If democratic legitimacy is seen as synonymous with elections, representative institutions which are not directly elected face a double problem: their authority can be questioned and their role is unclear. This is the case whatever the constitutional provisions specifying their function and the extent of their formal powers. And these are precisely the charges which have dogged the Canadian Senate. Whatever the merits of its activities, they have been undermined by a lack of institutional legitimacy.

Yet to argue that elections are the sole source of political legitimacy is too broad a claim; the courts in Canada, particularly since the Charter, are an obvious example of an institution whose widely accepted authority does not derive from popular representation. Elections are only one component of liberal democracy; constitutionalism, the rule of law and an institutional framework both to implement and to monitor representative government are equally important. Nor do elections alone guarantee popular acceptance of an institution; the directly elected Australian Senate for much of the first half-century of its existence was regarded as a chamber of little significance. Political legitimacy for a representative institution requires more than elections; the question is whether legitimacy can be achieved without elections. If the answer is yes, this assumes there is some institutional design which can generate enough political authority to sustain the legitimacy of a representative institution.

The many schemes for the reform of the Senate through the modification of the selection process of senators without using direct election have been motivated by the belief that political legitimacy can be achieved by institutional engineering of the selection process. The recent experience of the UK House of Lords adds weight to this view — the drastic reduction of hereditary representation and the introduction of balanced partisan appointments have enhanced the public acceptance of the role of the chamber. But many questions remain about the changes which could be applied to the selection process of senators and their consequences for the role of the Senate and the parliamentary process. These issues have been very well canvassed in the literature on the Senate. The chamber, and the possibility of its reform, has produced a great deal of excellent scholarship,1 and it could be argued that there is little to say on the topic that has not already been said by previous generations of scholars. The contribution of this paper is to put past debates into a contemporary context and, by looking at the experience of bicameralism in similar parliamentary democracies, to provide some suggestions about the reconfiguration of existing ideas about Senate reform — and to set out another set of proposals.

The paper begins with an attempt to avoid the confusion which sometimes surrounds debates over Senate reform by setting out the differing assumptions about the purpose of the Senate, the goals for reform and the methods of achieving them.2 This part of the paper also looks at the meaning of legitimacy as an attribute of institutions, and its association with democracy and representation. The following part examines the relationship between the power and the legitimacy of parliamentary institutions, and argues that these two attributes can contribute to the creation of a stable role for an appointed upper house in the parliamentary process. This is supplemented by an examination of the way in which the House of Lords and the Australian Senate have evolved to achieve such a role in their parliamentary systems.

The analysis then moves to the Canadian Senate and examines the role of the prime minister in the appointment of senators and the implications of the Senate’s fixed size and regional constraints on appointments in shaping its partisan dynamics. This is illustrated by an examination of the way in which Senate reform can be used as a tactical device for partisan gain. A survey is then undertaken of the issues raised by the reform of an appointed Australian state upper house and its broader implications.

Finally, the paper sets out six propositions concerning the Senate which argue that the most plausible course for reform is a choice from the wide repertoire of appointment procedures to achieve partisan balance. These must be combined in a way which enhances the legitimacy of the Senate while not unduly threatening the government or the partisan ambitions of the House of Commons. All but one of the propositions could operate without constitutional change; the remaining one requires a removal of the Senate’s legislative veto as part of a package of changes to ensure the acceptability of a reformed Senate.

Conflicting Assumptions about Senate Reform

There are few commentators on the Senate who think that it should be left as it is — that reform of any kind is undesirable. But, putting aside those who wish to abolish the Senate,3 the wide acceptance of the need for Senate reform masks disagreement over the extent and direction of desirable change. Contributing to the difficulty of discussion are profound differences in assumptions about the purpose of the Senate, the goal of Senate reform and the method of achieving reform. This makes for a confusing debate, so it may be helpful to set out the different assumptions in each of these areas.

The purpose of the Senate

Examining the purpose of the Senate is not a matter of historical accuracy but of current assumptions about what the Senate is supposed to achieve (see Stilborn 2003). For some, the Senate should represent the social and regional diversity of Canada as a way of bringing within the parliamentary process voices which might not be heard in the partisan debates of the House of Commons. One version of this view is eloquently stated by Ajzenstat (2003), who accepts that a parliamentary executive with a lower house majority will prevail eventually, but feels that delay and the canvassing of alternative proposals in the Senate is an appropriate and necessary parliamentary function. She argues that this is consistent with the founders’ assumptions that the Senate’s ability to reflect sectional and regional interest was what made the Senate an integral — and logical — component of a national parliament (but see Smith 2004, 47-50). Ajzenstat might admit (2007) that this view of the Senate rests on notions of limited government, parliamentary representation and the need for executive restraint which are now not widely shared, but similar views about the proper purpose of the Senate underpin David E. Smith’s extensive work on the chamber (2003a, 2003b).

