Senate reform proposals in comparative perspective (BP-316E)

This paper reviews a representative selection of major proposals for reform of the Canadian Senate, in order to take stock of what has been achieved during the several cycles of public and political interest thus far provoked by this issue.

While an extensive range of options concerning, for example, methods of selection, seat distribution and powers has been developed, these are not the only, or necessarily the most important, resources yielded by previous proposals. They also provide an inventory of considerations and arguments which are useful in evaluating more recent proposals, and which the authors of future proposals will ignore at their peril. Finally, the selected proposals, when considered together, provide important clues about the evolving expectations and unstated assumptions that have shaped, and predictably will continue to shape, public discussion of this issue.

Following a historical overview of Canadian Senate reform initiatives, this paper compares a representative selection of major reform proposals under four headings: method of selection, electoral system, distribution of seats and powers. A concluding section on evolving views of the purpose of the Canadian Senate makes some general observations about the shifting assumptions and expectations among Canadians in the various regions concerning the Senate.

In the sections just described, special attention is given to the most recent addition to the national "inventory" of major Senate reform proposals: the proposal in the Charlottetown agreement of 28 August 1992. This agreement is a record of a first ministerial decision, and does not set out the rationale for its Senate reform proposal. Some tentative conclusions about this rationale and broader reflectiveness of the "conventional wisdom" that has emerged in Canada concerning the Senate can, however, be drawn when the proposal is viewed in the context of other recent proposals. These conclusions, in turn, shed light on current and probable future directions of the Senate reform debate.

The issue of Senate reform is not new to Canada. Indeed, the proposal in the constitutional package announced by first ministers on 28 August 1992 in Charlottetown is but the latest addition to a tradition of public discussion which dates back to at least 1874. In that year – just seven years after the B.N.A. Act had been adopted – the House of Commons heard a proposal that it consider amending the Constitution to allow each province to choose senators. Proposals, focusing either on the limitation of terms of appointment (appointments were originally for life) or on abolition of the body, appeared regularly during ensuing years, and in 1909 the Senate itself first debated reform. A proposal that terms be limited to seven years, and that two-thirds of senators be elected, was rejected.(1)

Beginning in the 1960s, the issue of Senate reform began to be pursued with new urgency. Heightened public and governmental attention reflected developments both in Quebec, where the Quiet Revolution was fostering new autonomist pressures, and in Western Canada, where perceptions of the unresponsiveness of central institutions were exacerbating longstanding resentments.

More broadly, the new prominence attached to Senate reform in the '60s reflected the innovative style, and focus on institutions and processes, characteristic of the federal government of that era. The emergence of the provinces as independent policy-making entities also required a broad adjustment of federal-provincial relationships.

Three decades of sporadic discussion have produced some important shifts in the content of Senate reform proposals. During the 1960s and 1970s, the emphasis was on the rehabilitation of the appointed Senate, by means of some degree of provincial involvement in the appointment of senators.

One group of proposals would have involved the provinces in one half of Senate appointments, while leaving the other half exclusively in the hands of the federal government.(2) A second group of proposals for an appointed Senate relied heavily on the West German Bundesrat model, and would have given exclusive provincial control over appointments.(3) The Senate thus would have become a kind of proxy for the provincial governments, and it was anticipated that the process of accommodating provincial concerns might shift from the domain of intergovernmental relations and First Ministers' Conferences to the federal legislative process itself. Controversy over the desirability of this development became, indeed, a central element in debate over the proposals, with some proponents of decentralization opposing a ‘bundesrat’ Senate on the grounds that it would shift powers to the federal government and reduce the scope of provincial independence.

Beginning in the early 1980s, the prevailing fashion in Senate reform proposals underwent a substantial shift. The assumption underlying successive proposals came to be the equation of elected status with legitimacy, from which it seemed to follow that the Senate could not have a major and durable impact on the legislative process unless it, like the House of Commons, was a product of democratic electoral choice-making.

Among the major proposals calling for an elected Senate are the 1981 Canada West Foundation proposal, which laid the basis for the "Triple E" reform movement; the 1984 report of a special joint committee of the Senate and House of Commons; the 1985 report of the Royal Commission on the Economic Union and Development Prospects for Canada; the 1985 report of the Alberta Select Special Committee on Upper House Reform; and, more recently, the 1992 proposal of the Special Joint Committee on a Renewed Canada.(4) This last, it should be noted, was broadly reflective of the criteria and guidelines set out in a general framework proposal released by the federal government in September 1991, which the Committee had been created to study.(5)

The Charlottetown proposal of 28 August, as will be seen, is very much in the tradition of the proposals for an elected Senate of the early 1980s.(6) Indeed, its central novelty is that it combines within one proposal many of the procedural and structural options that had previously been seen as competing alternatives.

Proposals of the '60s and '70s for a reformed appointed Senate do not oppose elected upper houses universally. They argue, instead, that the election of the second chamber is incompatible with the parliamentary system, or would sacrifice other attributes whose loss would outweigh the advantages gained.

Thus, for example, a 1978 Government of Canada paper setting out the rationale underlying Bill C-60 argues that an elected upper House is a viable option in a congressional system founded on the separation of powers, such as that in the United States.(7) It is claimed, however, that the dependence of the executive on the confidence of the elected lower House in a parliamentary system makes the election of the upper House inappropriate. An elected Senate, simply by virtue of being elected, would undermine the supremacy of the House of Commons and foster confusion. Two elected Houses would complicate the question of ultimate responsibility, and thus undermine parliamentary government.

This basis for rejecting an elected Senate was repeated in the 1979 report of the Pepin-Robarts task force.(8) It argues that the Australian experience (which involved serious deadlock between the upper and lower Houses during the mid '70s) illustrates the problems that can arise in a parliamentary system when there is a conflict between two popularly elected Houses. It also argues that an electoral process would tend to enhance the role of political parties, so that party concerns, rather than regional concerns, would tend to dominate the activity of an elected upper House.

These arguments were also made by the government of British Columbia, in 1978.(9) This proposal argues, as well, that giving an elected Senate modest powers (perhaps in the form of a suspensive veto) would not necessarily be a permanent solution to the problem of conflict between the Houses. The point is made that elected senators might justifiably feel entitled to unrestricted powers, and would exert continuing pressure in this direction.

An alternative approach is taken in the 1972 Molgat-McGuigan report, which argues that an elected Senate would be less able to represent minorities and Canadian diversity than an appointed Senate;(10) representatives of even the least numerous ethnic and other minorities could be systematically appointed to the Senate, while an electoral process would be less likely to achieve their representation.

