

76

68

(264)

155

72

22

15

0

1972

38.5

35

17.7

7.6

1.2

3.2

2.8

7.5

77

69

(264)

109

107

31

15

1

1974

43.2

35.4

15.4

5.1

0.9

3.0

2.8

5.1

71

63

(264)

141

95

16

11

1

1979

40.1

35.9

17.9

4.6

1.5

3.1

2.5

3.6

76

68

(282)

114

136

26

6

0

1980

44.3

32.5

19.8

1.7

1.7

2.9

2.4

6.3

69

64

(282)

147

103

32

0

0

1984

28

50

18.8

0.1

3

2.7

1.7

18.8

76

67

(282)

40

211

30

0

1

1988

31.9

42.9

20.4

2.1

2.6

3.0

2.3

7.2

75

68

(295)

83

169

43

0

0

1993

41.3

16

6.9

18.7

13.5

6.6

3.9

2.4

41.6

69

66

(295)

177

2

9

52

54

1

1997

38.5

18.8

11

19.4

10.7

0.3

4.1

3.0

7.8

66

59

(301)

155

20

21

60

44

1

2000

40.9

12.2

8.5

25.5

10.7

2.2

3.8

2.5

9.2

61

54

(301)

172

12

13

66

38

0

Notes: Top figure in each cell: share of vote won by each party; bottom figure in each cell: number of seats won by each party in Canadian House of Commons. Figures in parenthesis in ‘year’ column = total number of seats in House. ENEP—Effective number of electoral parties; ENPP—Effective number of parliamentary parties. Volatility; Pedersen index. Turnout ‘A’: % of registered electors voting; Turnout ‘B’: % voting age population voting

Party: L—Liberal; PC—Progressive Conservative; NDP—New Democratic Party; SC—Social Credit; R—Reform (in 2000 Canadian Alliance); BQ—Bloc Québécois; O—Other.

(p.349)

Table 12.2. Party complexion of Canadian national governments, 1962–2000

Election year

Dates of government

Status of government

Name and party of premier

From

To

Premier

Party

1962

21‐06‐57

22‐04‐63

Minority

John Diefenbaker

Conservative

1963

22‐04‐63

Minority

Lester B. Pearson

Liberal

1965

20‐04‐68

Minority

Lester B. Pearson

Liberal

1968

20‐04‐68

Majority

Pierre Elliott Trudeau

Liberal

1972

Minority

Pierre Elliott Trudeau

Liberal

1974

04‐06‐79

Majority

Pierre Elliott Trudeau

Liberal

1979

04‐06‐79

03‐03‐80

Minority

Joe Clark

Conservative

1980

03‐03‐80

30‐06‐84

Majority

Pierre Elliott Trudeau

Liberal

1984

30‐06‐84

17‐09‐84

Majority

John Turner

Liberal

1984

17‐09‐84

Majority

Brian Mulroney

Conservative

1988

25‐06‐93

Majority

Brian Mulroney

Conservative

1993

25‐06‐93

04‐11‐93

Majority

Kim Campbell

Conservative

1993

04‐11‐93

Majority

Jean Chrétien

Liberal

1997

Majority

Jean Chrétien

Liberal

2000

Majority

Jean Chrétien

Liberal

Note: All governments are single‐party administrations.

Canada is a country where geography often overwhelms history and nowhere is this more true than in its electoral politics. This is in no small measure a consequence of the continuing use of single‐member electoral districts that privilege geographic party organization and political appeals. The great electoral realignments of Canadian history have been cast in geographic rather than social structural terms, and the political equations of the successive party systems have been written geographically. From the First World War to the 1960s, the Liberals' winning formula was based on a partnership between the western prairies and French‐speaking Quebec. Then, after 1960, with the west in the hands of the Conservatives and Quebec still dominated by Liberals, politics came to be centred on Ontario where national elections were won or lost. The realignment that accompanied the 1993 electoral shockwaves broke that pattern, leaving the national system regionally fragmented with Reform in the west, the Liberals strongest in the centre, and the Bloc in Quebec. Of the principal parliamentary parties, only the Liberals could plausibly claim to be a genuinely national party, but their majorities now depended on the electoral system delivering them virtually all the seats in populous Ontario.

Recent decades also saw the disintegration of the party system in a different sense as federal (which Canadians call national) politics became increasingly disentangled from provincial political life. Canadians came to live in ‘two political worlds’ (Blake 1985), managed by competing levels of government, which were organized by distinctive and distinct political parties and party systems. Thus, a citizen in British Columbia would have her federal party politics structured by Liberals and Conservatives while Social Credit and the New Democrats organized her provincial electoral choices. Even in provinces where a party was strong at both (p.350) levels, the respective party organizations were distinct and often pursued competing goals. Provincial parties of the same name often stood for quite different ends. The result was a sometimes perplexing pattern of party activity across the country with one (regionally‐varied) national and ten provincial party systems tied together in only the most tenuous fashion. During the 1988 general election the Quebec Liberal leader (and provincial premier) urged his colleagues not to support the national Liberal party; in 1998 the leader of the national Conservative party resigned so that he could be acclaimed the leader of the Quebec Liberals.

As evidence of the extent to which federal and provincial parties have become separated, the 1993 political earthquake that broke the three parties' hold on the national party system made no impact on existing provincial party systems or the parties that made them up. They continued as they were with their distinct patterns reflecting the idiosyncrasies of the separate provincial political systems (Carty and Stewart 1996). The attack on the federal party system was a complex one, for it was mounted by parties that challenged the existing practices as well as the very legitimacy of the country, while simultaneously working a major electoral realignment in the geography of Canadian national politics. As Table 12.1 indicates, the level of electoral volatility experienced in 1993 was extraordinary, greater than for any other regular democratic election in the century (Mair 1997: 79, 216), and it produced a degree of parliamentary turnover remarkable even by Canadian standards. In the decade from 1984, over 80 per cent of the country's electoral districts were represented by members of more than one party.

Clearly there is something unique about Canada's political parties and the party systems they have produced. Organizationally they are fragmented, both regionally and by level of government, resulting in a complex of organizational ties that vary over time both within and across individual parties. Individuals, be they citizens, partisans, or even activists, have multiple and shifting partisan identities. Electoral volatility is comparatively very high. But for all these striking differences from parties in most advanced democracies, Canadian parties perform the same manifest functions of nominating candidates and conducting election campaigns on their behalf. If they are so different as organizations, or in the dynamics that drives their competition, it is because the challenges of linking society to state in Canada have been different. In this essay we are concerned with identifying how Canadian parties have gone about their particular business in recent decades and assessing how they will carry these traditions into the new century.

In the early 1990s Canada's national parties began, as they had on the occasion of previous party system collapses, to reinvent themselves. New patterns of membership, organization, and leadership politics were experimented with, new communication strategies were developed, and distinctive representational practices were adopted (Carty et al. 2000). Yet, in all this change, the parties' basic cadre structure has persisted, for it continues to provide them with the most satisfactory organizational solution to the problem of tying a deeply fragmented society to a regime of disciplined party governance where a geographically based electoral (p.351) system persists. The new party system is still in transition: it is not clear which of the parties will survive, or how they might do so; the parties have yet to institutionalize new structures and practices; and the new regional dynamics that echo the politics of interwar Canada have yet to be fully assimilated into national electoral competition.

A Crisis in Party Legitimacy?

Did the political earthquake that hit the party system in 1993 signal Canadians' wholesale rejection of party politics, or at least the political parties then dominating the system? Assessing the legitimacy of a system dominated by cadre‐style parties is difficult for that organizational form is designed to maximize flexibility, facilitate shifting allegiances (by voters and politicians), and so almost inevitably stimulates political volatility, all the while operating under the façade of great parliamentary umbrellas. When one or more cadre parties give way to others it is not always clear what, if anything, has really changed.

Canadians have always had an ambivalent relationship with their national political parties. In part this comes from the fact that Canada was itself a political creation, put together by party politicians acting from party‐driven motives. The first common Canadian experience was a national election, held to choose the new country's first parliament, for which political parties were required. Thus, from the very beginning state and then nation‐building were deeply entwined with party‐building: political parties remain among the few genuinely national institutions Canadians have. This also means that they carry a heavy load. When Canadians get disillusioned and dissatisfied with the world they find themselves in, and decide something must be done about it, their instinctive response seems to be to start by remaking their political parties. The result has been an ongoing tension between the populist impulses of a new nation and the country's dependence on parties for survival. In 1990, three‐quarters of the population told a Royal Commission on Electoral Reform and Party Financing that ‘we would probably solve most of our big national problems if decisions could be brought back to the people at the grass roots’ but the same proportion then agreed that ‘without political parties there can't be true democracy’ (Blais and Gidengil 1991: 16).

