PDF
Geographic Representation and Electoral Reform
Hilary Pearse

The close relationship between parliamentarians and residents of the geographic district they represent is an essential element of the electoral system in Canada, the United Kingdom, Australia and New Zealand. Debates on electoral reform in these countries have typically accepted the importance of the relationship without question. This essay looks at  the basis for the continuing attachment to geographic representation. It concludes that there is evidence to support the importance of the role played by directly elected constituency representatives but suggests that the attachment to geography comes at a cost, by restricting electoral reformers in their choice of alternative options and constraining parties and representatives from exploring the full potential of new electoral systems. 

Canada, the United Kingdom, Australia and New Zealand share a tradition of single-member districts with plurality or majority electoral formulas. Due to their common history as British colonies, the modern political systems of Canada, Australia and New Zealand are all derived from the Westminster model of parliamentary democracy. However, there is some variation between the systems. The first-past-the-post plurality electoral system is currently used in the United Kingdom and Canada, and was used in New Zealand from 1946-1993. Under this formula, the candidate in a single-member constituency need only win a plurality of the votes to be elected. In contrast, the majoritarian Alternative Vote system used to elect members of the Australian House of Representatives requires candidates to win an absolute majority of the vote. Since 1996, just more than half of the members of the New Zealand parliament have been elected using the traditional plurality system in single-member districts, while the remainder are indirectly elected via party lists in proportion to the nation-wide support for their party. All four countries currently elect one candidate per geographic constituency, although various forms of multi-member districts have been used in the past in the Canadian provinces, and two- and three- member districts for national elections existed in the United Kingdom until 1950. 

Vernon Bogdanor describes the plurality system as it developed in Britain and the British colonies as being “profoundly linked to the notion of territorial representation.”¹  As representatives of constituencies, MPs were “attorneys seeking the redress of grievances before committing their constituencies to the payment of the expenses of government.” The concept of parliament as an assembly for the representation of constituency interests was later eclipsed by Burke’s notion of the parliament as a “deliberative assembly of one nation, with one interest, that of the whole; where, not local purposes, not local prejudices ought to guide, but the general good, resulting from the general reason of the whole.”²  Rather than acting as a ‘delegate’ for the will of the constituency, the representative is a ‘trustee’, elected by their constituents in recognition of their wisdom, to exercise their judgment as they see best. The rise of modern cohesive political parties has further challenged the concept of representation. A description of representation in Australia is equally applicable to the other three countries studied in this paper: “Within major parties, the popular images of elected representatives are neither as trustees nor delegates of their voters, but as partisans.”³ 

Even in an era of party dominance, the perception of the representative as a delegate for their constituency persists. 

Eulau and Karp define the modern understanding of representation as comprising four possible ‘components of responsiveness.’  In addition to ‘policy responsiveness’ and ‘symbolic responsiveness’, they include ‘service responsiveness’ – “the efforts of the representative to secure particular benefits for individuals or groups in his constituency” – and ‘allocation responsiveness’ – “the representative’s effort to obtain benefits for individuals or groups in his constituency through pork-barrel exchanges in the appropriations process or through administrative interventions.” ⁴  Similarly, Searing describes MPs in the United Kingdom performing two tasks in their constituency service role, acting as ‘welfare officers’ for individual constituents or as a ‘local promoters’ for the collective interests of their constituency. ⁵  Representatives in Canada, the United Kingdom Australia and New Zealand now identify with three broad representational roles, defined by Studlar and McAllister as “locals, who focus on articulating local concerns and interests; partisans, who see their role in party political terms; and legislators, who emphasise the parliamentary role of an elected representative.” ⁶ 

Constituency Casework 

The constituency casework performed by representatives, assisting their constituents with specific problems and interceding on their behalf when necessary, is regarded as an essential part of the job of an MP in Canada, the United Kingdom, Australia and New Zealand. Fenno defines the core activity of constituent service as “providing help to individuals, groups and localities in coping with the federal government.” ⁷  This role is regarded as the most important, and also the most enjoyable aspect of their job by MPs in all four countries. As one Canadian MP has commented, “once you realise that [constituents] have no place else to turn to, then after you help them, you understand it is the most vital service we provide.” ⁸ 

Some scholars have questioned the effectiveness of MPs’ casework role, given the small proportion of their constituency that uses their services, and the availability of other agencies for assistance. 