Earlier proposals for an elected Senate share the view that elected status is a prerequisite for the legitimate exercise of influence, and several make this claim explicitly. The 1985 Alberta Select Committee, for example, argues that only a directly elected Senate "would enjoy legitimacy and would be able to exercise fully the significant political and legislative powers necessary to make a valuable contribution to the Canadian Parliament."(11) A similar view is apparent in the 1984 Molgat-Cosgrove report, which argues that the central objective of enhanced regional representation can only be achieved by reforms which "ensure that senators have more political authority."(12)

The assumption that representative legitimacy and elected status are closely connected also underlies the 1985 Macdonald Commission proposal. It is argued that:

The Beaudoin-Dobbie proposal of 1992 is explicitly based on the arguments in favour of election contained in proposals of the early and mid-'80s. It concludes:

The Charlottetown proposal would enable provincial governments to choose between two methods for the selection of senators: direct election by the people of the province, or election by the provincial legislature.(15)

The first option – direct election by provincial populations – conforms with major Senate reform proposals since the early '80s. Like them, it assumes that, in order to be effective, a reformed Senate must be an active participant in the legislative process, from time to time amending or denying passage to legislation from the House of Commons. Directly elected senators, it is assumed, would have sufficient legitimacy to be able to do this.

The second option – election by the provincial legislature – was included during the final stages of negotiations, in order to obtain agreement by Quebec. A Charlottetown agreement Senate would, therefore, likely consist of two cohorts of senators: indirectly-elected senators from (at least) Quebec and senators directly elected by the people in other parts of Canada.

One possible way of reconciling the two conceptions of representation implicit in the Charlottetown agreement Senate is to argue that provincial legislatures (that of Quebec, at least) represent the people of the province on national as well as provincial issues. If this were the case, then indirectly elected senators would be akin to directly elected senators, in their capacity to represent the people of the province on issues within the jurisdiction of the federal government. It could then be claimed that a Charlottetown agreement Senate would represent the people of the provinces by two different methods, rather than providing two different types of representation.

On the other hand, one may accept the argument, made in other recent proposals, that there is an important distinction between representing provincial legislatures or governments (mandated to deal with issues in the provincial jurisdiction) and representing the people of the provinces in national decision-making. Acceptance of this distinction could result in questions concerning the legitimacy of indirectly elected senators, particularly where they voted on federal legislation that did not affect provincial jurisdictions. More broadly, questions about the legitimacy of indirectly elected senators appear unavoidably to be implied by the rationale for directly electing other Senators.

The status of the representation provided by indirectly elected senators would, in practice, depend partly on the procedures used to elect them. They could be selected according to procedures which would ensure that a province's cohort of senators would mirror its political diversity. Alternatively, procedures placing the power to select entirely in the hands of a government majority could result in senators more closely akin to proxies of the provincial government. The Charlottetown agreement does not preclude either of these possibilities. It therefore appears to open the door to a considerable variation in the kind of representation provincial populations would have in the Senate.

The main development among proposals for selecting senators is a decisive shift in fashion from appointive systems to elective systems, which appears to have taken place in a very short time during the early '80s. This was not the result of a sudden revelation concerning the relation between elected status and legitimacy. The authors of appointive proposals had routinely acknowledged this relation, before going on to advocate appointive systems on the grounds that they would achieve the desired results, without threatening responsible government. Something other than a recognition of the relation between legitimacy and election must have occurred, therefore, in the early '80s, to turn the tide of fashion in favour of elected Senates. Attention will return to this question in the concluding section of this paper.

Previous proposals for an elected Senate yield a number of possible electoral systems. The rationales for these systems focus, in general, on their effectiveness in reflecting the complexion of political opinion within the electorate, comprehensibility to the electorate, and administrative complexity.

In 1981, a Canada West Foundation study proposed the adoption of a single transferable vote (S.T.V.) system – a variant of proportional representation, in which voters rank individual candidates in order of preference, after which a formula taking account of these rankings is employed to identify winners. It is argued that the central criterion governing choices among possible electoral systems should be the extent to which a system contributes to the role of senators as members of a chamber devoted to regional representation. This means, in particular, that systems emphasizing the role of political candidates as party representatives should be avoided.

It is argued that plurality voting systems, whether involving single- or multiple-member constituencies, tend to elect groups of regional representatives dominated by one party, and are therefore oversensitive to party affiliation. As well, most proportional representation systems are claimed to rely on central party organizations to generate party lists and therefore to be too party-dominated and unreceptive to the autonomy of local party organizations in choosing candidates. They also discourage independent candidates, a factor which is viewed negatively in a system intended to encourage regional representation rather than party representation.

Single transferable vote systems are advocated on the grounds that they permit voters to decide whether to support the candidates affiliated with a particular party, or a group of candidates affiliated with various parties (or none) and chosen on the basis of individual characteristics. Choices from among candidates of one particular party are also possible. This system is claimed to induce candidates to develop personal positions on the issues. It is argued that its complexity has not made it unworkable in countries such as the Republic of Ireland, and would not do so in Canada.

The Canada West Foundation proposal further recommends province-wide constituencies. Sub-provincial constituencies would require the redistribution of senatorial districts after each ten-year census, and differing electoral districts could confuse voters. More fundamentally, province-wide constituencies would underline the differences in role of M.P.s, who represent localities, and senators, who should represent regional concerns.

The Canada West proposal defends the simultaneity of Senate and House of Commons elections on the grounds that senators are intended to be regional spokesmen on national issues. Senate elections coinciding with provincial elections would tend to be overshadowed by provincial campaigns, and reduce the credibility of senators with respect to national issues. Since some provincial party systems differ from that at the national level, simultaneous Senate and provincial elections could also generate confusion in the minds of voters. Also, the Senate would have to cease operations whenever a provincial election required senators to be absent for campaigning (or give up membership during a campaign). While entirely separate elections have certain advantages (they would foster a clear focus on regional issues) they would also involve substantially increased costs and inconvenience to the public.

The Molgat-Cosgrove report argues for single-member constituencies and plurality voting on the grounds, first, that proportional representation involving the selection of candidates from lists drawn up by their political party would foster greater dependence of candidates on their party. This, in turn, would reduce the likelihood of non-partisan conduct in the Senate, and increase the probability that partisan considerations would prevail over regional considerations in the voting of senators. The report also argues that most Canadians have no experience with such a system, and that its operation would not be transparent to voters. This, in turn, could undermine the legitimacy of a reformed Senate.

The Molgat-Cosgrove report argues that its proposed single-member constituencies should be drawn with special attention to geographic, community, linguistic and cultural factors in order to embody natural communities. This is presented as self-evidently desirable.