Earlier transformations of the party system have been marked by crises of legitimacy in which new parties have arisen to challenge the dominance of the old, stimulate organizational innovation, and mobilize Canadians in new ways. In each, deeply populist impulses threatened old ways of doing politics but were eventually absorbed by the system. While the major parties survived eruptions in the 1920s and 1960s, they did so by reconfiguring themselves and accommodating the system to new patterns of competition. 1993 was no different. The Bloc challenged the very legitimacy of the system by advocating that the country should be broken (p.352) into two; Reform argued that completely new kinds of parties, embracing populist norms and practices, were needed to transform the country's parliamentary and political life. Their success in capturing a third of the vote and dominating the opposition benches in the Commons crystallized an anti‐party sentiment that had been growing for some years.

Public opinion data reveals that Canadians have become increasingly cynical about politics. The proportions that report that ‘the government doesn't care what people like me think’ or ‘those elected to parliament soon lose touch with the people’ steadily grew (by 40 per cent) over the past quarter century (Blais and Gidengil 1991: 37–8). A good deal of this growing dissatisfaction has been aimed at parties. By the early 1990s, 78 per cent of Canadians were opposed to parties disciplining MPs in parliament, 47 per cent thought they offered no choice as all parties were basically the same, 81 per cent that there was too much party squabbling for good government, and 87 per cent agreed that parties simply confused the issues. Blais and Gidengil (1991: 41–4) concluded that, despite their recognition of the need for political parties, ‘the main thrust of Canadians' reactions is inescapable: they mistrust parties.’ Table 12.3 illustrates this growing alienation of the electorate from the parties: the numbers reporting ‘a good deal’ or ‘quite a lot’ of respect and confidence they had in parties fell from 30 per cent of the population in 1979 to just 10 per cent by the century's end, while the average party score on a feeling thermometer (running from 0 to 100) dropped from 56 in 1968 to a failing grade of 39 in 1997. This rejection was widespread: Clarke et al. (1998: 4) report

Table 12.3. Feelings Towards Canadian Parties, 1968–99

Year

Feelings about partiesa

Confidence in partiesb

1968

56

1974

53

1979

53

30

1980

51

1984

50

22

1989

18

1990

47

14

1993

43

9

1995

12

1996

11

1997

39

10

1998

13

1999

11

(a) average score on a thermometer (0–100) rating scale for the parties;

(b) aggregate percentage of respondents answering either ‘a great deal’ or ‘quite a lot’ to the question ‘Would you tell me how much respect and confidence you have in political parties?’.

Sources: Canadian Election Surveys; Gallup poll data.

(p.353) that ‘negative party performance judgments transcend demographic, partisan and various evaluative cleavages in the electorate.’

Despite this reaction against their parties, and the system of competition they fostered, Canadians' inherent ambivalence about, or perhaps dependence upon, political parties shone through for they responded by starting more of them. In Canada, as befits a country with cadre parties, a national political party is simply defined(by law)as an organization running at least fifty candidates under a common label in a general election. In 1972 there were four, by 1980 this number had more than doubled to nine and then, another ten years later, grown again to fourteen. In the aftermath of the system‐shattering earthquake the number fell back to ten for the 1997 election. While many of these parties hardly registered on the electorate's consciousness (nine of them together received only 2.6 per cent of the vote in 1988, see Table 12.1), two of them were instrumental in breaking the system up in 1993. Doing so in the face of the obstacles provided by the single‐member plurality electoral system was a remarkable feat, but one which spoke to the crisis of legitimacy faced by the established parties and the inherent fragility of a system of cadre‐style parties.

Growing antipathy towards the national political parties was not initially reflected in any systematic decline in Canadians' willingness to vote (refer again to Table 12.1). From the 1960s through the 1980s turnout varied but averaged about 75 per cent of the registered electorate (turnout in the provincial party systems differed and varied by province). But then, in the 1990s, turnout took a precipitous drop, falling to a record low of 61 per cent in the 2000 general election. Though the voter registration system was changed in 1997 (from election period enumeration to a permanent voters' list), that did not halt the erosion in the numbers of Canadians participating: in 1997 the proportion of those of voting age who bothered to vote fell to below 60 per cent for the first time in modern Canadian history. This sharp decline occurred despite the fact that both Reform and the Bloc were offering voters new and distinctive choices with the Bloc attracting many to the polls who had previously eschewed participation in federal politics (Johnston et al. 1996b). While the crisis in the party system was driving some Canadians to new parties, it was leading many others to abandon party politics all together. The result was a majority government elected by just 25 per cent of the electorate.

The very volatility of the electorate (Table 12.1) indicates the fragility of the ties between voters and parties. The 1993 explosion of the system was but the last of several occasions in the century at which the electorate, its shifts magnified by the electoral system, reconfigured the party system. To some degree the movement of Conservative voters to Reform may have represented a shift within the same party family, but Reform also attracted votes from populist New Democrats while the Bloc was redefining the very basis of national party competition in Quebec. With such a volatile electorate these new parties could shrink as fast as they have grown as the history of earlier protest parties reveals. Determined to persist, Reform sought to reconfigure and then institutionalize a united right by folding itself into (p.354) a new Canadian Alliance party in the run‐up to the 2000 general election. The Conservatives' refusal to cooperate with this attempt to restructure the party system suggests that large numbers of (central) Canadians see Reform/Alliance as an illegitimate interloper into the system.

Assessing the strength of the parties through the extent and intensity of the party identifications held by the electorate is difficult. First, there is the problem of measurement, for in a parliamentary system current vote intentions and partisan identifications can easily be blurred. Panel studies in the 1970s and 1980s revealed that 30 per cent or more of Canadians switched identifications between elections but whether this reflected fundamental identification changes or was in part simply tracking changes in party popularity is not entirely clear. The second problem flows from the fact that most Canadians live in two political worlds, structured by different party systems, with the result that they may need two sets of identifications. Thus, over the decade of the 1970s only about a third of the electorate maintained a consistent partisan identification at both levels of party life (LeDuc et al. 1984). Both this instability and inconsistency suggest relatively weak ties between voters and party.2

But did the pattern of party identification contribute to the collapse of the party system? Table 12.4 reveals that there was a slow erosion in the proportion of ‘very strong’ identifiers over the last third of the century: Clarke and Kornberg (1993: 303) estimate it shrank at the rate of 0.57 per cent each year between 1965 and 1991. At the same time there was no significant trend in the incidence of nonidentification, suggesting not so much a withdrawal from parties per se as a simple weakening of the hold that the parties had on Canadians. This left them particularly vulnerable to the appeal of new parties promising to do politics differently.

Cadre parties do not need members; they just need small groups of supporters to maintain their inter‐election organizations and then larger numbers of election period activists. As a result Canadian parties have never, until the 1990s, attempted to develop national membership programmes. No two parties have organized their members in the same way, and no party is organized the same way in all parts of the country. Their basic pattern has been to enrol members through local associations organized in the electoral districts.3 These associations have been more (Conservatives and Liberals) or less (New Democrats) permeable, but have typically grown in election periods only to shrink again in quiescent inter‐election years, and both of the new parties quickly fell into this same pattern. Table 12.5 illustrates the election‐year centred cycles of surge and decline as individuals renew (p.355)

Table 12.4. Federal Party Identifiers in Canada, 1965–97

Year

Very strong identifiers (%)

No party identification (%)

1965

24

11

1968

26

9

1974

27

13

1979

26

13

1980

31

10

1983

23

15

1984

22

14

1985

17

20

1986

15

27

1987

18

20

1988

23

11

1989

16

16

1990

20

14

1991

13

30

1993

9

22

1997

22

25

Note: Election years indicated in italics.

Source: Clark and Kornberg 1993; C.E.S. 1993, 1997.

Table 12.5. Canadian Party Memberships

Year

Liberal

Conservative

New Democrat

Reform

Total

1987

148,680

132,750

180,835

462,265

0.84

0.75

1.03

2.62

1988

279,365

312,995

193,225

785,585

1.58

1.77

1.10

4.45

1989

199,715

192,930

189,390

582,035

1.13

1.09

1.07

3.29

1990

317,125

111,510

200,010

628,645

1.80

0.63

1.13

3.56

1992

212,695

114,460

198,535

525,690

1.07

0.57

1.00

2.64

1993

242,195

223,610

227,150

692,955

1.22

1.12

1.14

3.48

1994

139,240

84,075

148,680

371,995

0.70

0.42

0.75

1.87

Notes: Election years indicated in italics; Membership estimates are calculated from surveys of local associations. Top figure in each cell is absolute membership, bottom figure is M/E ratio.

their memberships in order to vote in candidate nomination meetings and work in campaigns. The big Liberal growth in 1990, a non‐election year, reflects this same pattern for the party had a leadership contest that year and memberships were reactivated, or taken out, just for it. In Canada's cadre parties, membership (p.356) has traditionally been valued for the vote it provided—for a local candidate or national leader—but little else. These really are skeleton organizations, little more than Duverger's ‘groupings of notabilities for the preparation of elections.’