Only a small minority seek assistance in the first place, and at least some go away frustrated. The numbers feeling they have been helped as a proportion of the entire constituency inevitably turn out to be tiny: 6 per cent according to the Kilbrandon Commission, 5 per cent in the Granada study, 7 per cent in the Attitudes to Government survey. Thus the overall impact and reputation of the MP as a source of assistance looks modest in comparison with other agencies, in particular those with deeper local roots, such as councillors, neighbourhood groups and associations, and even the family doctor.  The MP does not loom large in the electorate’s mind as a source of help; he is an institution of last resort.⁹ 

In defence of the casework role played by representatives in the US and the UK, others have argued that, regardless of whether constituents actually use the casework service provided by their representative, having the option to do so, should the need arise, is important to citizens. “The sheer number of constituents who have personally received assistance from their representative may not be the crucial consideration. Because of differences in their sociodemographic composition, some constituencies may have a much greater basic demand for assistance than others, but constituents without current need might still believe that, should the need arise, their representative would be there to help.” ¹⁰  According to Cain, Ferejohn and Fiorina, constituents’ appreciation for this service is demonstrated by the additional personal vote that rewards dedicated representatives. In the United Kingdom, “variations in constituency work apparently account for swings of something between 1.5 and 2 per cent for Conservatives and between 3 and 3.5 per cent for Labour.” 

The electoral benefits of constituency service are contested in Australia. Studlar and McAllister demonstrate that local constituency work reduces an MP’s vote at a rate of 0.09% per hour for each extra hour of work per month, other things being equal. “For example, an MP who said that he or she worked 40 hours per month dealing with constituents’ problems and attending local functions could expect to be short 1.8% of the vote compared with an MP who devoted 20 hours per month, net of other things.” They attribute the negative relationship between constituency work and vote to the displacement of other more electorally beneficial activities, such as working for the national party in the capital, or developing a national media profile, that occurs when MPs are devoting their time to casework. 

It appears that the casework role played by representatives may be neither as effective, nor as valued by voters as has been traditionally assumed. 

The existence of alternative agencies for the redress of constituents’ grievances such as community centres, welfare rights groups and local law centres or Ombudsmen reduces the MP’s casework role to an avenue of last resort.  In 2004, the Law Commission of Canada concluded that “the limited research on geographic representation suggests that the link between constituents and their elected representatives may not be as important as we initially thought.” 

Direct Election and Accountability 

The direct election of a representative enables the constituency to hold them accountable for their actions during the parliamentary term.  If constituents consider the performance of their incumbent MP to have been inadequate, or believe that they would be better represented by a challenger, they have the electoral power to replace their representative. “The indispensable condition for an MP’s ultimate accountability to his electors and for close ties between citizen, locality and Parliament is the single-member constituency.”¹¹  For electors accustomed to directly electing their representatives in geographic districts, the indirect election of representatives via a party list in closed-list proportional representation (PR) systems, appears to undermine the capacity of voters to hold their representatives to account. During the campaign preceding the 1993 referendum on electoral reform in New Zealand, the pro-FPP lobby group, the Campaign for Better Government, exploited this fear in order to weaken support for MMP. Television advertisements claimed that “party list members would be unaccountable to public opinion, being in effect appointed by their respective parties.” ¹²  While campaign advertisement portrayals of list MPs as faceless, suited party apparatchiks have not been borne out, it has become standard practice for senior members of the party caucus to be placed high on the party list as a ‘back-up’ in the event that they lose their electorate seats, effectively denying voters the capacity to directly prevent the re-election of particular representatives.  In this situation, an MP’s re-election becomes dependent on the success of their party in the separate ‘party vote’. ¹³ 