Some put more stress on the representation of regional diversity, but also accept that being able to manifest this diversity in the Senate, rather than to force compliance, is its most important attribute. This perspective sees the Senate as a chamber which could be reformed without the need for popular elections. Elections, even with the use of electoral systems designed to encourage regional and partisan diversity, would still limit representation to a mechanical process rather than encourage the variety of opinions possible from a modified appointment process.

Another view is that the purpose of the Senate should be to complement the parliamentary activities of the House of Commons (Smith 2003a, 159). The Senate should examine the fine details of legislation, use the committee system to provide further information relevant to decisions on legislation and government administration and hold inquiries on matters of public interest (Thomas 2003, 215-26); but all these deliberative activities should be in the form of advice to the executive (Parkinson 2007). The Senate is intended, in other words, as a chamber of review and expert commentary; it should not be the forum for partisan confrontation between the Senate and the government. This makes its representative function subordinate to its role as a facilitator for the efficient functioning of the parliamentary process. Again, this perspective sees reform of the Senate as a process which need not involve direct elections; the greater the stress on the importance of the Senate as a source of expertise and informed review, the less relevant elections become.

A very different analysis of the purpose of the Senate is held by those who see the parliamentary process in more combative terms. For them, the Senate’s principal purpose is to act as a check on the executive.4 Party discipline has long reduced the ability of the House of Commons to subject legislation to the extensive examination it deserves and, with the partial exception of minority governments, to sustain effective inquiries into the activities of government. The role of the Senate is to fulfill those functions of parliamentary scrutiny that the House can no longer perform adequately.

This perspective challenges many assumptions about the Canadian parliamentary process, critical ones being the dominance of an executive supported by a House of Commons majority and conventional views of the operation of responsible parliamentary government. It implies that the government must not be able to rely on a partisan majority in the Senate. Of greatest significance, it assumes that the Senate needs a source of legitimacy to underpin forceful resistance to the government of the day, and that direct popular elections are likely to be the only way in which this can now be achieved. Smiley (1985, 356) posed the consequent dilemma:

No case at all can be made for a weak elected Senate in Canada, a body whose powers can easily be overridden by governments sustained by majorities in the House of Commons. Yet the existence of any other kind of elected second chamber is almost impossible to reconcile with the operative rules of responsible government as Canadians have come to understand them. However, responsible government should not be a shibboleth, both the empirical and normative assumptions on which this regime is based are questionable. Thus an elected Senate with the power and assertiveness to protect regional interests effectively will inevitably challenge legislative supremacy which is in essence a check for executive power.

Finally, there are those who see a powerful elected Senate as one way of achieving real parliamentary supremacy in a Canada which can finally cast off the executive dominance claimed to be inherited from its colonial past (Sproule-Jones 1984; and note Sharman 1990).

These three perspectives on the purpose of the Senate — as the parliamentary representation of social and regional diversity, as a vital component of an efficient parliamentary process and as a necessary check on the executive — explain why proposals for Senate reform generate disagreement. Each suggests a different dynamic for the operation of the Senate and its relationship with the lower house and with the government of the day. Not surprisingly, these differing perspectives suggest contrasting goals for Senate reform.

The goals of Senate reform

The key factor in distinguishing among the goals of those who argue for Senate reform is their attitude toward the role of partisanship and political parties. If the Senate is to be a body of expert legislative advice, party is largely irrelevant. Decisions should be made on merit, and partisan differences do nothing but complicate the deliberative process. This is an overstatement, but it points to the dilemma of reforming the Senate as though it were a committee of inquiry or a debating society. The chamber has real power, and some organizing principle must be used to coordinate majorities so that decisions can be made. As long as the decisions of the Senate impinge on the political priorities of the government by amending legislation, scrutinizing government actions or public inquiries, partisan politics will play a critical role. It may not always trump decisions made on other criteria, but it will be the dominant consideration in the operation of the Senate.