The extraordinary fluidity of these patterns of party life extends to the notables themselves around whom activists are gathered. It is not uncommon for individuals wanting to become a Member of Parliament to join a party on the eve of their announcement that they are planning to contest a local nomination. For their part, the parties themselves will often scavenge the neighbourhood looking for a popular local figure who might get elected: in 1997 Reform took to advertising for candidates in local newspapers. Preoccupied with electoral advantage, the parties' openness extends to recruiting individuals previously associated with their opponents. Thus, in the 1993 landslide election 27 per cent of the established parties' candidates had once been active in some other party while the figure was much higher in the ranks of the new challengers.

Table 12.5 also reveals just how shallow is these cadre parties' penetration of the electorate. They only manage to enrol somewhere between 0.5 and 1.5 per cent of the electorate so that even in years of the greatest political activity less than 5 per cent of the electorate hold even these most nominal sorts of memberships in any national political party. In cadre parties, members are drawn to successful politicians so it is not surprising that constituency memberships are typically about twice as large in associations where there is a sitting Member of Parliament. Despite its façade of a mass membership organization, this same phenomenon has also characterized the New Democrats whose incumbents traditionally had local memberships six times the size of the party's non‐incumbent associations (Carty 1991: 40). It is also important to note that the tie between local activists and the parties is a highly personalized one: over half of the local associations claim that with a different candidate they would attract a different set of volunteers (and members for the moment) to work on their election organization (Carty 1991: 175). Finally, this partisan involvement is fleeting: only half of those who worked in the 1988 general election campaign for their party had participated in the previous election four years earlier (Carty 1991: 173).

The existence of a system of income tax credits for donations to political parties means that ordinary Canadians can support parties financially without taking out membership. Table 12.6 reports the number of individuals who have made contributions to the national parties over the past two decades and provides a second measure of the extent to which the parties are rooted in the population. It indicates that only about 1 per cent of the electorate contributes to a national party in any given year, but that this too varies across the electoral cycle, typically peaking in election years. For their part, local associations report that only between one‐fifth and one‐third of their members make regular contributions, confirming our portrait of them as weakly tied to their party (Carty 1991: 82–4). The New Democratic Party, with its more bureaucratic mass‐style structure, has developed a more consistent set of donors but that advantage has been offset by somewhat (p.357)

Table 12.6. Number of Individuals Contributing to Canadian Parties, 1975–99

Year

Liberal

Conservative

New Democratic

Reform

Bloc Québécois

1975

13,373

10,341

58,889

1976

18,261

23,409

56,142

1977

21,063

20,339

60,169

1978

22,350

35,615

67,133

1979

13,025

34,952

63,655

1980

17,670

32,720

62,428

1981

24,735

48,125

56,545

1982

27,968

52,694

66,665

1983

33,649

99,264

65,624

1984

29,056

93,199

80,027

1985

28,545

75,117

97,364

1986

35,369

52,786

90,487

1987

28,972

39,320

87,927

1988

30,642

53,893

118,390

1989

19,970

40,191

89,290

7360

1990

36,361

27,702

116,448

23,462

1991

26,396

27,391

94,080

43,176

1992

29,025

27,823

75,213

55,760

1993

41,058

44,728

65,301

49,488

1994

36,880

14,532

54,446

28,970

29,084

1995

39,019

15,870

55,438

32,982

25,848

1996

37,471

18,859

48,972

66,982

17,030

1997

34,429

23,352

50,434

75,587

18,885

1998

32,710

17,908

45,385

55,405

10,533

1999

30,735

16,437

41,760

53,262

13,104

Notes: Election years indicated in italics. The parties use different financial systems so that the numbers may not be strictly comparable: in some parties membership fees go to local associations, in Reform they may be counted as contributions to the national party; the New Democrats' party finances are so deeply entangled that it is difficult to distinguish between contributions to their provincial and national organizations.

smaller average gifts. The 1993 collapse of the Conservatives was prefigured in the collapse of their donor base: it fell from almost 100,000 in 1983, the year before its great electoral victory, to just 28,000 in 1992. But if Canadians were no longer supporting the Conservative Party the same could not be said of the party system in general: as many Canadians (200,000) were prepared to reach into their pockets to support a major party in 1993 as had been in 1988 or 1984.

Ambivalent partisans, ambivalent because the party system is an imperfect mechanism for linking them to their government, populate Canadian parties. While the fragmentation of the party system contributes to this ambivalence it is also true that ambivalence contributes to fragility and fragmentation, as Canadians opt for exit over loyalty. The 1990s crisis in the party system was in part a legitimacy crisis, but it was also part of a process of challenge and change by which cadre parties are (p.358) forced to rebuild their organizations and rethink their practices, and by which the party system reorders the shape of competition. In rejecting the pan‐Canadianism common to all the parties over the third (1960–90) party system, the new parties have redefined the national political agenda and stimulated organizational innovation. What they have not done is alter the basic cadre style of Canadian party life. Indeed, their easy acceptance of this form of party organization testifies to how deeply rooted it remains.

Cadre Organizations

Cadre parties operate at two levels, at the local grass roots where individual members join and participate, and at the national level which focuses on the parliamentary life of the caucus. Sitting rather precariously between them are modest national offices that try to organize and coordinate local party activity in the interest of the central leadership. The relationship between these two levels of party life is uneasy. The constituency associations jealously guard their autonomy and prerogatives in the face of a national party that would like to impose common practices and be able to reach into the associations to influence their decisions. These conflicts most often arise over the constituency associations' long‐standing right to choose (or remove) its local candidate. Attempts by the national party to impose its will, though sanctioned by party rules and the law, are seen as illegitimate by the membership. Too heavy a hand will simply result in members deserting. Given that serious national parties believe they must mount at least a nominal campaign in each constituency, this makes their organization heavily dependent on the enthusiasm of the local volunteers they can attract and hold. For their part, to protect themselves, incumbents work to make the party organization in their constituencies a personal one and try to keep the national office out of their affairs.

Individual party members have little to do between elections. In the constituencies there are routine tasks of organizational maintenance as well as purely local communications and policy work, but none of this takes many members. It is only at election time that the organizations spring back into life and in the process enrol large number of members: a contested nomination may see thousands registered in a matter of weeks. However, where there is no contest for the nomination, memberships may change little. Thus, membership levels can be very misleading guides to organizational strength for they often simply reflect the desire of individuals to participate in a contested party meeting. The very fluidity and openness of cadre‐style organizations, coupled with minimal inter‐election demands, means that membership is a relatively low priority for the parties.

Most constituencies across the country have local units of at least four of the (post 1993) national parties. Their strength and activity vary sharply across both space and time within the same party. As vote‐gathering machines, a party's electoral (p.359) victories stimulate healthy local associations, for access to local representatives provides members with an incentive to remain active. This is why continuing electoral success is directly related to membership strength. But does weak organization explain electoral defeats, or strength account for victories? In the case of Canadian parties the answer is more likely to be the reverse—electoral defeats produce weak organizations, wins lead to strong local associations. Thus, when the Conservatives suddenly made a breakthrough in Quebec in 1984 it quickly developed a full set of strong local organizations where none had existed for most of the century. Within a few years of the party's defeat in 1993 that organizational presence had vanished so that many local associations were not even able to find candidates on their own for the 1997 election (Woolstencroft 1997: 77).

The national party organizations maintain central offices but these are largely charged with routine organizational maintenance tasks and running the periodic national conventions. In some parties the headquarters staff manage ongoing direct mail fund‐raising campaigns but in others it is assigned to separate organizations. In none of the parties is there a highly developed permanent staff of career party functionaries. For most of the country's history the parties were only able to raise funds at election time and so no tradition of strong or substantial party offices developed. Now that the parties have more resources the local associations do not want a party bureaucracy that would inevitably interfere with them, while the elected politicians have taken care of themselves (parliamentary and constituency staffs supported by the state have grown over the past two decades) and do not want to see permanent party officials who might rival their pre‐eminence in the party. When the parties have lots of money party offices tend to expand but then quickly contract when revenues fall. The result is little continuity in party field workers and the absence of a permanent organizational staff whose work might transform the cadre cast of the parties.4

Canada's cadre parties exist to contest elections and there are three organizational corps that constitute the real sinews of their electoral machine and so are critical to their working and success. None are part of the formal apparatus of the party; all exist as informal, personalized extensions of electoral politicians. Directed and staffed almost entirely by shifting cadres of volunteers activated for a contest, they spring into life for electoral contests and then retreat into quiescence between contests. In the subtle, continually shifting networks of relationships among those who make them up lies the organizational heart of the cadre party. First are the organizational teams put together in the constituencies to fight local contests. These teams centre on the candidate and reflect his or her approach to politics and place in the constituency. Run by amateur local campaign managers, and official agents, who are appointed by the candidate and answer to him or her, (p.360) they are staffed by volunteers who are there to support the particular candidate (Carty 1991: ch. 7; Sayers 1999). Traditionally autonomous, these campaign teams are only loosely integrated into the national campaign despite the fact that together they spend more than the national party during a general election campaign. The second key organizational group is the national campaign committee and those who work for it. These are the ‘war‐time generals’ brought in to organize the parties' national electoral campaigns in place of the ‘peace‐time generals’ who run the party offices between elections but are displaced for the critical business of fighting elections. The chair of the campaign team, and the key members of it, are appointed by the party leader and owe their positions and influence to their personal ties to him or her. These individuals generally hold no formal position in the party but their key electioneering role, and their strong ties to the leader, make them among the most powerful figures in the nether worlds of Canadian parties. Finally, the leadership selection politics of Canadian parties forces those who aspire to become party leader, and thus dominate the national organization and parliamentary caucus, to build party‐wide networks of supporters. The result is the development of personal factions within the parties, organizations that can reach well beyond the circles of the national capital down into the local organizations of as many constituency associations as possible. These highly personalized networks, defined by their loyalty to an individual, may lay dormant for long periods of time but can be quickly activated for leadership challenges or contests. It is the competing loyalties of these three sets of organizational groups, first to their patron, second to the party, that drive both intra‐party and inter‐party competition in Canada and leave the parties with extraordinarily fluid, candidate‐centred organizations.