However, the accountability of representatives in plurality single-member districts has also been questioned. In the UK, Crewe has argued that “the large majority of seats are impregnable: consecutive landslides of 1945 proportions for Labour and 1983 proportions for the Conservatives would still leave 70 per cent of seats in the same party hands. The safe seat increases the Member’s incentive to neglect his constituency and deprives constituents of an effective electoral sanction if he does.” In a party-dominated political system, ‘safe seats’ will nearly always be won by the same party, regardless of the performance of the individual representative. In addition, pre-selection of candidates in safe seats may depend more on the party selection committees’ perception of candidates’ loyalty to the party than their dedication to the constituency. Bogdanor argues that “the MP’s career depends more upon his party than upon his constituency. An MP can often survive unpopularity in his constituency provided that relations with his local party remain good; but, if he loses the support of his local party, his political career will usually be at an end.” 

In Canada, the United Kingdom, Australia, and pre-1996 New Zealand, the nature of the electoral system requires voters to express both their constituency candidate preference and their party government preference in the one vote. “Voters cannot separate their choice of member from their choice of government.” Where these preferences differ – the voter’s preferred candidate is not a representative of the preferred party – the voter’s desire to vote for their government of choice may prevent them from holding their representative to account. In Westminster-derived democracies, where party identification is the “pre-eminent attitudinal influence on voting behaviour”, the probability that voters will use their vote to reprimand an unpopular MP or reward a popular one is diminished. 

Party Dominance and the Personal Vote 

The influence of party identification, attitudes to party leaders and attitudes to constituency candidates on voting behaviour has been measured in all four countries. Although it is accepted that party identification dominates voters’ electoral decisions, scholars argue over the extent to which a ‘personal vote’ for candidates exists independent of party preference. In one of the seminal texts on the topic, The Personal Vote, Cain, Ferejohn and Fiorina define the personal vote as: 

…that portion of a candidate’s electoral support which originates in his or her personal qualities, qualifications, activities, and record. The part of the vote that is not personal includes support for the candidate based on his or her partisan affiliation, fixed voter characteristics such as class, religion and ethnicity, reactions to national conditions such as the state of the economy, and performance evaluations centred on the head of the governing party.¹⁴ 

The influence of a candidate’s personal qualities and activities on mass voting behaviour in parliamentary democracies has traditionally been considered to be small. In 1955, Milne and MacKenzie described the belief “that the personal qualities of a candidate are of little importance in winning votes”, as so widespread that it was “no longer a paradox but a platitude.”¹⁵  Similarly, Cross argues that Canadian MPs lack a personal mandate. “All studies of Canadian political behaviour tell us that the vast majority of voters use their single ballot to express their preference for a governing party (and preferred Prime Minister) – even though the only names appearing on the ballot are those of the candidates for the their local riding.” ¹⁶ 

However, a number of scholars have demonstrated that the personal vote does exist in parliamentary democracies, and that while it may be small, it can play a significant role in marginal constituencies.  In their comparison of the personal vote in the US and the UK, Cain, Ferejohn and Fiorina conclude that although “the electoral advantage which accrues to a hard-working congressman is far greater than that occurring to a similarly hard-working MP”, “the efforts of the MP do have a discernible effect … which may be growing in importance.” In Australia, Bean has demonstrated that the personal vote is a “real, if modest, component in the mix of factors that combine to determine party choice in Australian federal elections in the lower house”, with an impact of between 2 and 3 per cent of the vote. Similarly, Ferejohn and Gains have found “some evidence of the development of the personal vote in Canada”, where the opportunity for incumbent MPs to “develop a favourable personal reputation in their constituencies” can affect their electoral success.  Docherty estimates that the effects of incumbency average “between 3% and 5%” in Canada, which may provide the winning margin in a close race. 