The Molgat-Cosgrove report proposes triennial elections, each of which would renew one-third of the Senate's membership. The elections would be held on fixed dates. Separate elections are supported on the grounds that elections coinciding with national elections would be overshadowed by campaigns focused on electing a government, while senators elected separately would have more authority as regional representatives. Separate elections would also, it is claimed, increase the likelihood of successful campaigns by independents.

The Alberta Select Committee proposal recommends plurality voting in multiple-member province-wide constituencies. This is defended, first, on the grounds that plurality voting is already a part of the Canadian electoral tradition. It is claimed that an elected Senate would involve changes which are radical enough, without being accompanied with experimental electoral systems. It is also argued that proportional representation systems have been tried in various Canadian jurisdictions, and subsequently abolished.

Province-wide constituencies are defended on the grounds that smaller constituencies would render senators less distinguishable from M.P.s, and that provinces would be best represented as wholes, by representatives not tied to localities or distracted by purely local concerns.

The Alberta Select Committee argues that Senate elections should be simultaneous with those for provincial legislatures. A central basis for this is the view that provincial perspectives would be at the forefront of public attention during provincial elections, and would thus be highlighted in Senate elections. It is also argued that if Senate elections were to coincide with provincial elections, the ties binding senators to national political parties would be reduced, making senators likely to represent regional interests rather than national party positions.

Proportional representation is called for, to correct the tendency of plurality voting, or "first-past-the-post" systems, to underrepresent minority parties. In Canada, the report argues, there has been a persistent tendency for political parties in the House of Commons to have very few members from some regions, even though their popular vote in these regions is not vastly less than that of the successful political parties. This can lead, in turn, to cabinets in which some regions are not adequately represented. The relatively large constituencies involved in Senate proposals would, it is further argued, exaggerate these distortions if a plurality voting system were employed. Senate reform along such lines could turn out to be a step backwards in representation.

It is recognized that a proportionally-elected Senate would frequently not give a majority of seats to political parties governing with a majority in the House of Commons. The central response to this is to curtail the powers of the Senate, so that the House could ultimately prevail. It is also argued that a distinctive electoral process could enable the Senate to compete with the House of Commons for legitimacy in the eyes of the Canadian public, and that this might have interesting long-term consequences.

The Macdonald Commission report recommends six-member constituencies, on the grounds that this number is large enough to allow a system of proportional representation to operate.

The Macdonald Commission report proposes that Senate elections should be simultaneous with those for the House of Commons. This proposal is based on the claim that simultaneous elections would reduce the danger of major contrasts between the party composition of the House of Commons and that of the Senate, and thus of the possible "paralysis" if a House dominated by elected members of one party were to confront a Senate dominated by elected members of another.

The Beaudoin-Dobbie proposal recommends a proportional representation system involving (a) the nomination of slates of candidates by parties; (b) the ability of independent candidates to run; (c) the promotion, by parties, of gender equality and the representation of Canada's diversity; and (d) the option, for voters, of selecting candidates from several party slates. Rationales for this proposal are: that it would make it unlikely for the party composition of the Senate merely to duplicate that of the House of Commons; that it would reflect accurately the party preferences of voters in the various regions and avoid translating regional votes into monolithic single-party blocks of senators; and that it would better represent minorities.

The Beaudoin-Dobbie committee argued that excessively large constituencies distance candidates from voters and heighten the reliance of voters on party affiliations. At the same time, they make candidates more dependent on party assistance, giving an advantage to widely-known party notables, and creating barriers for independent candidates. The committee thus recommended that constituencies be no larger than required by an electoral system achieving proportional representation: multi-member constituencies electing at least four senators.

The Beaudoin-Dobbie Committee called for fixed electoral terms, and elections separate from those of either the House of Commons or provincial legislatures. It argued that maximizing the effectiveness of regional representation should be the central criterion in decisions about matters such as timing and length of terms, and that separate elections would distance Senate elections from those for the House of Commons, and enable them to be less partisan, thus reducing the likelihood that party affiliation would dominate the legislative behaviour of senators.

Staggered terms are opposed on the grounds that they would reduce the size of the pool of senators to be elected at each election, and thus impede the capacity of the electoral system to deliver proportional representation. The length of the non-staggered terms recommended is six years, which is claimed to avoid the risk of isolating senators from electors implicit in systems employing longer terms.

The Charlottetown agreement does not specify an electoral system for the proposed Senate. It merely prescribes that elections to the Senate should be simultaneous with those for the House of Commons, and indicates that federal electoral legislation will be "sufficiently flexible" to allow provinces and territories to provide for gender balance in their cohorts of senators.(21)

The absence of an electoral system proposal makes it impossible to determine which of the lines of argument in preceding proposals may be applicable to the Charlottetown agreement. It may be noted, however, that previous proposals have defended the simultaneity of Senate and House of Commons elections on three main grounds: economy (it would be less expensive than separate elections); practicality (it would avoid problems created by elections for different provincial cohorts – or separately elected cohorts – of senators occurring while Parliament was sitting); and appropriateness (it would elect senators in the context of party campaigns relating to national issues, rather than provincial elections relating to provincial issues).

The Charlottetown agreement does not exempt provinces in which senators are elected by provincial legislatures from the simultaneity requirement. It may be that the third of the justifications (on national issues) just cited has a special significance in these cases. Senatorial selection by provincial legislators in the context of a national election campaign might help to "federalize" the focus of the selection, and mitigate the possible concerns (raised in the preceding section) relating to the representational status of senators in this category.

Concerns about partisanship, and how it is to be handled in a body intended to represent regional interests, have resulted in contrasting approaches to the design of an electoral system among recent Senate reform proposals.

The Canada West Foundation and Molgat-Cosgrove proposals reflect the view that voting along lines of party affiliation by senators would tend to undermine the representation of regional interests and concerns, and that ways must therefore be found to reduce the prominence of political parties in Senate politics.

The Alberta Select Committee proposal takes a somewhat different tack. It focuses on discouraging the influence of the national political parties while assuming, apparently, that the affiliation of senators with provincial political parties would foster the representation of provincial interests and concerns.

The Macdonald Commission takes the opposite approach, arguing that political parties are important, and deeply-entrenched, representational vehicles in Canada, and that a reformed Senate should use them to enhance regional representation rather than attempting to minimize their influence.

The Beaudoin-Dobbie report does not address this issue explicitly. It does, however, emphasize the value of proportional representation in distinguishing the composition of the Senate from that of the House. The capacity of the Senate electoral system to do this by reflecting party preferences accurately is, indeed, the decisive consideration in the Committee's choice of an electoral system. This could not have been the case if it had been assumed that party affiliations would wither away in a reformed Senate.

The absence of an electoral system recommendation in the Charlottetown agreement prevents this proposal from confirming an otherwise modest trend towards favouring political parties, rather than the displacement of party affiliation by regional affiliation. The way in which this issue is dealt with remains, however, of key importance in establishing the credibility of any future Senate intended to enhance regional representation.