To all this the social democratic New Democratic Party sought to be an exception. With its mass‐membership structure, and its provision for the affiliation of trade union locals, it provided for an organization that was less electorally focused and less dependent on personality. However, as a federation of provincial party organizations, its strength is dependent on the vitality of the party in the provinces, and not being a player in several of the provincial party systems has weakened its national organization significantly. Where the party was most successful it succumbed to the cadre instincts of Canadian party organization with personal machines growing up around party leaders and its incumbent politicians.

For most of their history Canadian parties were relatively poor, able to raise money at election time but then doing without during the inter‐election periods. The increasing cost of electronic politics forced the parties to realize that that model of financing was no longer workable and in the mid‐1970s major reforms to election and party financing were passed by parliament that: provided for state subsidies (through both direct payments and indirect tax expenditures); limited campaign (but not inter‐campaign) period spending; and established a regime of public disclosure and accountability. These changes quickly changed the parties, significantly increasing their organizational strength and capacity. However, the new arrangement was limited by the decision not to regulate the financial affairs (p.361)

Table 12.7. Canadian Party Income, 1976–99

Year

Liberal

Conservative

New Democrat

Reform

Bloc Québécois

1976

5823*

4084

2925

1977

4587

3774

3525

1978

5018

5465

4184

1979

6302

8376

6020

1980

7457

7564

6101

1981

5592

6950

6003

1982

6746

8521

7108

1983

7736

14,767

8669

1984

11,598

21,979

10,513

1985

6163

15,073

10,152

1986

10,719

15,639

14,639

1987

8882

13,058

12,608

1988

16,358

25,231

18,754

1989

6397

14,521

13,865

13,511

1990

12,038

11,045

15,439

2212

1991

6776

12,037

19,933

6588

1992

7555

11,542

13,819

8543

1993

14,723

22,276

18,227

8357

1994

11,764

4131

10,722

5675

2053

1995

13,229

5576

12,394

4337

1683

1996

14,113

6708

12,074

7162

1160

1997

17,481

10,982

14,012

8800

2140

1998

13,714

5813

5527

5772

727

1999

14,627

5140

6422

6284

1297

Notes: All figures in thousands of $Cdn.

(*) Includes the party income for both 1975 and 1976. Election years indicated in italics.

of the local associations or leadership contests. Given their vital importance in a cadre system, this has left a large ‘black hole’ in Canadian party finance and our knowledge of the parties' organizational life.

Tables 12.7 and 12.8 report the ebbs and flows of party income and expenditure since 1975 and the introduction of a partially regulated and publicly subsidized party finance regime. As Stanbury (1991) has demonstrated, the parties' finances are anything but transparent and no two of them operate in the same way so that the data have to be read with considerable caution.5 But that said, there are several distinct features of the national parties' finances that stand out. First, there was a noticeable increase in the total amount of money in the party system beginning in the mid‐1980s; this has been broadly sustained in real terms since then. This increase was led by the Conservatives' success in raising money by direct mail (p.362)

Table 12.8. Canadian Party Expenditures, 1976–99

Year

Liberal

Conservative

New Democrat

Reform

Bloc Québécois

1976

4707*

3497

2381

1977

4187

4233

3105

1978

5283

5470

3514

1979

2771

5184

4678

1980

3702

4923

5992

1981

5116

7542

6491

1982

6781

8521

4871

1983

6277

13,199

8009

1984

11,999

20,777

7407

1985

8149

11,654

11,071

1986

11,166

14,141

15,188

1987

9274

13,490

14,012

1988

10,176

21,124

14,933

1989

7115

12,824

12,507

966

1990

11,587

10,384

14,262

1721

1991

6769

11,534

18,711

6289

1992

6931

10,697

14,237

89,211

1993

15,517

22,197

13,632

6338

1994

11,707

2403

9411

2921

2053

1995

10,687

4191

14,506

5661

2001

1996

13,193

6590

13,575

7220

1289

1997

26,664

21,965

17,397

11,138

4807

1998

12,790

5102

10,770

6156

711

1999

12,605

5138

15,088

6246

1052

Notes: All figures in thousands of $Cdn.

(*) Includes the party expenditure for both 1975 and 1976. Election years indicated in italics.

from supporters, but not necessarily members, and soon emulated by the other parties. These contributions from individuals carried an income tax credit, so were effectively being partially financed by the state. This practice pointed to the distinction between supporter and member in which the former was often more valuable to the politicians.

The second regular feature of the parties' finances is the sharp increase in election years when expenditures typically double inter‐election year amounts. This reflects the relative ease with which the parties are able to raise funds for election campaigns as well as the public subsidies that the state now provides for their election expenses. It is also a powerful reflection of the fact that these parties are election machines that conserve their energy for electoral activity and do not believe that their national organizations have a significant role to play in public life between elections. There are non‐general election years in which a party will spend large amounts of money (e.g. the Conservatives in 1983 or the Liberals in 1990), but they too are election‐centred for those instances of high spending are typically periods in which the party is seized in the throes of a leadership contest.

(p.363) The third point to note is that the governing party is more successful at raising money, and hence spends more, than any of its opponents: the Conservatives were the richest party from the mid‐1980s until 1993 when they were passed by the Liberals after they came to office. This pattern is surely typical of cadre parties whose principal difference, as far as many individual and corporate donors are concerned, is that one is in power, the other not. The long‐standing practice of many corporations has been to give to both parties—but more to the one in government. It is also true, however, that the Liberals and Conservatives, the only parties ever to have held national office, are regularly better financed than the other parties. Despite Reform/Alliance's considerable success in displacing the Conservatives from their western parliamentary strongholds and as the principal party of the right, the country's economic establishment did not desert the Conservatives as quickly as the voters and so the Conservative Party had the resources to weather the worst fallout of the 1993 electoral earthquake.

A fourth aspect of Canadian party finances is the state's approach to regulation and subsidization. That Canadian party (and candidate) expenditures are regulated, and subsidized, only for the few weeks in the multi‐year election cycle testifies to the public conception of them as cadre organizations whose only significant activity is electoral. That the parties do not attempt to exploit these long periods when they are unregulated to spend more in them testifies to their own conception of themselves as essentially vote‐gathering machines. This focus on the parties' public, electoral, activity has left their internal affairs completely unregulated. The result has been leadership contests in the major parties, at which the prime ministership was at stake, in which the total amounts spent by the candidates exceeded those allowed the whole party in a general election, and local nomination contests within individual constituency associations in which would‐be candidates spent more than the winner was then allowed to spend in the election contest itself (Carty et al. 2000: ch. 7). These monies do not always make their way into party accounts and it is difficult to know what the true magnitudes, or sources, are. The commitment to local association autonomy that has been integral to the operation of these cadre parties has meant that local finances are both unregulated and unreported. Considerable sums are believed to exist in many local bank accounts and trust funds but their size and use remains one of the mysteries of Canadian party life.

Cadre parties are focused on the politicians so it should be no surprise that the spending limits that now govern Canadian elections permit the local candidates, as a whole, to spend more than the national parties in their quest for office. In the 1997 general election, a national party's candidates could spend up to $18.8 million, about 65 per cent more than the $11.4 million allowed by the party for its national campaign. While the national parties could actually exceed those limits due to loopholes in the law (e.g. polling expenses are not considered election expenses), the bias in favour of the local campaigns, which are rarely integrated into the national efforts, is considerable (Sayers 1999). The continuing ability of many local party associations to raise more than they are allowed to spend (p.364) helps them maintain their organizational independence from their party's national office. Thus, the overall pattern of party finance that has evolved in the past two decades is one that seems particularly well designed to meet the needs of the local associations and more especially the local politicians who dominate them. At the same time, the relatively modest limits on individual candidates keeps the system permeable to new entrants and so intrinsically volatile.

Canada's cadre parties have not been transformed into modern cartel party organizations (Young 1998). Though both parties and candidates receive some public subsidies to support their election expenditures, and while individuals can receive a (steeply graduated) tax credit for political donations, the proportion of public funding to both has been estimated at about a third of their total spending (Young 1998; Carty et al. 2000: ch. 7). This is unlikely to change as the opposition Canadian Alliance Party is opposed, in principle, to state support for political parties, although until the law is changed their politicians continue to take full advantage of it.