Electoral reform has provided a rare opportunity to examine the extent to which the support for electorate MPs is independent of support for their party. In an MMP system, each voter casts two votes: one for their preferred candidate and one for their preferred party. Three elections in New Zealand using this system have established a pattern where the two major parties, Labour and National, win most of the electorate seats, while the minor parties win most of their seats via the party vote. In 1996, 37 per cent of voters split their vote, dropping to 35 per cent in 1999 and rising to 39 per cent in 2002. In practice, this means that the majority of Labour and National electorate MPs win more votes than their parties in each electorate. Particularly popular MPs may win twice as many votes as their parties. 

In most Western democracies, with the notable exception of the United States, representative government is party government. 

However, the high proportion of split-ticket voting is by no means all about personal voting. A large proportion of split votes are due to tactical voting where voters recognise that a minor party candidate is unlikely to win in their local district so split their votes between a minor party and a major party candidate. 

Despite evidence to support the existence of a personal vote in Canada, the United Kingdom, Australia and New Zealand, it is important to be reminded of the dominant influence of party identification plays on voting behaviour. For example, in Bean’s test of variables influencing voting in the 1987 Australian Federal Election, the influence of party identification on voting was over six times stronger than the effect of attitudes towards local members. 

Implications for Electoral Reform 

Over the past decade, the electoral reform debate has entered the mainstream political agenda in Canada, the United Kingdom and New Zealand.  In Canada, five provinces and one territory are currently considering electoral reform. The federal government has also pledged to investigate reform options at the national level. In the United Kingdom, the Independent Commission on the Voting System recommended electoral system reform in 1998.  In New Zealand, a majority of voters in the 1993 referendum on electoral reform supported changing from the existing FPP plurality system to the more proportional MMP system.  Australia has been the exception to this trend. Arguably, the lack of salience of this issue in Australia may be due to introduction of a PR electoral system for the Australian Senate in 1948, which has facilitated the representation of minor parties in the upper house and tempered the major party dominance of the political system that has contributed to calls for electoral reform in the other three countries. 

The desire to maintain the close relationship between representatives and their constituencies has influenced the electoral system options that have been considered. In British Columbia, the Citizens’ Assembly on Electoral Reform selected local representation as one of its three key criteria for assessing potential reform options for the province. The Assembly’s interim report states that: 

Our tradition has long valued a system of representation that provides for local representation – for its politicians to speak for and answer to distinctive communities that make up the whole province. Citizens believe it is important that the interests of their particular communities be represented in public debate and policy-making. This is accomplished when MLAs have an intimate knowledge of the communities they represent and the concerns of the people in them. ¹⁷ 

In particular, members of the Assembly from large, rural ridings were wary of any electoral system that would increase the size of their electoral districts. The Assembly’s decision to recommend a Single Transferable Vote (STV) system, rather than a Mixed Member Proportional (MMP) system was partly based on the fact that an STV system, despite having multi-member districts, would maintain the same ratio of representatives to constituents as the current system. In contrast, under MMP, district MLAs would each represent a larger number of constituents in significantly expanded ridings. Assembly members hope that MLAs in multi-member districts would continue to act as local representatives. Since the adoption of STV for national elections in Ireland in 1922, a majority of Irish voters have twice voted to retain the electoral system in referenda held in 1959 and 1968. Members of the Dáil are renowned for their emphasis on local constituency issues, although scholars remain divided over the extent to which this is due to the electoral system or the localism of Irish political culture. 