But to stress the importance of party does not require the persistence of the kind of party warfare that typifies the current operation of the Senate. Some suggestions for reform deal explicitly with the question of party and seek to moderate its role. One way is to produce a balance between the two major parties and to change the way senators are chosen so that parity is maintained even when there is a change of government. In addition, reform could ensure that parties other than the two largest are regularly represented in the Senate, together with, perhaps, the occasional independent. The goal of these reforms is to recognize the central role that the governing and major opposition parties play in the parliamentary process, but to temper the operation of the Senate by having a significant cross-bench component. If this went so far as to require the governing party to gain the support of cross-bench senators to achieve a majority, the major parties would retain their dominant role but would be subject to the moderating effect of compromise and coalition building with other parties and independents.

Other reforms see party playing a different role. For those whose principal goal is to make the Senate responsive to the political priorities of the provinces, party is the agency for regional responsiveness — party as determined by the government in each province or as decided by popular election. In the German Bundesrat model, under which all the senators from a province would be selected by the provincial government, partisanship for each delegation of senators would be defined as loyalty to the governing party in the province. Direct election of senators from each province would inject provincially based parties into national politics. This would certainly not reduce the importance of party, but might work to blur the difference between provincial and national parties (the German experience is surveyed in Detterbeck and Renzsch 2003). Whether appointed or elected, the partisan attachments of province-based senators would transform the operation of the Senate in ways which are unpredictable.

For a few reformers, the most important goal of Senate reform is to enhance the role of party rather than to reduce it. In addition to ensuring that the governing party would not likely control a Senate majority, reform would entail providing nongovernment parties with much greater political legitimacy so that they could use to the full the Senate’s powers to force negotiation with the government over its legislative proposals. At the other end of the scale, there are those who aim for the Senate to be guided by a much weakened partisanship; the selection of senators would reflect regional diversity by appointing people who have status in their communities. In this case, partisan attachment would be blended with personal attributes to ensure that party considerations were moderated by local experience and personal beliefs.

Whatever the goals of Senate reform, consideration of the role of parties is unavoidable in the selection of senators, in the operation of the Senate and in its role in the parliamentary and governmental process.

Method of achieving reform

The focus of this paper is not on the circumstances in which successful reform of the Senate might be achieved or on an extensive history of proposals for reform since the 1960s (see Seidle 1992; Stilborn 2003). Rather, the issues to be canvassed here are the institutional changes to the Senate that reformers advocate. Chief among these are changes to the procedures for selecting senators. This might appear to provide a relatively limited set of choices, but it has not prevented the topic from being one of the most contentious. The big divide is between the direct election of senators and all other methods of selecting senators.

Those who argue for direct election argue that the Senate needs the political legitimacy that only direct election can bestow. This raises the question of what the added legitimacy is to achieve. For those who wish the Senate to use its powers to confront a government supported by a majority in the House of Commons, the purpose of direct elections is clear. But few reformers regard this as the primary role of the Senate. Those who want the Senate to reflect the interests of regional political communities see elections as a way of both expressing and legitimizing their concerns. But a Senate elected on this basis is likely to succumb to the dictates of party even if it is a different constellation of parties than those which now populate the Senate benches. Regional elections are just as likely to lead to confrontations between the Senate and the government on partisan grounds as a Senate elected on some other basis. Legitimacy intended for one purpose may end up supporting a very different one.

Elections are assumed by many reformers to be the only way in which popular support for the Senate can be achieved; democratizing the Senate through an electoral process is seen as a cure-all. The hope that the direct election of senators can, by itself, solve the problems of the role of the Senate is certainly misplaced. It is easier to argue that direct election without other institutional changes would exacerbate the problem of the Senate — this would certainly be the view of the government of the day. As mentioned above, it is hard to imagine direct elections, however organized, without political parties playing a central role, and it is the likely goals of those parties rather than the electoral process that is the critical variable. To propose direct elections brings the debate back to the question of the purpose of the Senate and the goals of reform.

If indirect elections are ignored — an arrangement enabling some or all of the sitting members of the federal or provincial parliaments (or a convention elected for the purpose of choosing senators) to elect senators — the other instrument of reform is to amend the appointment procedure. At the moment, recommending the appointment of senators is the prerogative of the prime minister, subject only to the availability of Senate vacancies and the rules specifying the regions from which senators are to be appointed. This process has been roundly criticized as an exercise in partisan patronage which robs the Senate of the public support necessary for it to exercise its powers effectively.