One hundred years ago Canadian parties were deeply enmeshed in the media but that era has long passed. In the absence of any enduring press–party linkages or support patterns, the political parties must compete with other groups and individuals for the media's attention. Between elections their national organizations (as opposed to their parliamentary caucuses which seem to live for the front page or lead story) make relatively little effort to do so. At election time they require that broadcasters provide some free‐time for election broadcasts, allocated by rules agreed to by the parties and administered by a Broadcasting Arbitrator working through the independent Chief Electoral Officer, and make other time available for purchase. In recent years the parties have experimented with the private media—direct mail, the Internet—and it is clear that the use of these technologies will grow in the future. Public opinion polling is, of course, an important medium (one in which information flows upwards to politicians rather than downwards from them) and the parties have maintained their ability to use it by excluding its costs from the limits otherwise imposed on election expenditures. This is one area where the parties have established an advantage over their local associations, for the costs associated with polling continue to exceed the reach of most constituency parties.

Everything is done twice, and often differently, in Canadian political organizations, once in English, once in French. As serious national parties must reach Canadians in both major language groups this forces them to establish two distinct media communications groups in their organizations.6 During election periods the French‐speaking and English‐speaking campaigns are organized quite separately from one another and develop distinctive messages for the two language communities. The very limited integration of the media in the two parts of Canada (p.365) makes this possible, while the different agendas and partisan balances driving politics in the two sides of the country makes it necessary.

The portrait we have of Canadian parties' organizational strength is opaque. It is so because they remain modern examples of nineteenth century institutions. These cadre parties thrive when they win, but struggle when they do not. Indeed defeat often stimulates an ‘opposition party syndrome’ (Perlin 1980) in which the politicians that survive the defeat fall prey to internal factionalism and leadership conflicts, which in turn makes candidate recruitment and local organization‐building difficult, leading to defeatism and the inability to present a convincing face to the electorate and so another election defeat which is then, more often than not, followed by another round of the syndrome. This is a pattern that persists because the style of party organization persists. What has changed over the years is the role Canadian parties have played in the country's governance.

Cadre Parties in a Modern Polity

The parties' basic manifest functions remain: they continue to monopolize the task of nominating candidates and conducting election campaigns, though the latter role is being challenged by the increasing involvement, at both national and local levels, of interest groups in the electoral process. By comparison, the parties' latent functions have shifted dramatically as Canada has evolved (Carty 1997). In the early decades (1867–1920) the parties were engaged in state‐building as they controlled the public service. The patronage politics that first party system produced eventually saw the parties consume the state. The succeeding party system (1920–60) forced the parties to become nation‐builders, operating as the agents of national accommodation through a politics of regional bargaining and brokerage. By the end of that second era the state had consumed the party and the Liberals were left as the ‘Government Party’ (Whitaker 1977). The third party system (1960–93), now crushed by the earthquake of 1993, pushed the parties into the role of pan‐Canadian agenda setters as the effective centre of Canadian governance was moved to the apartisan setting of the intergovernmental conference room where federal and provincial governments bargained public policy. It was the very flexibility of their cadre organizational style that allowed the parties to respond to the very different functional demands the political system made of them in each of those different periods. At issue now is the capacity of this old form to serve modern needs.

Governance

The hard question is to what extent does Canada enjoy party governance? The easy answer is that by their very nature cadre parties can provide the form, but rarely the (p.366) essence, of party government. In reality, any interpretation of Canadian parties' role must be more nuanced than that simple judgement allows. In his recent book on the organization and operation of the national government, Donald Savoie argues that Canadian government is organized as prime ministerial government: the parties ‘are essentially election‐day organizations’ incapable of even ‘articulating a plan of action’ let alone implementing it ‘should it come to power’ (Savoie 1999: 344). Bakvis' study of the role of ministers corroborates this picture for he argues that ‘party is still important, but primarily as one of the arenas in which ministers jockey for position and influence’ (Bakvis 1991: 286). Yet, despite consigning party to the narrowly electoral dimension of national politics, Canadian politicians persist in practising a classic Westminster form of party government. Five of the twelve elections since 1960 failed to return a parliament with a majority party but in none of those instances was a coalition government ever seriously contemplated let alone created. A strong commitment to the norm of one‐party government (in the name of electoral accountability) leads party leaders to form minority governments in such situations, confident that the plurality electoral system will soon return the system to normal by delivering a majority at the next election.

The long‐standing party bargain of local organizational autonomy for parliamentary discipline produces national party caucuses that are highly cohesive. Though recent studies are beginning to challenge the myth that all MPs are nothing but ‘trained seals’, loyalty is the overwhelming principle (Massicotte 1998; Wearing 1998). Elected members who are unhappy with their party, or the limitations imposed on their parliamentary life, are as likely to choose the exit as the voice option. The result is a relatively high turnover rate, and correspondingly short parliamentary careers (Docherty 1997), a pattern that strengthens the propensity to prime ministerial over party government. With neither the power to select nor remove the leader, the caucus can provide little in the way of a party check on the government.

Disciplined one‐party majority government does mean that Canadian governments can pursue distinctive agendas and can move national public policy in their preferred direction. The Liberals demonstrated this after the 1993 election when they sharply reversed the direction of the country's fiscal policy with dramatic reductions in expenditures, which quickly turned large government deficits into surpluses for the first time in almost three decades. Whether there was anything distinctively Liberal about these policies is more contentious; it was the Liberals who had first created the huge deficits during an earlier turn in office. Canadians who voted for the Liberals in 1993 should have been surprised for they had run on a programme calling for more government spending and the abrogation of two of the then Conservative government's main policies—a new Goods and Services Tax and the North American Free Trade Agreement. In fact, they did neither and soon had embraced both. Thus, while governments can, and do, make a difference, it is rarely a party difference. It is precisely this that has led increasing numbers (p.367) of Canadians to believe that those they elect soon lose touch, and provided the impetus to the rise of the Reform party.

Canadian government may be party government but it is also federal government in a country in which the national and provincial party systems have become detached. For constitutional reasons many of the areas of major (i.e. expensive) public policy require the close cooperation of both levels of government in both the instigation and implementation of public services. This has meant that over the past several decades the most significant public policy decisions have been made in a complex system of intergovernmental competition and accommodation. Besides strengthening the hands of government leaders vis‐‘à‐vis their own parties, this has led to something of an apartisan decision‐making process. With representatives from four or five different parties inevitably involved in intergovernmental bargaining, partisan considerations must often be put aside in the interest of reaching agreement. Not only does this process drain most public policy of a distinctly partisan colouring, but it also shields most decisions from any clear process of democratic accountability. Since cadre parties seek to provide for party government in only the narrowest sense they do not see this as a problem.

There are three ways in which this modest form of party government is now under challenge. The first is from the Charter of Rights and Freedoms, which was written into the constitution in the early 1980s. It has reinforced a political discourse of rights, stimulated more Canadians to take part in politics through interest groups as opposed to parties, and inserted the Supreme Court as a major actor in the political process. The result has been that parties are no longer able to dominate, or always even to define, the national agenda, that they must compete for resources with interest groups which now actively intervene in election campaigns, and that the courts may strike down their policies or force governments to adopt policies that divide their party supporters.

The second challenge to traditional forms of party government comes from the Reform/Canadian Alliance Party with its commitment to altering the very character of parliamentary life in Canada. The party, drawing on deep‐seated populist impulses in the Canadian west, argues that the very nature of representation should be changed. It seeks to replace politicians’ fealty to their party with obedience to the will of their constituents. This attack on the sanctity of party discipline, which is inherently precarious in a cadre party, strikes at the very heart of party government as it has been practised in Canada for over a century. Though it has not yet been successful in its campaign against party discipline, nor does its caucus particularly practise what it preaches, the movement's electoral successes do testify to the virulence of anti‐party sentiment among many political activists.

Finally, Quebec nationalism, organized by the Parti Québécois provincially, and the Bloc Québécois in national politics, provides a direct challenge to the very survival of the country. Those parties' commitment to support a series of provincial referendums on the issue of Quebec independence, until they win, has kept this fundamental issue at the top of the national political agenda, but in a way that (p.368) isolates most Canadians, and hence the national parties, from the debate. In doing so it fosters a perception among many that the national government is preoccupied with Quebec and is comparatively indifferent to the interests of other regions. It is this sense that national parties, especially when in government, turn their backs on the west that contributed to the rise of Reform and the breakdown of the party system in 1993.

Political Recruitment

Canada's cadre parties exist to do little more than manage elections so it is not surprising that they monopolize the routes to elected office. The parties' privileged position as the gatekeepers of the country's electoral life leaves little room for independents in Canadian politics. But their nature means that they are as much vehicles for the politically ambitious as they are disciplined recruitment agencies. Virtually all political recruitment is done through the local constituency associations that jealously guard their traditional power of nominating their parliamentary candidate. These local associations often do not have any formal search committees, or enforceable rules governing their nomination contest, so that self‐selection is a dominant force in the process (Carty and Erickson 1991). Individuals, following their ambition, or the urgings of a few friends, can organize a personal campaign team, sign‐up sufficient new members and win a nomination meeting vote (Sayers 1999). So deeply entrenched is this practice that even sitting MPs are not immune from challenge at the constituency party nomination meeting that proceeds every election: in 1997 four sitting MPs lost their places to individuals who recruited themselves and captured the local party nomination.