Many of the proposals for a reformed appointed Senate developed during the '60s and '70s called for a redistribution of Senate seats. Typically, there was minimal explanation of this. It appears to have been regarded as self-evident that the existing distribution of seats needed to be adjusted to reflect population growth in the Western provinces, while remaining politically acceptable elsewhere. This assumption is reflected in the proposed distributions shown in the table below.

	Molgat- McGuigan	Bill C-60	Ontario (1980)	Beige Paper	Lamontagne	Pepin- Robarts
Ontario	24	24	26	20	24	12
Quebec	24	24	30	20	24	12
B.C.	12	10	12	9	12	8
Alberta	12	10	10	8	12	6
Sask.	12	8	8	5	10	4
Man.	12	8	8	5	10	4
N.S.	10	10	10	4	10	4
N.B.	10	10	10	4	10	4
Nfld.	6	8	6	3	8	4
P.E.I.	4	4	4	2	8	2
N.W.T.	2	1	1	*	1	0
Yukon	2	1	1	*	1	0
TOTAL	130	118	126	80	126	60

* The Beige Paper indicates that this distribution is only illustrative, for the purpose of establishing proportions. It is stated that the territories should have full voting rights, although a seat distribution is not specified.

The 1978 proposal of the government of British Columbia asserted that equal regional representation was still the best basis for distributing Senate seats, but that British Columbia's emergence as a region needed to be recognized and reflected. A numerical distribution was not, however, included in the proposal.

More recent proposals for an elected Senate fall into two groups with respect to the criteria employed for distributing of seats. Proposals in one group call for an adjustment to the current distribution, to reflect demographic realities that have emerged since 1867. A second group allocates an equal number of Senate seats to each province. The contrast between the two groups is apparent below:

	Molgat- Cosgrove	Macdonald	Can-West*	Alberta	Beaudoin- Dobbie**	Current***
Ontario	24	24	6-10	6	30 / 20	24
Quebec	24	24	6-10	6	30 / 20	24
B.C.	12	12	6-10	6	18 / 12	6
Alberta	12	12	6-10	6	18 / 12	6
Sask.	12	12	6-10	6	12 / 8	6
Man.	12	12	6-10	6	12 / 8	6
N.S.	12	12	6-10	6	10 / 8	10
N.B.	12	12	6-10	6	10 / 8	10
Nfld.	12	12	6-10	6	7 / 6	6
P.E.I.	6	6	6-10	6	4 / 4	4
N.W.T.	4	4	1- 2	2	2 / 2	1
Yukon	2	2	1- 2	2	1 / 1	1
TOTAL	144	144	62-104	64	154 / 109	104

* Proposal sets out ranges, specifying that all provinces should have the same number of seats.
** Proposal sets out two possible distributions (with a third being recommended in a Liberal Party dissent).
*** Permanent seats only.

The Canada West proposal contains an extensive discussion of what should be the appropriate geo-political unit of representation in a reformed Senate. It argues that provinces have become the authentic regions of which Canada is composed, and that any representation scheme which ignores this "political reality" will be greeted with justifiable scepticism by many Canadians. The Canada West proposal then notes that the principle of equal representation is widely employed in other federally organized countries, and argues that any system deviating from this principle would not give convincing representation or protection to the concerns of all regions.

The Alberta report contains both negative and positive arguments in favour of an equal distribution of seats among the provinces. On the negative side, it is argued that representation by population would cause the Senate to duplicate the House of Commons, and would not provide a balance of provincial viewpoints. It is argued, as well, that equal regional representation is based on the assumption that regions are the authentic sub-national units of Canada. This assumption is rejected, and regions are portrayed as being merely "artificial" groupings of provinces.

The alternative to these approaches – the representation of provinces – is supported on the grounds closely akin to those set out in the Canada West proposal of four years earlier. It is argued that Canadians identify with their provinces rather than their regions. Equal representation of provinces is called for on the grounds that it is the only distribution which would give Canadians a balanced process of federal government.

These proposals oppose provincial equality in the Senate. The Molgat-Cosgrove proposal argues that disparities in size between Canada's largest and smallest sub-national units are much greater than in the United States, for example, where the largest state contains only about 10% of the population. Equal numbers of seats, it is claimed, would grossly overrepresent the smaller provinces. The Committee concludes that substantial overrepresentation is desirable, but that this must be balanced against demographic realities, and that the appropriate distribution should be based on a principle of equality weighted according to population.

The Macdonald Commission proposal incorporates the Molgat-Cosgrove distribution of seats, although it does not provide a specific rationale for this.

The Beaudoin-Dobbie proposal accepts the position of the Canada West and Alberta proposals, that the central basis upon which Senate seats should be distributed is provincial (or by territory), rather than regional. It goes on to argue that, since the purpose of a directly elected Senate is to represent the people of the various provinces, the distribution of seats cannot entirely ignore the principle of representation by population. Instead, it is claimed that the distribution should seek to balance requirements of this principle against the need for enhanced representation for people living in the smaller provinces.

It is argued, as well, that there is no general principle of federalism which requires provincial equality in upper Houses, and that, if such a principle were to be applied consistently in a directly elected Senate, it would require equal numbers of seats for the territories as well as the provinces. Two possible distributions are then recommended for consideration, on the grounds that they each respond to the various considerations that must be reflected in a distribution of seats which is fair to everybody.

The Charlottetown agreement proposes a Senate in which seats are distributed on the basis of provincial equality, with each province receiving six Senate seats and the Northwest Territories and the Yukon each receiving one.

There are two possible bases for adoption of an equal distribution of seats in the Charlottetown Senate reform proposal. The proposal could, like the equal distribution of Senate seats adopted by the designers of the American constitution, express an essentially political gesture towards the smaller sub-national units.(24) Alternatively, it could express convictions about the importance of equality, either as a matter of principle or as a practical advantage for smaller provinces.

While the Charlottetown agreement does not present a rationale for the equality of seats, background notes subsequently released by the federal government claim that:

This comment may suggest that the Charlottetown agreement reflects an endorsement of the arguments made in previous proposals which favour provincial equality, rather than simply an accommodative gesture by the larger provinces. This possibility is strengthened by the explicit commitment to "the principle of equality of the provinces" contained in the Canada clause proposed in the Charlottetown agreement.

The evolution of attitudes to the distribution of seats in a reformed Senate may be seen by comparing the average regional distributions of the earlier appointive proposals with those of the more recent elective proposals.