Alternative appointment procedures are variations on two themes. The first accepts the partisan nature of appointments, but argues for senators to be nominated by party committees with rules that permit all parties to select senators, even if the governing party can ensure a majority on the floor of the Senate. This process could be combined with opportunities for regional participation and review by a body similar to the House of Lords Appointment Commission, set up in 2000 as an independent statutory body to vet party nominees for their “propriety”5 and to make its own recommendations for nonpartisan appointments. Again, the goals of this reform become an issue: what is a Senate reformed along these lines likely to achieve that the current one cannot?

A more adventurous change would be to have senators nominated by a nonpartisan panel, perhaps a panel of notables chosen by the government, with guidance as to what sort of person should be selected. Or it could be a standing commission like the House of Lords Appointment Commission, which selects appropriate people from a long list of names suggested by the public (see United Kingdom 2000). Both of these options and any system which was based on partisan recommendations, however, might be subject to the approval of the prime minister.

To list these options is to reinforce the depressing circularity of debates about Senate reform. If the Senate were elected, it could threaten the current dynamics of the Canadian parliamentary process in ways which could have profound effects on the operation of government and the role of the House of Commons. If the Senate were appointed, despite any good work it might do, it would be seen as either a vestigial attachment to Parliament or, if the Senate were controlled by a nongovernment majority, as an irritant lacking the legitimacy to do anything but harass the government at the behest of the opposition in the House of Commons.

A note on legitimacy

As a supplement to the discussion above, those who wish to reform the Senate often refer to their goal as democratizing the chamber, increasing its representativeness and, perhaps most frequently, enhancing its legitimacy (see Stilborn 2003, 58-9). In a study focused on the debate over the reform of the House of Lords, Kelso (2006) makes some helpful distinctions between these terms, and points to the confused way in which they have been used to describe and justify the many schemes for Lords reform. Democratization, she suggests, is intimately related to having an elected component — there have been several proposals for changing the composition of the House of Lords so that some, most or all of its membership is directly elected (see McLean, Spirling, and Russell 2003). But elections may not enhance the representativeness of the chamber; if political parties are the main organizing agents for elections, the diversity of views, social attributes and occupations in society are unlikely to be fully represented. If representativeness — as opposed to popular representation — is the major goal, some form of appointment may be a more effective way of achieving it. Of course, representativeness may not matter to those who believe that electoral democracy is of overriding importance in the design of legislative institutions, but the distinction is important.

Kelso’s (2006) treatment of institutional legitimacy is especially useful. She points out that there are two kinds of legitimacy: input legitimacy, which relates to the way in which the members of an institution are selected, and output legitimacy, which refers to the public assessment of the relevance and quality of the institution’s performance (for an extended treatment of this topic, see Beetham 1991). Both forms of legitimacy express public assessment of the worth of an institution, but input legitimacy is a matter of the design of the institution while output legitimacy must be earned by the institution’s performance. Assessments of the Senate are an excellent example of the gap between the two: the Senate clearly performs many useful functions in the Canadian parliamentary process (see Franks 2003) but the public, largely unaware of this, focuses almost exclusively on its lack of input legitimacy. The whole thrust of the debate on the reform of the Senate is not the quality of its institutional performance but the way it is chosen.

Power and Legitimacy

The elephant in the room in debates about the composition of the Senate is the extent of the Senate’s powers. The Senate has the same powers as the House of Commons, with the exception that appropriation and tax bills “shall originate in the House of Commons” (Constitution Act, 1867, section 53). This means that the Senate has the power to veto any legislation proposed and passed by the lower house, to introduce a bill on any topic other than a money bill, to set up committees of inquiry on any subject and to act as a forum for wide debate on any matter of public importance. Of greatest significance, although the Senate does not have the power to originate financial legislation, it may block money bills, including the budget and other key appropriation bills essential for the operation — and the life — of a government. This power makes the Senate a potential threat to the government’s control of the parliamentary process, notwithstanding those who argue that there is something vaguely unconstitutional about the Senate’s use of the legislative veto (Rémillard and Turner 2003, 126-7).