In recent years party leaders and managers have attempted to exercise more control over the recruitment process—in order to run more women and minority candidates, to defend the party against single‐issue group penetration, to attract ‘stars’, and sometimes to defend sitting MPs. The Liberals have tried leader's appointments, the New Democrats quota systems, and Reform/Canadian Alliance elaborate screening mechanisms to do this, but none of them have been particularly successful at altering the basic pattern of recruitment. Local party associations see the right to choose their candidates as a matter of party democracy and strenuously resist interference from other parts of the party. On occasion, when the leadership has insisted, an entire local executive has resigned on the eve of an election leaving the party candidate to survive as best they might.

Local associations give the highest priority to recruiting individuals who are ‘committed to the constituency’, and rate qualities like ‘experienced party worker’, ‘political experience’ or ‘nationally well known’ at the bottom of a list of twelve qualities they look for in a candidate (Erickson 1997: 53). Primarily interested in winning, these local organizations will often recruit a candidate who has an attractive constituency profile with little consideration for his or her past party activity. In 1997, one New Democrat party association recruited a former Conservative (p.369) government minister, the Liberals had former New Democrat legislators and MPs as candidates, a Conservative association nominated a former Reform MP, and both Reform and the Bloc had former Liberals and Conservatives. All this gives evidence of a very catholic approach to recruitment and the absence of a disciplined or tightly managed career ladder by the parties.

The short political lives that the Canadian parliamentary process allows most politicians (Docherty 1997: 42–7) means that the parties may be forced to recruit individuals into positions of leadership from outside the caucus, sometimes from outside the party itself. All three of the leaders of the old parties (at the end of the century) were not in parliament when they won their leadership, and some ministers of the government were drafted into the party in order to join the government. The separation of the national and provincial party systems, and the almost complete absence of party politics at the municipal government level, contributes to the difficulties that parties have in dominating the routes to political office. Whether the parties are recruiting the politicians, or the politicians are using the parties, is an open question. Seen from the perspective of the system's gatekeepers the parties appear to dominate. But on the ground, where the politicians live, the cadre style of the parties leaves the politicians comparatively free to come and go through whichever gate suits them at the moment.

Canada, like many other nineteenth‐century democracies, was once governed by the politics of patronage. Most of that was done away with in the early decades of the twentieth century (though it demonstrated a remarkable persistence in several of the eastern provinces), leaving only limited traces in the governance of the country. The government still controls appointments to the Senate, as well as a wide range of public advisory boards and commissions, and enthusiastically appoints its supporters to them, especially where such individuals can do no ready harm. There is also some minor patronage (contracts, legal work, etc.) available to governing parties, but for the most part these appointments and commissions are designed to reward electoral and financial supporters, as much as to shape or control public policy and government programmes in any systematic fashion. Using these resources to recruit both people and money allows the parties to sustain their electoral vocations.

Interest Articulation and Aggregation

For much of the century Canada's cadre parties have specialized in finding ways to aggregate competing interests rather than articulate underlying social cleavages. This propelled them towards an accommodative style of politics dominated by the imperatives of regional brokerage. Then, as a new party system emerged after 1960, the parties adopted a new pan‐Canadian approach to governing which saw them work towards building support from individual Canadians for their conception of the national agenda (Smith 1985: 31). This politics had the three national political parties adopt essentially similar approaches to issues of representation (p.370) (Carty et al. 2000: ch. 5) while sharply increasing the distance between federal and provincial party organizations.

In their move to establish a pan‐Canadian politics the Liberal and Conservative parties did not question their long‐standing cadre‐organization instincts that led them to bridge cleavages and mask differences in the interest of building a broad electoral base. For their part the New Democrats traded their socialist principles for the electoral promise of populist appeals (Bradford and Jenson 1992). The consequence was that there was little difference among the parties: all adopted the same vision of a modern welfare state, all proclaimed and practised a politics of pluralism and openness in party life and public policy, all bought into a conception of Canada as bilingual and multicultural, and all sought to accommodate the constitutional demands of Quebec. Each of the national parties accepted that their ambitions for national policy initiatives clashed with the constitutional realities of the federal division of powers and so committed themselves to government by executive‐federalism. But that system made it impossible for voters to hold any government, any party, electorally responsible. Inevitably this eviscerated much of the point of the parties' electoral competition and three decades of seemingly endless debate and constitutional wrangling over Quebec–Canada relations rubbed an increasingly frustrated electorate raw. The result was the collapse of the party system in the 1993 election.

Both the Bloc Québécois and the Reform parties promised a new kind of politics, a new approach to how interests ought to be promoted, and in doing so represent a fundamental attack on the very essence of the pan‐Canadian party system that had existed since 1960. Their ‘anti‐system’ messages are different. The Bloc, articulating Québécois nationalist aspirations of independence, seeks an end to the politics of accommodation between English and French speakers by breaking the country apart. But it is a second division team with its political direction being driven by the Parti Québécois and the politics of the party system in Quebec. How long the Bloc will remain in national politics, or what course it will follow, are more likely to be dictated by events in Quebec than the wider Canadian party system. The party's success at capturing the majority of Quebec seats in parliament has increased the regionalization and fragmentation of the national party system. It has meant that for the first time in the country's history a national Liberal government had to be formed without substantial support from Quebec. It has fundamentally altered and stretched the national political agenda by putting Quebec independence on the ballot in one‐quarter of the country.

Reform's populist politics rejected the politics of pan‐Canadian accommodation and executive‐federalism in favour of simple majoritarianism practised by smaller and more decentralized government. It advocated the end of party discipline in the interest of making parliamentarians responsive to their local electorates, and it rejected a politics of interest‐group pluralism and the recognition of special rights in the name of individualism and equality. It was in this, more than in its determinedly right‐wing economic policies, that the party challenged the existing (p.371) national political parties and broke the mould that had governed them for over three decades. Carrying this anti‐system message while playing in the system has not been easy. Does the party have to accommodate to the pressures of the system to increase its chances of winning? But if it does so will it have lost its soul, and its very reason for being? Can it be the ultimate cadre party while reforming the very nature and scope of Canadian governance? The decision to reinvent itself as the Canadian Alliance reflects its belief that it can. The historical record suggests it cannot, but it also suggests that with its strong regional base the Canadian Alliance can persist as a major actor in the party system.

The eruption of these two new parties in 1993 followed a 1992 national referendum on constitutional change that had repudiated the country's political class (Johnston et al. 1996: 277). That referendum asked Canadians to accept a comprehensive package of constitutional amendments and it was supported by all the national parties, most provincial party leaders and leading spokespersons for business, labour, and aboriginal interests. For perhaps the first time in Canadian history since Confederation, the country's political elites were united and their defeat at the hands of the people, led by Quebec nationalists and western Reformers, signalled the growing gap between Canadian elites and masses. It is this gap that the new parties have continued to exploit. They reject the consensus that the elites of the old parties had fashioned on the constitution and bilingualism, they question the legitimacy of the rules of the game governing party and election financing, and they argue for new representational norms. Reform/Alliance and Bloc success in forcing social and economic divisions into the party system shrank the differences between the political agendas of the party elites and the mass public, but only by increasing the diversity among the party elites themselves and so expanding the range of choice offered by the party system.

Not all Canadians find cadre parties particularly satisfying political instruments for promoting their policy objectives and there has been an increase in the public and electoral activity of single‐issue groups. These groups compete with the parties for resources, human and financial, and increasingly attempt to push their issues onto the agenda. The focus and discipline with which they articulate their demands makes it increasingly difficult for the parties to engage in the traditional brokerage that promises something, but rarely everything, to everyone. These groups have invaded the electoral process in at least three ways over the past decade. First, single‐issue groups sometimes mobilize their members to swamp a local party's membership in order to influence the nomination of a candidate (or selection of delegates to a party convention). The parties have fought back and their leaders now veto local nominations engineered by obviously single‐issue groups. Second, these groups increasingly participate in local constituency campaigns in an attempt to heighten the salience of their issue and to support or oppose particular candidates. In the 1988 general election over half the candidates of the major parties reported that some interest group had intervened in their local campaign in a deliberate attempt to influence their election or defeat (Carty 1991: 186–8). The (p.372) third approach taken by interest groups has been to launch large‐scale mass media campaigns during the election in an attempt to shape the national campaign agenda.