The appointed Senate proposals of the '60s and '70s recommended only a modest shift of seats from central Canada to outer Canada. They would have increased the proportion of seats assigned to the West by transferring seats from Atlantic Canada in greater proportions than from central Canada. In contrast, the elected Senate proposals of the '80s have recommended a more pronounced shift of seats from central Canada to outer Canada, and have involved increases in the proportion of seats allocated to both Atlantic Canada and Western Canada.

It should be noted, however, that the apparent trend towards more radical distributions of seats away from central Canada has not been entirely uniform. The second most recent elected Senate proposal considered in this paper – the Beaudoin-Dobbie proposal – allocates 40% or 37% of the Senate's seats to central Canada, 20% or 24% to Atlantic Canada, and 39% or 37% to the West, for a breakdown very similar to those of earlier appointive proposals.

The Beaudoin-Dobbie proposal's departure from the pattern of other recent reform proposals may be related to the fact that it, again unlike the others, criticizes the assumption that there would be a significant difference, in practice, between a provincially equal Senate and one embodying a less extreme degree of overrepresentation of the smaller provinces. The Beaudoin-Dobbie report argues that, in either case, the senators of an individual province will usually not be able to determine legislative outcomes. Their success as regional representatives (and thus the effectiveness of a reformed Senate in performing this role) will therefore depend more on their ability to enlist support from senators representing other provinces, and to create cross-provincial voting coalitions, than on their number.

Broadly similar powers are recommended in the various proposals for a reformed appointed Senate that emerged during the '60s and '70s. The proposals involving appointments by both federal and provincial governments favoured suspensive vetoes, ranging from 60 days in the case of Bill C-60 to six months in the case of the Molgat-Cosgrove report.(26) Bill C-60 also originated the idea of a double-majority voting procedure for measures of linguistic significance. It was proposed that these measures require majorities of both French-speaking and English-speaking senators, and that the rejection of such a measure by the Senate require its subsequent passage by a special majority in the House in order to become law.

The general direction of the "bundesrat" (provincial appointment) proposals is similar to that of earlier proposals. They would replace the inclusive power to veto legislation with suspensive vetoes, in some cases accompanied by a power of absolute veto over a limited class of legislation. Thus, for example, the Pepin-Robarts report recommends that approval of the upper House not be required for legislation within exclusive federal jurisdiction; that suspensive vetoes be available with respect to matters of concurrent federal and provincial jurisdiction; and that upper House approval be required for treaties dealing with matters in the provincial jurisdiction, or (with a two-thirds majority) for the exercise of the federal spending power within the federal jurisdiction.(27) The establishment of a role for the second chamber in the review of federal appointments was also popular.

The underlying rationale for these powers is consistent with that of more recent proposals. Both the proposals for a reformed appointed Senate, and more recent proposals for an elected Senate, undertake to replace the extensive (but normally unused) powers of the existing Senate with more limited powers whose full use is anticipated, and would be generally accepted.

In most of these proposals, it is argued that the powers recommended would enable the Senate to be effective, while at the same time preventing it from being so powerful as to be capable of deadlocking the legislative process or complicating the practice of responsible government.

As well, it is proposed that the Senate not be able to consider motions of non-confidence, and not be able to ratify foreign treaties, or appointments of ambassadors and Supreme Court Justices.

The Canada West proposal is based on two principles: The first is that the government should continue to be responsible only to the House of Commons, on the grounds that the responsibility of a government to two chambers is unworkable. The second is that an elected Senate represents the people rather than provincial governments, and should therefore have no special powers over intergovernmental matters. It is argued that the proper role for an upper chamber is ensuring that national policy is sensitive to regional concerns, and that a "stand-in" for provincial premiers is not needed, since they are entirely capable of representing their concerns in intergovernmental forums. This means that the powers of the Senate should apply to national legislation, where its status as an elected body provides a basis for powers similar to those of the House of Commons, which are required if the Senate is to be effective in its fundamental role. Its powers should not be restricted to (or distinctively focused upon) intergovernmental matters.

It is argued, further, that an elected Senate should not have a ratification power over (1) foreign treaties (because, in Canada, there is no exclusively federal treaty-making power); (2) appointments of ambassadors (because there is nothing specifically regional about these); and (3) Supreme Court appointments (because these should have provincial input).

After affirming the principle that the Senate should not be a confidence chamber, the Molgat-Cosgrove report argues that an absolute veto over ordinary legislation would make the government, in effect, responsible to both Houses. This would substantially complicate the parliamentary process, by requiring the government to serve two (potentially opposed) masters.

The example of Australia is used in support of the claim that deadlock between the Houses is the central danger created by an absolute veto. A suspensive veto is recommended on the grounds that it avoids this danger, and the need for double dissolution procedures (with the probability of proliferating elections) to which the possibility of deadlock gives rise.

It is argued, as well, that powers of delay over supply bills could paralyse public administration, and that the denial of power to overturn the government requires denial even of a suspensive veto over appropriations bills.

An absolute veto over legislation of linguistic significance is proposed in conjunction with a double-majority voting procedure, and is justified on the grounds that it is necessary in order that the procedure can be effective in its purpose of providing "additional protection for the French language and culture," – one of the original purposes of the Senate.

The rationale for the Alberta Select Committee proposals is similar to that developed by the Canada West Foundation. It is argued that a suspensive veto is insufficient to ensure effective influence on the legislative process, and that if the Senate were to become an elected body, the central basis for limiting its powers in this way would disappear.

Constitutional change is not excepted from this argument. The Committee argues that the amending formula ensures that provincial interests will be protected, and that additional protection is unnecessary.

The power to ratify non-military treaties is proposed on the grounds that social and economic impacts within particular provinces may be involved. Military treaties are treated as an exception, however, on the grounds that delays in this area may threaten national security. It is stated, as well, that possible powers such as the ratification of public service and judicial appointments properly belong to "other federal bodies."

Finally, the Alberta Committee argues that the Senate has a role in protecting language and cultural rights, and that a majority of its witnesses were supportive of a double-majority voting procedure as a means of accomplishing this.

Declaring that the aim of Senate reform should be to ensure regional sensitivity and temper majority rule, not override the principle of responsible government, the Macdonald Commission report argues that the present powers of the Senate, if actually used, could seriously complicate the practice of responsible government. It is argued that the potential contribution of the Senate to regional responsiveness can be achieved with a suspensive veto, and that a veto of six months would ensure that regional viewpoints were fully considered and would be sufficiently long to "give pause" to unrestrained majorities in the House of Commons.

The Macdonald Commission report claims, in addition, that the Senate can function as both a regionalist and a dualist chamber, and that in this latter mode it can contribute to representing and reconciling the interests of French- and English-speaking Canadians. An absolute veto over legislation of special linguistic significance is proposed (in conjunction with double-majority voting) a means of enabling the Senate to perform this second role.