The need to maintain local representation also influenced the decision of the Prince Edward Island Reform Commission in 2003. Like British Columbia, the Commission narrowed the final choice for a reform option to the STV and MMP systems. However, the small geographic size of the province facilitated the Commission’s final recommendation in favour of MMP. The Carruthers Report identifies the desire of Prince Edward Islanders to maintain a single, local representative as a key factor influencing this decision.  “The one thing that did come across loud and clear at the public meetings was the request to keep District Members of the Legislative Assembly even though their number may be reduced. Islanders want to be able to identify with their District MLA and the Commission respects this fact.” ¹⁸ 

Recommendations to introduce MMP in both Quebec and New Brunswick not only strive to maintain geographic representation by retaining a smaller number of single member constituencies but also by advocating the use of regional party lists. Rather than allocating compensatory list seats on the basis of a party’s support throughout the entire province, distributing list seats on the basis of the party’s strength in smaller regional districts sacrifices an element of the electoral system’s proportionality in order to ensure that all representatives, whether elected directly in a constituency or via the party list, serve a geographically defined region. Massicotte acknowledges that province-wide list seat allocation reduces the level of distortion involved in translating votes to seats but argues that that “this solution will seem undesirable to most Quebec citizens. Deep down, they are accustomed to linking an MNA [Member of the National Assembly] to an identifiable territory.” ¹⁹  Similarly, the New Brunswick Commission on Legislative Democracy envisages that regional list MLAs will also act as geographic representatives. List MLAs would undertake constituency service, and provide voters who did not support their constituency MLA with an alternative local representative. The Commission predicts that “this may spawn competition among MLAs from the same region to provide better constituency service, something voters would likely appreciate.”²⁰  Although tension between constituency and regional list members in Wales has recently prompted the Richard Commission to recommend replacing the Welsh mixed system with STV, proponents of regional lists for mixed systems in Canadian provinces remain confident that such competition will benefit voters, if not politicians. 

The Law Commission of Canada has also recommended the introduction of MMP at the national level. In order to maintain the current number of seats in Parliament while creating a number of party list seats to compensate parties that do not win a number of ridings that is proportionate to their share of the votes, the number of ridings must be reduced by increasing the size of riding boundaries. However, diluting the link between MPs and their constituents through larger ridings was of less concern to the Law Commission than to provincial reformers. The Commission explained: 

…at various points throughout the Commission’s consultation process, we heard from citizens who suggested that although the link between constituents and Members of Parliament is important, this concept might not fully reflect contemporary Canadian values and experiences. Today’s highly mobile and diverse citizens often identify themselves with communities of interest that are not geographically determined, or that lie outside their community of residence. It may therefore be somewhat limiting to conceptualise our electoral system primarily on the basis of territorial constituencies. ²¹ 

Electoral reform commissions in the United Kingdom and New Zealand have wrestled with similar concerns. Crewe argues that the 1976 Hansard Society Commission rejected the traditional reform option in the United Kingdom, the Single Transferable Vote (STV), in favour of a mixed system, in deference to a strong feeling for the single-member constituency. In 1998, the Jenkins Commission also recommended a “two vote mixed-system” as the best alternative for Britain. The Commission recommended that 80-85 per cent of MPs should continue to be elected on an individual constituency basis, although using the majoritarian Alternative Vote system rather than the current FPP method, while the remaining ‘top-up’ members should be elected with a second vote for an open party list. Rather than using a more proportional nation-wide constituency for the election of ‘top-up’ members, as is the case in Germany and New Zealand, the Commission recommended that “in the interests of local accountability and providing additional members with a broad constituency link, additional members should be elected using small top-up areas’, such as the existing counties or equivalently sized metropolitan districts.”  In New Zealand, the Royal Commission on the Electoral System also recommended a mixed system in 1986, in order to combine proportional representation with the traditional representation of constituents using geographically defined electorates. The Commission noted that, “the ability of constituents to take up their concerns through an MP with specific responsibilities to the local community is a healthy feature of our system which protects the rights of citizens and enhances the specific role of Parliament in attending to grievances.” 