Regulating single‐issue group electoral activity was not a concern when the electoral expense regime was put in place in the mid‐1970s, but the recent growth in their activity now worries the political parties who see their control of the electoral process in jeopardy. Twice in the last decade they have moved to limit interest groups' activity and influence by imposing severe restraints on their spending during campaigns. Twice that legislation has been found unconstitutional. The commitment of the government to legislate for a third time in the dying months of the century indicates that the old parties are determined to try and defend their traditional electoral ground and to reassert their monopoly over the channels of electoral participation. The sheer scale of interest group life, and the courts that give them a hearing and a place in the system, are likely to thwart that ambition.

Participatory Organizations?

Active participation in the life of Canadian parties is, for virtually all individual members, structured by and centred in local constituency associations. National parties believe they must maintain associations in every constituency and the reality is that many of these are little more than ‘paper branches’, organizations with no money, few members and are consequently inactive (Carty 1991: 98–102). Active local party organizations providing an opportunity for political activity are most likely to be found in areas of electoral strength, especially those where the party has a sitting incumbent. Given the high levels of electoral volatility, areas with an active membership can quickly be turned into participatory deserts.

Party members engage in a wide range of local activities, not all focused on national politics, but the principal tasks are those of organizational maintenance—fund‐raising, social events, and annual membership drives—with a concern to be ready for the next election. Large numbers of local associations report that they have their members engaged in policy study of one kind or another, but then admit that there is very little competition among their members to be delegates to their party's national conventions where policy will be debated (Carty 1991: ch. 3). Most experienced party members realize that despite their work, opinions, or convention resolutions, party policy remains the preserve of the professional politicians and especially that of the leadership. As a result, most local associations, whatever their nominal membership, are dominated by a small core of committed activists who keep the local machine on stand‐by or work to support the personal ambitions of a local MP. The high level of turnover on local executives suggests that these positions offer little in the way of influence or power in the party. Reform entered the system promising to break this pattern by giving local members a greater role in the party's policy‐making and its parliamentary direction. While its local associations were very active in its early years, the leadership continues to manage its membership as easily as in the traditional parties (Flanagan 1995).

(p.373) In the absence of any systematic membership data it is not possible to draw sharp profiles of just who these party members are. Studies of delegates to party leadership conventions all report that they are disproportionately middle class and well educated. They are also often young and new to the party. To the extent that these delegates reflect the membership, it is clear that a party's members are rarely typical of its electorate and that many of them do not have a record of long service or involvement in party activity.

The only good reason to join a political party in Canada is to participate in nominating a local candidate or in choosing the party leader. Membership conveys a vote and the opportunity to participate in making a decision of real, immediate, and obvious consequence. For most other partisan activity, including working on campaign teams that are appointed by the candidate, formal party membership is irrelevant. Only when there are internal party contests do Canadians join local party associations in substantial numbers, creating the membership peaks and troughs that mirror electoral cycles, their membership fees constituting little more than a party poll tax. Even these membership swells distort the extent to which there is a participatory membership, for turnout rates in party contests for candidates or leaders is normally well below 50 per cent (Carty and Erickson 1991: 125).

By the standards of a cadre party these are participatory organizations. Individual party members have a direct vote that allows them to choose, or replace, their local notable at election time. For much of this century party members have also had an indirect vote (via a delegate convention) in the selection and then removal of party leaders. By the 1990s growing dissatisfaction with the leadership convention process—its escalating costs, candidate manipulation of majoritarian delegate‐selection processes, and multiple‐voting provisions—led the parties (first at the provincial, then the national level) to adopt party leadership‐selection processes that give every member a direct vote (Carty and Blake 1999). While a consensus has yet to emerge among the parties as to how best to organize such contests, the principle seems now firmly entrenched for all but the New Democrats who are unclear as to how to incorporate their affiliated (local trade union) members.

In 1998 the Conservatives were the first nationwide party to use a direct membership vote to select a new party leader. In fact, the vote was constrained by the party's decision to weight votes so that each constituency's membership, irrespective of its size, had an equal influence on the outcome.7 The contest led to a large increase in the party's national membership—from 18,000 to 90,000—in a couple of months, but only half of the members actually voted in the first ballot, 40 per cent in the second. The party had recruited just 3.7 per cent of its 1997 voters in anticipation of the contest and then only 1.9 per cent of them participated. True to form the membership soon collapsed and by 1999 was back down to 18,000 (p.374) (Stewart and Carty 2001). The Canadian Alliance followed suit in 2000 when it too chose its leader by a vote of a newly, and temporarily, inflated membership. With the country's next prime minister likely to emerge out of some sort of membership vote, the parties appear to have lost control over one of the most critical internal decisions they make, leaving parliamentary parties to adjust to a membership and leadership far removed from their world.

Despite the Conservatives' modest success, some of the provincial parties have stimulated greater participation in direct membership votes for their leadership, particularly when in office. In Alberta the governing provincial Conservative Party even allowed new ‘members’ to join between the two ballots, effectively eviscerating the notion of party membership and turning the contest into an American‐style primary election. But this is only to recognize and extend the logic of cadre politics by making the party a loose framework within which notables conduct their leadership struggles among the electorate.

The nomination and leadership processes of the parties have remained unregulated by law and the parties have demonstrated only the most limited capacity to constrain the excesses of ambitious politicians. As it becomes clear that these are no longer private internal party matters to be settled by a disciplined membership, but are public contests open to any interested individual,8 calls for their regulation by the state will grow. To date the parties have resisted pressures to do so, but in this the parties may yet succumb to another round of the American contagion Siegfried observed a century ago.

Political Communication and Education

Walter Bagehot, in his study of the English constitution published in the year of Canada's creation, argued that the education of the nation was one of the primary tasks of parliament and the party politicians who inhabit it. Since the death of the partisan press in the early decades of the twentieth century it has not been a high priority for Canadian parties. Beyond the efforts of individual campaigns, they have not sustained any regular communication programmes and few of them have managed even to maintain regular party newsletters for extended periods of time. When the 1991 Royal Commission on Electoral Reform suggested the creation of party foundations to do policy research and education work, the recommendation merely seemed to puzzle the few party politicians who read it. Party communication remains electoral campaigning. To try and do more would restrict the politicians' freedom to run on the issues of the moment, or to reverse direction to catch a prevailing electoral wind.

(p.375) This narrow conception of their responsibility to political education has supported the parties in a preoccupation with regulating election campaign rules in their favour. To ensure that they have the ability to get their campaign messages to the electorate the parties have used the Election Act to require that multiple dwelling units provide them access, that the press not charge them higher than average advertising rates, and that broadcasters provide free‐time and guarantee the purchase of additional paid‐time. Coupled with their attempts to limit the campaign period advertising of single‐interest groups, these are all designed to strengthen their hold on election agenda‐setting. Once the campaign is over, Canadian party organizations retreat to their quiescent stage, adopting a largely passive and reactive response to the mass media and the shifting tides of public opinion.

Canadian parties define their biggest communication challenges as those of operating in at least two languages, though local election campaigns can often involve several more. The parties have traditionally dealt with this by having two distinct communication and campaign teams responsible for French‐ and English‐ speaking Canada. Though these teams are appointed by the leader, and report to the centre, there is often very little coordination between their activities. Preoccupied with helping politicians get elected from distinct electorates, they are not concerned with ensuring the party is speaking with one voice.

Plus Ça Change, Plus C'est La Même Chose?

The great French political geographer, André Siegfried, visited Canada at the beginning of the twentieth century and discovered a party system whose elections were characterized by unmatched ‘fury and enthusiasm’ (Siegfried 1906: 117). He reported that four issues dominated electoral contests: ‘general prosperity’, which all the parties supported; ‘local advantage’, promoted by each candidate seeking to be their constituency's representative; the ‘race question’, by which he meant the tension between French‐speaking Catholics and English‐speaking Protestants; and ‘leadership’. With all for prosperity and local patronage, and none for inflaming the social division, electoral politics more often than not came down to questions of personal leadership. Parties were reduced to those gathered around leaders who were responsible for creating whatever national organization existed, and ‘whose mere name is a programme in itself’ (Siegfried 1906: 136).

If Siegfried were able to return at the beginning of the twenty‐first century he might be surprised at how little elections had changed. Parties all promise to have the formula for prosperity, constituency candidates still run as local champions, the relationship between the two large ethnic communities still bedevils the country's politics, and elections revolve around the claims of party leaders. But more striking is how little the parties have changed. They remain cadre‐style organizations consisting of small numbers of local activists and politicians linked together at the (p.376) centre by a party leader. These leaders now command parties with all the trappings of modern ‘electoral‐professional’ political machines—specialized media and campaign organizers, pollsters, computer‐driven fund‐raising and communication technologies—to help them fight elections, but their relationship to the organization in the constituencies and its local notables has hardly changed.

The Social Democrats tried to build a party organization on the model of a mass party but with limited success. While they never enrolled large enough numbers to create a true mass base, their real obstacle was the enormous social and economic heterogeneity of the country that made cadre forms the most efficient structure for linking the imperatives of local diversity to the impulses of national government. Their most successful organizations have been the local associations built by successful local MPs who epitomize the cadre notable. While party finance reforms have made the parties' electoral resources more dependent upon the state than in the past, there is little evidence that this has led to a cartelization of Canadian party organization (Young 1998).