The central rationale for these powers is that they would work in concert with other features of the proposed Senate to enable effective representation of regional interests while avoiding the dangers of deadlock and the paralysis of Parliament. In order to achieve this balance, the Senate is given relatively unrestricted powers to amend or defeat legislation, while deadlock is avoided by means of legislative overrides and limitations on the power to delay. Throughout the argument, the continuity of this proposal with elected Senate proposals of the mid-'80s is repeatedly stressed.

The proposal to make legislation relating to the French language and culture subject to a double majority voting procedure is defended on the grounds of widespread consensus, both among previous reports and among Canadians. The proposal to endow the Senate with a power to ratify appointments is supported on the grounds that such agencies can have significant impacts on all regions of Canada.

The Charlottetown agreement recommends the following powers for a reformed Senate:

The structure of the powers proposed for the Charlottetown Senate establishes a number of distinctions between classes of legislation, accompanied by attempts to tailor powers to the various classes.

Such an approach holds the promise of enabling the Senate to be relatively influential over the fate of some types of federal legislation, while ensuring that the House of Commons will prevail elsewhere. While it can thus be defended, generally, on the grounds that it preserves the supremacy of the lower House, it also requires specific defences of the powers established for each special category of legislation. It will be seen, in the concluding section of this paper, that some of these categories of legislation may provide useful clues about the expectations that the designers of the Charlottetown agreement Senate were attempting to meet.

A second noteworthy feature of the powers proposed in the Charlottetown agreement is the reliance on joint sittings of the House of Commons and Senate as the ultimate means for determining the fate of most bills. The joint sitting procedure would place substantial power in the hands of senators under conditions of minority government, or small government majorities, while tending to minimize their impact where large government majorities exist in the House. This is because the smaller size of the Senate would render it unable, in the context of a joint sitting, to overcome large government majorities in the House, even where voting patterns in the Senate differed significantly from those in the House.

The tendency of this proposal to vary the Senates's legislative "clout" according to the size of House majorities could be supported by the argument that there should be particularly stringent limits on the capacity of a government with weak national support, and possibly minimal support in some minority regions, to pass legislation impacting on these regions. Alternatively, it could be argued that minority/small majority government situations do this anyway, and that the Charlottetown Senate would have been least effective when it was most needed, in situations where governments had large House majorities.

The most obvious general trend relating to powers in Senate reform proposals developed during the past three decades is the gradual increase in the legislative powers seen to be required by an effective Senate. The earlier appointive proposals generally reflect the view that the federal legislative process can be made responsive to regional interests through the use of suspensive vetoes, in some cases involving delays. In the case of the "bundesrat" proposals, suspensive vetoes are supplemented, or in some cases replaced, with absolute vetoes over narrowly defined classes of legislation affecting provincial jurisdictions. The elected Senate proposals, with certain exceptions, broaden the absolute veto powers of the Senate, while relying on House overrides to maintain the supremacy of the lower House, and responsible government. The most significant strengthening of powers in the elected Senate proposals occurs, however, with respect to defined classes of legislation. The establishment of a class of legislation relating to Canada's official languages, for example, was first proposed in Bill C-60 of 1978, although that proposal envisioned an absolute Senate veto which could be overridden by a special majority in the House of Commons. With the sole exception of the Canada West proposal, the elected Senate proposals of the '80s were for absolute Senate vetoes over legislation relating to official languages (or the French language and culture), subject to a double majority voting procedure and not reversible by the House of Commons.

In addition to prescribing special powers over matters relating to the French language or culture, and retaining the distinctions between ordinary legislation, supply bills, and the ratification of appointments that have become conventional, the Charlottetown agreement establishes a new class of legislation involving tax policy changes relating to natural resources. The Senate is made predominant over the House of Commons with respect to this class, since there is no provision made for House overrides of Senate vetoes.

This proposal, which attempts to give Alberta the vital core of what has been advocated in the movement for a "Triple E" Senate, represents a significant further step in the gradual evolution, seen in previous proposals, towards increased Senate powers. It also suggests the possibility of analogous provisions relating to sectors seen to be of central importance in other regions. Indeed, the reappearance of a natural resources tax policy power in any future reform proposal is likely to result in strong pressures for the establishment of further special classes of legislation, unless a convincing case for the unique treatment of natural resources tax policy is developed.

Canadian Senate reform proposals have typically given rather cursory attention to the purpose of the Senate before turning to the details of reform. This has left relatively unexplored a crucially important group of questions relating to the problems that Senate reform is intended to address, and the expectations of reformers about what Senate reform can achieve. In this concluding section, purposes apparently envisioned for the Senate in the various reform proposals are examined for the light they can shed on the broader expectations and assumptions of Senate reformers.

According to the Molgat-McGuigan report, released in 1972, regional alienation and the emergence of separatist movements, combined with new concerns about domination by the central provinces, reinforce traditional views of the need for a regionally representative upper House.(34) Bill C-60, of 1978, took a similar approach. The purpose of the upper chamber envisioned was to protect the interests and represent the concerns of provinces and regions.(35)

It is evidently assumed, in these proposals, that appointed senators could exercise enough influence upon the legislative process to be seen as effective representatives of the regions, and could achieve this through powers limited to a suspensive veto. This suggests that effective regional representation was believed to require only a relatively modest alteration of the status quo. Modest intentions and expectations are further suggested by the minimal degree to which seats are redistributed (implying, perhaps, that outlying regions need only a moderate increase in "clout"), and by the fact that one half of the Senate would have continued to be appointed by the federal government.

The mixed appointment feature may suggest, indeed, that regional representation is not the dominant purpose of the Senate in these models. The effect of these models would likely to have been to achieve a slightly different balancing of federal and provincial perspectives within the Senate (and within the legislative process as a consequence). They would not have resulted in the emergence of the Senate as a single-purpose advocate, in Ottawa, of regional interests. Nor was this likely thought, at the time, to be needed.

The view that regional representation is a distinctively important purpose of the Senate achieves clearer prominence in proposals for a Senate wholly appointed by the provinces.

In 1979 the Pepin-Robarts Commission, for example, argued that a council appointed by provincial governments could combine the representation of regional interests with the institutionalization of the processes of executive federalism, in which Members of the council would act as proxies for their provincial governments on matters affecting provincial jurisdictions.(36) The 1980 Quebec Liberal Beige Paper adopted a similar approach, arguing that increased functional interdependence between the levels of government creates a need for an institution that enables the provinces to influence federal initiatives that impact on provincial jurisdictions.(37)

In some provincial appointment proposals, regional representation is explicitly equated with provincial representation. In its 1978 proposals, for example, the B.C. government argued that the purpose most clearly articulated by the Fathers of Confederation was "the representation of provincial interests in the making of national laws."(38) It is claimed that the Senate has failed to do this effectively.