The power of the Senate was amply demonstrated during the Progressive Conservative government of 1984 to 1993, when the Liberals were usually able to muster Senate majorities to harass the government’s legislative program and to block the passage of the occasional bill when they believed they had the support of the public. Even Prime Minister Mulroney’s unprecedented use of section 26 of the Constitution Act, 1867 in 1990 to appoint eight additional senators to enable the passage of legislation to introduce the goods and services tax can be seen as strengthening, rather than weakening, the power of the Senate (Franks 2003, 155-65). The Mulroney government could have challenged the Senate directly by calling an election on the issue, but the unpopularity of the measure and fear of defeat forced the government to admit that the Senate was a critical player in the legislative process whose consent had to be gained for the government to implement its financial measures.6

Governments usually have partisan majorities in the House of Commons, which, when combined with strong party discipline, give the government of the day a free hand in the running of the lower house and deny the opposition parties the ability to force changes on government legislation or to pursue inquiries into matters the government does not want aired. This is not to deny the importance of the House as a forum for debate on matters of public importance and for providing numerous opportunities for embarrassing the government. But this is a long way from fulfilling the nominal role of the chamber: the independent scrutiny of legislation and government activities. With the partial exception of periods of minority government, the House has long ceased to be an autonomous actor in the parliamentary process.

Executive control of the business of the House of Commons is now taken for granted, and even celebrated as a way of ensuring the electoral accountability of governments. Parties in government have a period between elections to follow their preferred policies and pass legislation with few formal parliamentary restraints on their activities, and must then justify their activities at the next election. This view of the parliamentary process is so ingrained that those who suggest electoral reform which might deny the government a solid partisan majority on the floor of the lower house are accused of fostering government instability. The thought that governments might have to justify legislation on its merits in parliamentary debate, and to persuade members of Parliament other than government partisans to support it, is seen as almost revolutionary. Executive dominance has become the defining characteristic of the political process in Canada.

This explains why the position of the Senate is anomalous. All provincial legislatures are now unicameral,7 and the idea that parliaments should be bicameral to provide an upper house which can act as a continuing check on governments supported by a lower house majority has little currency in Canada. But the Senate continues to hold the power to disrupt the government’s control of the parliamentary process. This anomaly has been papered over by degrading the status of the Senate so that its members do not feel they have the authority to use their powers to challenge governments, especially if they face a government determined to get its way.

Most of this diminution of legitimacy is the result of the nature of the Senate’s composition: the appointment, rather than election, of senators, the permanent nature of their tenure until the age of retirement, the lack of all but the most cursory review of senators once they are appointed, the fact that appointments are made by the prime minister alone and the manifestly partisan way in which the process of appointment has been used. The fixed size of the Senate enables a prime minister who has access to sufficient Senate vacancies to further enhance the government’s security by creating a majority of partisans in the Senate so that the chamber’s political weakness is compounded by partisan indifference. This situation has not occurred by accident; it is the response of an executive which wishes to reduce the uncertainty of having a powerful Senate.

Before discussing the implications for Senate reform of this disparity between formal power and political legitimacy, it may be useful to look at the position of the Senate in comparison with other possible combinations of power and legitimacy. Power in this context is equated with the constitutional power of the institution to take independent action in the legislative process. Legitimacy is a more elusive concept and deserves further examination, but for present purposes, it is the ability of a chamber to count on popular acceptance of the use of its formal powers to limit government action, whether or not they agree on the policy at issue.

In figure 1, the United States Senate is clearly located in the high-power, high-legitimacy quadrant and the Canadian Senate in the high-power but lowlegitimacy quadrant. Over time, both the UK House of Lords and the Australian Senate have changed their status. The Australian Senate remains a powerful chamber with very similar powers to those of the Canadian Senate, but it has moved steadily from low to high legitimacy since the loss of government control of the chamber after the adoption of proportional representation (PR) for the election of senators in 1948. The House of Lords originally had powers similar to those of the House of Commons, but its power to veto legislation was removed in 1911 and its power to delay legislation further reduced in 1949. Over the past 30 years, however, its legitimacy has grown as a consequence of changes to its composition and its willingness to challenge the government on measures facing widespread public opposition. The diagram may overstate the degree of change for the House of Lords and the Australian Senate and the extent of public acceptance of their legitimacy, but there is no question that change has been in the directions indicated. For both institutions, there is strong backing for their continuing independent action in legislative review and, in the case of the Australian Senate, its powers of inquiry.

Figure 1 suggests that there may be only one naturally stable position for an upper house: the highpower, high-legitimacy quadrant. High power implies that the chamber has strong constitutional backing, reinforced by public acceptance of the value of its role. These two attributes are congruent and likely to be self-reinforcing. The low-power, low-legitimacy quadrant may look stable, but it implies an institution whose future invites substantial reform or abolition.8

The quadrants with mixed values are more interesting because they indicate where tensions over second chambers are likely to arise. The low-power, high-legitimacy quadrant implies an institution which has gained wide public support through some combination of the way its membership is selected and the manner in which it carries out its functions. Its lack of constitutional clout means that it is dependent on its high legitimacy to remain an effective influence on the governmental process. To push the analogy, it is in unstable equilibrium — it can maintain its influence only by successfully maintaining its legitimacy. Working in its favour, however, is that a disagreement with the government and/or the lower house will be seen as a political conflict, not a constitutional challenge. Its lack of constitutional power means that the government may disagree with the policies of the upper house, but does not see disagreements as threatening the system of government. The upper house is a nuisance but its actions cannot prompt a constitutional crisis.