Evaluating electoral systems according to their capacity to maintain geographic representation restricts the options that can be considered, excluding, for example, the list-PR systems used in much of continental Europe. It also changes the nature of the inevitable trade-off that is involved when assessing alternative systems. Geographic representation can often only be achieved at a cost. “For example, more women are elected to the legislature in list-PR electoral systems than in systems based on single member ‘first past the post’ districts.”²²  This is because closed party lists can be used to increase the representation of women or members of minority ethnic groups through mechanisms such as quotas or ‘zipping’, where parties attempt to balance the representation of demographic groups in their selection of list candidates. In contrast, Carty has documented the resistance of decentralised Canadian constituency associations to national party reforms designed to increase the social representativeness of Canadian MPs. Over eighty per cent of both Conservative and Liberal constituency association presidents agreed with a survey statement that, “Complaints about a lack of opportunities for women are exaggerated.  They could easily get ahead in our riding if they just got more involved.” ²³  Carty concludes that the constituency-based structure of Canadian parties has made it more difficult for parties to respond to wider societal pressures, such as calls to address the under-representation of women and ethnic minorities. 

Arguably, geographic representation may come at the cost of greater social diversity within the legislature. 

In addition, as discussed above, the emphasis on geographic representation in New Zealand, Quebec and New Brunswick precludes the consideration of alternative representational roles for list members. Barker and Levine argue that the introduction of MMP in New Zealand provided an opportunity for list MPs to develop a distinctive parliamentary role. 

This new group of MPs could articulate the philosophy of their party; act on behalf of the interests of a particular sector of society; focus exclusively on a particular issue; or seek to exercise their judgement on behalf of the ‘national interest’. The existence of parliamentarians concentrating solely on the development of policy expertise could enable parties to utilise their resources – MPs – better in parliament and so carry out more efficiently their various parliamentary functions.²⁴ 

Instead, New Zealand list MPs function as shadow electorate representatives and are described as “55 MPs in search of a constituency.” Reform proposals in New Brunswick and Quebec explicitly identify list members as representatives of specific regions within each province. 

Conclusion 

Geographic representation has traditionally defined the link between MPs and their constituents in Westminster parliamentary systems. However, this essay has demonstrated that there is legitimate cause to question the cultural attachment to single-member constituencies in Westminster democracies. The close relationship between representatives and their constituencies is exaggerated. Alternative agencies exist to redress constituents’ grievances and it is questionable whether the majority of voters value the casework role undertaken by their local MP. The link between the vote for a Member of Parliament and the vote for the government in FPP and Alternative Vote systems limits the ability of voters to hold their representatives accountable, unless they are willing to sacrifice their opportunity to express their preference for the party of government. 

The desire to preserve the close relationship between representatives and their constituencies has influenced the electoral reform process in Canada, the United Kingdom and New Zealand by limiting potential options to electoral systems that are compatible with geographic representation. In New Zealand, the potential for list MPs to fulfil alternative representative roles by representing group interests or dedicating their work to particular policy issues, has not been investigated due to the dominant conception of representation as tied to a geographic constituency. Crewe has described the attachment to single-member constituencies in the United Kingdom as “based on sentiment rather than evidence.”²⁵  The idea of a close relationship between the representative and their constituents is a key component of the political culture in Canada, the United Kingdom Australia and New Zealand, regardless of the reality of party politics in all four countries. Is geographic representation still a sufficiently important feature of Westminster parliamentary systems to justify restricting the options for electoral system reform? 

Notes 

1. Vernon Bogdanor, ‘Introduction’, in Vernon Bogdanor and David Butler (eds), Democracy and Elections: Electoral Systems and their Political Consequences, Cambridge: Cambridge University Press, 1983, p. 2. 

2. Edmund Burke, ‘Speech to the electors of Bristol’, Works, London: S & C Rivington, 1774. 

3. Donley Studlar and Ian McAllister, ‘Constituency Activity and Representative Roles among Australian Legislators’, The Journal of Politics, Vol. 58, no. 1, February 1996, pp. 73. 

4. Heinz Eulau and Paul Karps, ‘The Puzzle of Representation: Specifying Components of Responsiveness’, in Heinz Eulau and John Wahlke, The Politics of Representation: Continuities in Theory and Research, London: Sage Publications, 1987, p. 62. 