The shift away from the mixed appointment model to provincial appointments, combined with the role (in several of these proposals) of provincial governments in selecting Senators, suggests an important shift in underlying purpose. The "bundesrat" Senates do not attempt to balance regional representation with other purposes. They first serve the purpose of regional representation, with other purposes playing a highly subsidiary role. Secondly, regional representation is clearly equated with provincial representation, in these models, and provincial representation is equated with the representation of provincial governments rather than the direct representation of the people in each province.

The potential impact of provincially appointed Senates is moderated by two features: they are appointed bodies exercising only (for the most part) suspensive vetoes, and their most significant (ie. longest suspension or, in some cases, absolute) veto powers are restricted to classes of federal legislation affecting provincial jurisdictions. These features, together with the fact that the major redistributions of seats proposed are from the Atlantic provinces to the West, rather than from the large provinces to the small provinces, have implications for the practical meaning of the regional representation which ‘bundesrat’ Senate models would have achieved.

In practice, this "regional representation" would have amounted to a modest increase in the influence of provincial governments, collectively, within the federal legislative process. Provincial influence relating to federal legislation affecting provincial jurisdictions would have increased more significantly. The relative numbers of Senators from "outer" Canada and central Canada would not, however, have been greatly altered (although representative weight would have migrated from the East to the West).

As with the mixed appointment models, one may infer that the practical meaning of "regional representation" emerging from "bundesrat" models tells us what their creators thought was needed. The central problem was seen as a problem of insufficient influence in Ottawa by the governments of the provinces, both large and small. More fundamentally, it was assumed that any feelings of alienation or exclusion then being experienced by people in Canada's regions could be addressed by increased provincial influence over, primarily, federal impacts within provincial jurisdictions.

More recent proposals for an elected Senate continue to assert that the need to enhance regional representation is the fundamental reason for Senate reform. Interestingly, some have undertaken explicitly to state and justify views about what regional representation is and why it is needed. This may suggest an awareness, on the part of authors, that they were expressing a new and possibly controversial conception of the Senate's traditional purpose.

The Canada West Foundation-sponsored study of 1981 argues that the Senate was originally created to serve the need for regional representation, but has not done so because its status as an appointed body has undermined its legitimacy. It is argued, as well, that the four Divisions referred to in the Constitution Act, 1867 are not bound together by sufficient ties of historical association, culture, economic similarity or geographic kinship to be more than arbitrary regions, and that regional representation should therefore be understood as provincial representation. Thus, in order to fulfil its original purpose, the Senate has to be reformed so as to provide representation on a provincial basis.(39)

In the 1984 Molgat-Cosgrove Committee report, 1985 Alberta Select Committee report and the Macdonald Commission report of the same year, substantially the same argument is adopted. The central purpose of the Senate is portrayed as regional representation, which is equated with representation on a provincial basis, rather than the representation of Canada's traditional regions. It is argued in all three reports, as well, that representation on a provincial basis does not mean the representation of provincial governments; representation within the federal legislative process is portrayed as having the potential to undermine the intergovernmental process (Molgat-Cosgrove, Alberta), and as incorrectly employing provincial governments, rather than federal M.P.s and senators, to represent regional interests relating to national issues (Macdonald Commission). The reports differ, however, on whether representation on a provincial basis requires equal representation; the Molgat-Cosgrove and Macdonald Commission reports take the position that it does not.(40)

The discussion of the purpose of the Senate in the 1992 proposal of the Beaudoin-Dobbie Committee reflects both the insights reached in previous studies and the assumptions underlying the federal government's 1991 framework proposal, which the Committee was created to study. It is argued, fundamentally, that regional representation should be the primary purpose of the Senate because enhanced representation, particularly of the outlying regions within the central government, remains a major need within the country. It is argued further that differences among the provinces make it difficult adequately to represent their populations on the basis of the traditional regions, and require regional representation to be understood as provincial representation. Finally, it is argued that provincial governments have no mandate to represent their populations on national issues, and that the form of representation sought from a reformed Senate therefore requires senators to represent the people of the provinces, rather than provincial governments.(41)

Architectural features of the proposals just reviewed confirm what is actually said about the purpose of the Senate. In recommending an elected Senate, for example, these proposals reflect a redefinition of the representational problem to be addressed by Senate reform. The focus is shifted from provincial governments to the people of the provinces, and the Senate comes to be seen as the institutional mechanism which will remedy popular feelings of alienation and exclusion.

With respect to the distribution of Senate seats, these proposals defend their recommendations on the basis of arguments about the fair treatment of the populations of the various provinces, and abandon appeals based on the treatment of Canada's traditional regions. They also, perhaps reflecting this, involve more radical redistributions away from central Canada. As well, the heightened representational status that would be possessed by elected senators is reflected in the recommendations concerning Senate powers contained in these proposals, which prescribe generally greater powers than were envisioned in earlier proposals.

In short, while elected Senate proposals retain the language of "regional representation" traditionally used to define the purpose of the Senate, they reflect a substantial shift in thinking about what the Senate should actually accomplish within our legislative system. This shift in thinking was reflected, more recently, in the reform proposal contained in the Charlottetown agreement.

The Charlottetown proposal does not spell out the purpose that it is supposed to serve. A fact sheet distributed by the federal government in September 1992 declares, however, that the upper House is "designed to give the less populous regions a greater say in the policy-making process of Parliament," and that the proposed Senate would give "the less populous provinces more clout in the legislative process."(42)

The architecture of the Senate proposed in the Charlottetown agreement further suggests that its underlying purpose is regional representation, which is equated with representation of the provinces. The proposal would redistribute seats on a provincial basis, rather than on the regional basis adopted in 1867. Indeed, with its recommendation that all provinces have the same number of Senate seats, the proposal repudiates the lingering attempt to maintain some degree of equality among the traditional regions which has characterized several other recent proposals.

Like other recent proposals, the Charlottetown proposal reflects the assumption that regional representation will not be truly effective unless representatives have real influence within the legislative process. It provides for the direct election of senators, and thus for their enhanced legitimacy as the representatives of regional interests.

Unlike other recent proposals, however, the Charlottetown proposal leaves the door open for elections by provincial legislatures as an alternative to direct election. It thus provides provinces and territories with the option of basing their selection of senators on a conception of regional representation which harks back to the "bundesrat" models, within a Senate otherwise based on a different conception of regional representation. This creates the possibility that a Senate modelled on the Charlottetown proposal would provide several kinds of representation, endowing its general purpose of regional representation with a plurality of specific meanings.