The position of the current Canadian Senate — high power with low legitimacy — can be seen as the most unstable quadrant. The upper house poses a continuing source of uncertainty to the government because of the chamber’s constitutional powers, but its low legitimacy reduces the chances that this will be a problem for the executive. But the executive is always apprehensive that a change in partisan alignment in the upper house, the emergence of an issue which places the upper house on the side of public opinion in its dealings with the government or pressures for constitutional change will give the upper house the political backing to match its formal powers. Moreover, a clash between the government and a high-power but lowlegitimacy upper house bent on blocking government legislation is not just a nuisance but can rapidly become a serious constitutional standoff.

The Australian Senate

The two non-Canadian cases show how this instability can be resolved or at least substantially reduced. The Australian Senate moved from the high-power but low-legitimacy position of the Canadian Senate to the high-legitimacy quadrant as a consequence of the government’s losing its partisan majority in the chamber and the balance of power being held by minor parties and independents. These minor party groups had an interest in fostering a brokerage role for the Senate in dealing with government legislation and using its powers of scrutiny to, as one minor party slogan had it, “keep the bastards honest” (Warhurst 1997).

Direct popular election had not, by itself, given the Senate high legitimacy until there were major changes in the chamber’s composition and, as a consequence, its mode of operation. This is of relevance to the Canadian case, since it demonstrates that direct election of an upper house does not automatically generate wide public acceptance of its role. The electoral systems used by the Australian Senate from its creation in 1901 — plurality voting in the states as multi-member districts until 1918, then the alternative vote until the 1949 election — had the effect of creating wild swings in the partisan character of the chamber, sometimes producing a Senate with more than 90 percent of the seats held by one party grouping.9 This usually meant that the government of the day had a comfortable majority in the chamber, but, on occasion, a government could face a Senate controlled by an opposition determined to be as obstructive as possible. The manifestly partisan character of the chamber, combined with candidate selection based on party loyalty rather than on ability or ambition and the general lack of parliamentary activity and initiative left the impression that the chamber had little constructive role to play. The move to adopt proportional representation (by the single transferable vote) in 1949 and a split in the Australian Labor Party in the mid-1950s triggered a series of developments which led to minor parties and independent senators regularly holding the balance of power in the chamber, and a huge change in the scope and consequence of Senate parliamentary activities (Sharman 1999a).

The result has not reduced the suspicion and resentment of governments towards the power and influence of the Senate; the memory of its exceptional, dramatic and highly partisan action in forcing a government to the polls in 1975 has not faded.10 But there is now sufficient public acceptance of its role for the Senate to be seen as an important independent player in the parliamentary process (Mulgan 1996). The crisis of 1975 certainly demonstrated the Senate’s power, but it was also an important factor in prompting the rise of minor parties that had no wish to see the Senate used merely as the tool of opposition parties in the lower house. The fact that minor parties have held the balance of power for most of the period since 1975 has been the major factor in legitimating the role of the Senate (Sharman 1999a). During a recent period when the government managed to regain a partisan majority in the Senate, there was a good deal of newspaper commentary to the effect that the loss of the Senate’s ability to moderate government legislation represented a substantial loss to the parliamentary process and that this worked against both the public interest and the long-term interest of the government.

The Australian Senate has achieved legitimacy through a sequence of events which can be summarized as:

• an electoral system change which reduced the dominance of large party groupings;
• the election of minor party and independent senators who held the balance of power;
• the focus of minor party and independent senators on the legislative role of the Senate and its committee work inquiring into government activities and matters of public interest;

• the growing willingness of the Senate to use its power to amend and, if necessary, reject government legislation;

• the change in the recruitment of candidates to favour able and politically well connected senators; and

• the evolution of protocols between governments and minor party senators to regularize consultation over the introduction of legislation.