5. Donald Searing, ‘The Role of the Good Constituency Member and the Practice of Representation in Great Britain’, Journal of Politics, Vol. 47, 1985, pp. 348-381. 

6. Donley Studlar and Ian McAllister, ‘The Electoral Connection in Australia: Candidates Roles, Campaign Activity, and the Popular Vote’, in Political Behaviour, Vol. 16, no. 3, 1994, p. 385. 

7. Richard Fenno, Jr., Home Style: House Members in their Districts, Boston: Little Brown, 1978. 

8. See David Docherty, Mr Smith Goes to Ottawa: Life in the House of Commons, Vancouver: UBC Press, 1997, p. 195. 

9. Ivor Crewe, ‘MPs and their Constituents in Britain: How Strong are the Links?’, in Vernon Bogdanor (ed), Representatives of the People? Parliamentarians and Constituents in Western Democracies, Aldershot: Gower, 1985, p. 57. 

10. Bruce Cain, John Ferejohn and Morris Fiorina, The Personal Vote: Constituency Service and Electoral Independence, Cambridge, Mass: Harvard University Press, 1987. p. 54. 

11. Ivor Crewe, ‘MPs and their Constituents in Britain: How Strong are the Links?’, in Vernon Bogdanor (ed), Representatives of the People? Parliamentarians and Constituents in Western Democracies, Aldershot: Gower, 1985, p. 44. 

12. Jack Vowles, ‘The Politics of Electoral Reform in New Zealand’, International Political Science Review, Vol. 16, no. 1, 1995, p. 110. 

13. Jonathan Boston, Elizabeth McLeay, Stephen Levine and Nigel Roberts, New Zealand Under MMP, Wellington: Bridget Williams Books, 1996. 

14. Bruce Cain, John Ferejohn and Morris Fiorina, The Personal Vote: Constituency Service and Electoral Independence, Cambridge, Mass: Harvard University Press, 1987, p. 9. 

15. R.S. Milne and H.C. MacKenzie, Straight Fight: A Study of Voting Behaviour in the Constituency of Bristol North-East at the General Election of 1951, London: The Hansard Society, 1954, p 121. 

16. Bill Cross, ‘Members of Parliament, Voters and Democracy in the Canadian House of Commons,’ Parliamentary Perspectives, Ottawa: Canadian Study of Parliament Group, Number 3, October 2000, pp. 9-10. 

17. Citizens’ Assembly on Electoral Reform, A Preliminary Statement to the People of British Columbia, Vancouver, Spring 2004, p. 5. 

18. The Hon. Norman H. Carruthers, 2003 Prince Edward Island Electoral Reform Commission Report, Charlottetown, 18 December 2003, p. 83. 

19. Louis Massicotte, In Search of a Compensatory Mixed Electoral System for Québec, Working Document, Montreal, 2004, p. 164. 

20. Commission on Legislative Democracy, Final Report and Recommendations, Fredericton, New Brunswick, 31 December 2004, p. 35. 

21. Law Commission of Canada, Voting Counts: Electoral Reform for Canada, Ottawa, 2004, p. 65. 

22. David Farrell, Electoral Systems: A Comparative Introduction, Basingstoke: Palgrave, 2001, pp. 165-66. 

23. R.K. Carty, Canadian Political Parties in the Constituencies, Royal Commission on Electoral Reform and Party Financing, Research Studies Volume 23, Toronto: Dundurn Press, 1991, p. 231. 

24. Fiona Barker and Stephen Levine, ‘The Individual Parliamentary Member and Institutional Change: The Changing Role of the New Zealand Member of Parliament’, in Lawrence Longley and Reuven Hazan (eds), The Uneasy Relationships between Parliamentary Members and Leaders, London: Frank Cass, 2000, pp. 112-3. 

25. Crewe, 1985, p. 22.