The suggestion of a plurality of purposes residing under the general language of "regional representation" is strengthened by the way in which the Charlottetown proposal deals with powers. As has been seen, the proposal establishes a new benchmark in the growth of numbers of classes of legislation and corresponding legislative powers. With respect to powers concerning tax policy legislation related to natural resources, for example, the Charlottetown Senate would appear to be more powerful than any of the other proposals examined in this paper. With respect to the category of "ordinary legislation," on the other hand, the Charlottetown Senate would appear normally to have somewhat less leverage than Senates proposed in other recent proposals, which would have established lengthy suspensive vetoes and/or House of Commons overrides requiring special majorities.

If the Charlottetown model is taken as an expression of what is needed in order to fulfil the purpose of regional representation, it thus may add something to preceding proposals. More clearly than its antecedents, it suggests that "regional representation" has come to express expectations so specific and divergent that they can no longer be met by an institution that handles all issues, relating to all regions, in the same way.

As has been seen above, Senate proposals have consistently focused on regional representation as the fundamental purpose of the upper House. Beneath continuities of language, however, some noteworthy shifts in concrete expectations have occurred. Expectations about the representation that a reformed Senate should provide have moved from the representation of Canada's traditional regions to the representation of Canadians on a provincial basis, and, in most proposals, from the representation of provincial governments (or, in a few cases, legislatures) to the representation of people by directly elected representatives.

This latter shift, in particular, implies a substantial increase in expectations about what the Senate, and Senate reform, can achieve. The nature of the shift itself implies rising expectations about the Senate, because its role is broadened beyond the intergovernmental arena to which "bundesrat" proposals would have largely confined it. As well, feelings of regional resentment appear to have been steadily on the increase during the early and mid-'80s when most of the elected Senate proposals were developed. This increased both the urgency and the difficulty of the challenge to existing federal institutions which proponents of reform undertook to meet through changes to the Senate.

It is probably no coincidence, as well, that the tide of fashion turned decisively from reformed appointed Senates with moderate powers to elected Senates with, in some cases, powers approaching those of the House of Commons during the early '80s, when resentment over the NEP reached its peak in the West. The Triple E Senate movement in particular, which crystallized around the proposal of the 1985 Alberta Select Committee, is evidence that Alberta's experience with the NEP has cast a long, and persisting, shadow over ensuing thought about Senate reform. When proponents of the Triple E Senate use the familiar language of regional representation, they appear to be expressing an expectation which is quite new: that outlying provinces should have the means to stop federal legislation that they perceive to be against their central interests.

This shift suggests perhaps the most important lesson to be learned from a review of the purpose of the Senate as expressed in successive reform proposals. Beneath the repeated general assertions that this purpose is regional representation, are progressively more specific and intense expectations. From contemplating a modest, but ill-defined, increase in regional representation in the early '70s, proposals have evolved through progressively more ambitious methods of giving people in the outlying regions more clout, with proposals from the Western provinces most clearly exhibiting this tendency. The Charlottetown proposal reflects the reality that there is now a considerable divergence among the various regions about what should be expected from Canada's Senate.

While the future course of discussion about Senate reform cannot be predicted, it is unlikely that the expectations that have propelled the evolution in reform proposals thus far will spontaneously diminish. Indeed, the most probable development is that they will continue to increase, and that future Senate reform proposals will reflect this increase, even as they foster it.

(1) Janet Marie McCauley, "The Senate of Canada: Maintenance of a Second Chamber Through Functional Adaptability," Doctoral Dissertation (unpublished), Pennsylvania State University, May 1983, p. 53-54.

(2) This group of proposals includes a White Paper (Government of Canada, The Constitution and the People of Canada, 1969); Bill C-60 of 1978 (the somewhat belated translation of the Paper into legislative language); Canada, Special Joint Committee of the Senate and the House of Commons on the Constitution of Canada, Report, 1972 (referred to below as Molgat-MacGuigan); and Canada, Standing Senate Committee on Legal and Constitutional Affairs, Report on Certain Aspects of the Canadian Constitution, 1980 (referred to below as Lamontagne).

(3) Variations on the "bundesrat" model were proposed in: British Columbia, Reform of the Canadian Senate, Victoria, 1978 (referred to below as B.C.); Canada, Task Force on Canadian Unity, A Future Together – Observations and Recommendations, Ottawa, 1979 (referred to below as Pepin-Robarts); Ontario, Advisory Committee on Confederation, Reports (First and Second), 1978 (referred to below as Ontario); and Constitutional Committee of the Quebec Liberal Party, A New Canadian Federation, 1980 (referred to below as Beige Paper). All these proposals envisioned a Senate appointed and recalled by provincial governments, and acting on their instructions.

(4) See, respectively, Peter McCormick, Ernest C. Manning and Gordon Gibson, Regional Representation: The Canadian Partnership, Canada West Foundation, September 1981 (referred to below as Canada West Foundation); Canada, Parliament, Report of the Special Joint Committee of the Senate and of the House of Commons on Senate Reform, Queen's Printer, Ottawa, January 1984 (referred to below as Molgat-Cosgrove); Canada, Royal Commission on the Economic Union and Development Prospects for Canada, Report, Volume III, Minister of Supply and Services Canada, Ottawa, 1985 (referred to below as Macdonald Commission); Canada, Alberta, Report of the Alberta Special Select Committee on Upper House Reform, Strengthening Canada: Reform of Canada's Senate, Edmonton, March 1985 (referred to below as Alberta Select Committee); and Canada, Parliament, Report of the Special Joint Committee of the Senate and of the House of Commons on a Renewed Canada, Queen's Printer, Ottawa, 28 February 1992 (referred to below as Beaudoin-Dobbie).

(5) See Government of Canada, Shaping Canada's Future Together – Proposals, Minister of Supply and Services Canada 1991, p. 16 f.

(6) See Consensus Report on the Constitution, Charlottetown, August 28, 1992 and Draft Legal Text, October 9, 1992.

(22) See Canada West Foundation, p. 110-111, and Alberta Select Committee, p. 26 f.

(23) See Molgat-Cosgrove, p. 28 f, Macdonald Commission, p. 89, and Beaudoin-Dobbie, p. 51.

(24) Jacob E. Cooke, ed., The Federalist, Wesleyan University Press, Middletown, Connecticut, 1961, p. 416-17.

(25) Government of Canada, Our Future Together – Fact Sheet: Responsive Institutions, 1992, p. 2.

(35) See Government of Canada, The Constitution and the People of Canada, Ottawa, p. 30 f.

(40) See, respectively, Molgat-Cosgrove, p. 19; Alberta Select Committee, p. 14 f; and Macdonald Commission, p. 87 f.

(42) Our Future Together – Fact Sheet, "Responsive Institutions," p. 1 and p. 2.