This last characteristic is important because it shows that governments have begun to see dealing with a Senate over which they do not have partisan control as something which represents normal, rather than abnormal, politics. Consultation between minor party senators and the government has been aided by the selection process for Senate candidates, which has ensured that some of the abler government ministers have been senators, providing opportunities for both formal and informal negotiations. This routinization of consultation is one of the strongest indications of the acceptance of an independent Senate as a legitimate component of the Australian parliamentary process.

The House of Lords

The House of Lords has had a more complicated history but, in its most recent phase, has much in common with the Australian experience. The loss of its veto power over legislation in 1911 and 1949 appeared to set the Lords on a trajectory to eventual abolition. As a picturesque feudal anachronism, it appeared to have little in common with the majoritarian, executive-dominated style of politics which had come to typify UK politics. Its place in the lowpower and low-legitimacy quadrant of figure 1 made it a prime candidate for having its power further reduced or being reconstituted as some advisory body or disestablished entirely.

Most of the many accounts of why this did not happen are based on divisions within the Labour Party when in government about the rival claims of reform and abolition (see Dorey 2006). The most graphic example was the Wilson Labour government’s 1969 bill to phase out the hereditary component of the Lords, which was defeated by a coalition of Conservative opposition members who resisted change and rebel Labour backbenchers who wanted more drastic change. This embarrassment discouraged further attempts at reform — the Lords had already ceased to be a serious obstacle to government legislative programs — and the pressure for change receded. But the increasing malaise about what is now called the “democratic deficit” in UK politics, the dissatisfaction with the arcane nature of the UK constitutional structure and the rising concern with individual rights all led to a resurgence of a broad interest in institutional reform, culminating in the Blair Labour government’s commitment to reform the Lords when it won office in 1997.

This had been preceded by several developments which had enabled the House of Lords to increase its profile and enhance its reputation. The Life Peerages Act, 1958, passed by the Conservative Macmillan government, in addition to permitting members to be appointed to the Lords without increasing the number of hereditary peers, made membership open to women for the first time. The longer-term significance of this change was the ability of governments to appoint members for their political contributions to the chamber and to the government’s image without getting embroiled in the broader social issue of hereditary peerages. There is no limit to the number of life peerages, with more than a thousand having been created since 1958 — more than 350 during Prime Minister Blair’s term of office alone. This gives the prime minister great power to shape the composition of the Lords, but the power has been used in a way which has increased both the political and the social diversity of the chamber’s membership.

This process has continued since the Blair government’s 1999 reform. The House of Lords Act removed all but 92 hereditary peers and set up a House of Lords Advisory Commission, which now vets government nominees for life peerages for their propriety and can also advise the prime minister on nonpartisan appointments.11 A fixed number of opportunities for the appointment of these independent peers is allocated by the government, and the Advisory Commission solicits suggestions from the public; the prime minister retains the final say. By April 2008, some 47 of these independently selected life peers had been appointed, further contributing to the partisan diversity of the chamber. These developments have reinforced the status of the Lords and made it more difficult for the government to brush aside opposition from the chamber. Despite its nonrepresentative and unelected nature, this has enabled the Lords to claim that it has a legitimate function to challenge the government in the parliamentary process and for this claim to be broadly accepted.

This is not the case in Canada, and it is the dynamics of the Canadian Senate and the procedures for the appointment of senators to which we should now turn our attention.

The Prime Minister and an Appointed Senate

In suitably arcane language, the Constitution Act, 1867 specifies that “The Governor General shall from Time to Time, in the Queen’s Name, by Instrument under the Great Seal of Canada, summon qualified Persons to the Senate” (section 24). By customary practice, sanctified by the minutes of a meeting of the committee of the Privy Council in 1920, the recommendation to the governor general for the appointment of senators is the “special prerogative of the Prime Minister” (Dawson 1933, 125). Membership of the Senate is thus the gift of the prime minister alone.

The exercise of this power has few substantive constraints but a great many procedural ones. Section 22 of the Constitution Act specifies that 24 senators are to be appointed from each of the four regions (divisions) into which the act divides Canada: Ontario, Quebec, the Maritime provinces and the Western provinces. There are additional specifications for individual Maritime and Western provinces, and provisions for six senators to be appointed from Newfoundland and Labrador and one each from the three territories, making a total of 105. In certain circumstances (sections 26-28), the governor general may appoint four or eight additional senators,12 one or two each from the four divisions set out in section 22, to create a Senate with a maximum complement of 113 members. This procedure for the appointment of supplementary senators has been used only once, when the Mulroney government appointed eight additional senators in 1990 to enable passage of legislation establishing a national goods and services tax over the vigorous objections of the opposition Liberals (for a summary of these events, see Franks 2003, 161-3).