The convulsions that began to restructure the Canadian party system in the early 1990s led to two new parties playing a significant role in parliament, to the fragmentation and regionalization of party support for all of the parties—the new as well as the old — and to an increased diversity in the parties' style and approach to articulating and aggregating interests. Within the parties, new patterns of leadership politics, giving all members a vote, began to emerge. As in past periods of party system change, the basic underlying electoral alignments are being reshaped producing new constellations of competition and different sets of electoral choices for individual electors. Yet, in none of this party system change has there been a fundamental challenge to the basic cadre form of party organization that has persisted since the nineteenth century.

Siegfried's characterization of successful Canadian parties as ‘machines for winning elections’, preoccupied with ‘the pure and simple continuation of their own existence’, seems as relevant at the beginning of the twenty‐first century as it did at the beginning of the twentieth. Given shape by the imperatives of the plurality electoral system and a majoritarian parliamentary politics, but operating in a new society, Canadian parties have fallen uneasily between their more structured European cousins and their idiosyncratic American neighbours. They have resisted any organizational contagion from the left and the developmental paths of European parties (Katz and Mair 1995). If there is to be any contagion in the new millennium then Siegfried was prescient in pointing to the south. In the last century America's cadre parties have been transformed into a set of regulated ‘public utilities’ to structure democratic competition (Epstein 1986: ch. 6). With the emergence of primary‐like contests driving their internal politics, Canada's parties seem poised to follow in that direction.

While Canada's cadre parties persist, even if reorganized and reshaped, it is clear that the party system, defined by those same parties, has not. At the beginning of the new millennium it is being recreated (for the second time in the last half‐century) to (p.377) serve a changing political system as the patterns of Canadian governance are being reordered. The parties' primary tasks—their manifest functions—of nominating candidates and conducting elections persist; what are again changing are the latent functions they perform for the wider polity. This is not the decline of party, for the roles of the political parties are not so much in decline as they are being transformed. They must be judged in terms of the demands of a new and different system rather than compared to their practices in an old one. The genius of Canadian cadre parties has been in their ability to survive and adapt to successive political systems that have made radically different demands upon them. That is why the country is entering the twenty‐first century with refurbished nineteenth‐century parties.

References

Bibliography references:

Bakvis, H. (1991) Regional Ministers: Power and Influence in the Canadian Cabinet (Toronto: University of Toronto Press).

Blais, A. and Gidengil, E. (1991) Making Representative Democracy Work: The Views of Canadians (Toronto: Dundurn Press).

—— —— Nadeau, R., and Nevitte, N. (1999) ‘Measuring party identification: Canada, Britain and the United States’. A paper presented to the annual meeting of the American Political Science Association, Atlanta.

Blake, D. E. (1985) Two Political Worlds: Parties and Voting in British Columbia (Vancouver: UBC Press).

Bradford, N. and Jenson, J. (1992) ‘Facing economic restructuring and constitutional renewal: social democracy adrift in Canada’, in Frances Fox Piven (ed.) Labour Parties in Postindustrial Societies (New York: Oxford University Press).

Carty, R. K. (1991) Canadian Political Parties in the Constituencies (Toronto: Dundurn Press).

—— (1997) ‘For the third asking: is there a future for National Political Parties in Canada?’ in Tom Kent (ed.) In Pursuit of the Public Good (Montreal: McGill‐Queen's University Press).

—— and Blake, D. E. (1999) ‘The adoption of membership votes for choosing party leaders: the experience of Canadian parties’. Party Politics 4.

—— and Erickson, L. (1991) ‘Candidate nomination in Canada's National Political Parties’, in Herman Bakvis (ed.) Canadian Political Parties: Leaders, Candidates and Organization (Toronto: Dundurn Press).

—— and Stewart, D. K. (1996) ‘Parties and party systems’, in C. Dunn (ed.) Provinces: Canadian Provincial Politics (Peterborough: Broadview Press).

—— —— (2001) ‘Leadership politics as party building: the 1998 conservative leadership contest’, in W. Cross (ed.) Political Parties, Representation and Electoral Democracy in Canada (Toronto: Oxford University Press).

—— Cross, W., and Young, L. (2000) Rebuilding Canadian Party Politics (Vancouver: UBC Press).

Clark, H. and Kornberg, A. (1993) ‘Evaluations and Evolution: Public Attitudes Towards Federal Political Parties, 1965–1991’. Canadian Journal of Political Science XXVI, 287–312.

(p.378) Docherty, D. C. (1997) Mr. Smith Goes to Ottawa: Life in the House of Commons (Vancouver: UBC Press).

Epstein, L. (1986) Political Parties in the American Mold (Madison: University of Wisconsin Press).

Erickson, L. (1997) ‘Canada’, in Pippa Norris (ed.) Passages to Power: Legislative Recruitment in Advanced Democracies (Cambridge: Cambridge University Press).

Flanagan, T. (1995) Waiting for the Wave: The Reform Party and Preston Manning (Toronto: Stoddart).

Johnston, R., Blais, A., Gidengil, E., and Nevitte, N. (1996) The Challenge of Direct Democracy: The 1992 Canadian Referendum (Montreal: McGill‐Queen's University Press).

Katz, R. S. and Mair, P. (1995) ‘Changing models of party organization and party democracy: the emergence of the cartel party’. Party Politics 1: 5–28.

LeDuc, L., Clark, H., Jenson, J., and Pammett, J. (1984) ‘Partisan instability in Canada: evidence from a new panel study’. American Political Science Review, 78, 470–84.

Mair, P. (1997) Party System Change: Approaches and Interpretations (Oxford: Oxford University Press).

Massicotte, L. (1998) ‘Party cohesion in the Canadian House of Commons 1867–1945’. A paper presented to the annual meeting of the Canadian Political Science Association.

Merton, R. (1949) Social Theory and Social Structure (Glencoe, IL: The Free Press).

Perlin, G. (1980) The Tory Syndrome: Leadership Politics in the Progressive Conservative Party (Montreal: McGill‐Queen's University Press).

Savoie, D. J. (1999) Governing from the Centre: The Concentration of Power in Canadian Politics (Toronto: University of Toronto Press).

Sayers, A. M. (1999) Parties, Candidates and Constituency Campaigns in Canadian Elections (Vancouver: UBC Press).

Siegfried, A. (1906) The Race Question in Canada (Toronto: McClelland and Stewart), reprinted, 1966.

Simeon, R. (1972) Federal‐Provincial Diplomacy: The Making of Recent Policy in Canada (Toronto: University of Toronto Press).

Smith, D. (1985) ‘Party government, representation and national integration in Canada’, in P. Aucoin (ed.) Party Government and Regional Representation in Canada (Toronto: University of Toronto Press).

Stanbury, W. T. (1991) Money in Politics: Financing Federal Parties and Candidates in Canada (Toronto: Dundurn Press).

Wearing, J. (1998) ‘Guns, gays and gadflies: party dissent in the House of Commons under Mulroney and Chrétien’. A paper presented to the Annual Meeting of the Canadian Political Science Association.

Whitaker, R. (1977) The Government Party: Organizing and Financing the Liberal Party of Canada, 1930–1958 (Toronto: University of Toronto Press).

Woolstencroft, P. (1997) ‘On the ropes again? The campaign of the progressive conservative party in the 1997 federal election’, in Alan Frizzell and Jon Pammett (eds) The Canadian General Election of 1997 (Toronto: Dundurn Press).

Young, L. (1998) ‘Party, state and political competition in Canada: the Cartel Argument revisited’. Canadian Journal of Political Science XXXI, 339–58.

Notes:

(1) On manifest and latent functions see Merton (1949) and his discussion of party machines.

(2) There is a considerable literature on party identification and its measurement. For a recent review see Blais et al. 1999. They show that party identifications are lower in Canada than either the USA or Britain and also note that in Canada, unlike the other two countries, ‘the relationship between party identification and voting for the party one identifies with is not monotonic’ (Blais et al. 1999: 10).

(3) One consequence of this is that there are no direct measures of annual party membership size. The figures reported in Table 12.5 are based on surveys of local associations reported in Carty 1991 and Carty et al. 2000.

(4) One indication of the comparative insignificance of party staff in Canada is the absence of any study of the parties' national offices and the lack of any data on national party staffing patterns over time.

(5) The impact of inflation over the period covered in the tables was considerable and this must be taken into account in assessing any trends over time. NDP finances are particularly difficult to decipher as they include provincial level accounts. On that see Stanbury 1991.

(6) In practice the parties and their local candidates, speak to Canadians in the many different languages that they find in the electorate.

(7) Each constituency was given 100 points to be distributed in proportion to the share won by the candidates in the vote. In this the Conservatives were attempting to mirror the logic of a national election contest in which it is the constituency that is the unit of importance.

(8) Most of the parties have traditionally opened their membership (and hence party contests) to individuals who were neither of legal voting age nor Canadian citizens. Reform set a new standard when it insisted that only voting age citizens be entitled to vote in its affairs.