Apart from these regional requirements — and additional rules for Quebec senators to be chosen from electoral divisions within the province — there are no special qualifications for becoming a senator except a minimum age of 30 and the need to fulfill citizenship, residence, solvency and property requirements which, at $4,000, may have been high in 1867, but would not be hard for most current Canadian citizens to meet (section 23).

The major constraint on the prime minister’s discretion to appoint senators is the limit on the size of the Senate. If a government inherits a Senate with no vacancies, there is no opportunity to appoint new senators; if the Senate has a large partisan majority hostile to an incoming government, there is no way to change its complexion. This limit is critical for understanding the dynamics of the operation of the Senate and the attitude of governments to Senate reform. If, for example, there were no limit on the size of the Senate, a government could, as an extreme measure, swamp a hostile Senate majority by appointing new senators sympathetic to it. Appointing new lords — both before and after the creation of life peerages — has always been an option for UK governments faced with a recalcitrant House of Lords; in Australia, during the time that membership was by appointment, neither the New South Wales nor the Queensland legislative council had constitutional limits on the number of members.13

Its fixed size intensifies the partisan edge to the attitude of governments to the Senate, and there is no recourse to the large-scale creation of senators beyond the eight provided for in section 26 of the Constitution Act, 1867. Without these limits, the Senate would still be important and would remain a significant actor in the legislative and parliamentary process, but its existence would not involve the fear that it might block key government legislation without redress for the government; in such case, the government could simply appoint enough new senators to give it a majority. But, as presently constituted, the Senate retains the power to wreak havoc on a government’s legislative program and to use its powers of investigation and inquiry to undermine government policies.

Of course, one person’s unprincipled Senate obstruction is another person’s necessary and effective parliamentary scrutiny of government and its legislation. The Mulroney Conservative government, particularly from 1985 to 1990, faced repeated challenges from a Liberal majority in the Senate. While acknowledging the partisan battle in the Senate during this period and the subsequent years to 1993, a respected scholar of the Canadian parliamentary process has argued that the Senate operated on many occasions as a chamber of sober second thought (Franks 2003, 165). This, however, does not make Senate challenges any less threatening in the eyes of governments.

These apprehensions aside, for the prime minister, the power over appointments to the Senate generates large benefits. The most obvious are patronage and partisan gain: the ability to reward those who have assisted the government and the prime minister, to signal government concern with a particular issue, group or locality, to provide a consolation prize for those who have missed out on other government positions and to enhance the government’s appeal by appointing a few symbolic third-party or nonpartisan senators. All of these appointments are likely to contribute to the ranks of those who support the party of the prime minister who made them senators. If these appointments build or maintain a partisan majority, the government can feel secure that the Senate is unlikely to cause it any serious problems.

Prime ministerial appointment has another, less obvious benefit to the government of the day. No government relishes the existence of a powerful and independent actor in the parliamentary process; anything which limits the significance of the Senate is to be welcomed. The arbitrary, partisan and personal nature of the selection of senators by the prime minister reduces the legitimacy of the chamber, no matter how illustrious the appointments. This helps to discredit the work of the Senate and undermines its committee inquiries when they oppose government policies. Governments supported by a House of Commons majority can always stress their popular mandate and the corresponding weakness of a Senate appointed on the personal recommendation of previous prime ministers.

Prime ministers who are able to create or maintain a partisan majority and leave office with few vacancies in the Senate can take satisfaction in another partisan bonus: the creation of a poison pill for an incoming government of a rival partisan colour. The larger the Senate majority, the greater the difficulty for the new government and the greater the opportunity for the former governing party, now in opposition, to use the Senate to embarrass the government.

The exclusive power which resides with the prime minister to recommend appointments to the Senate also provides the opportunity not to appoint senators. This may seem a perverse benefit, but it has been used by the Harper government to orchestrate a campaign to support legislation for Senate reform. At the end of April 2008, there were 14 unfilled Senate vacancies but 60 of the continuing senators were Liberals. It may be that, if the number of Liberal senators drops below 53 — the number required to control the Senate — while the Harper government remains in office, the prime minister’s resolve may weaken and he will appoint enough Conservative sympathizers to give the government a Senate majority, notwithstanding the lack of progress on a government bill providing for provincial elections to advise the selection of Senate nominees (Bill C-20, the Senate Appointments Consultation Act). And, should his government face imminent defeat, there is always the possibility, however distant and at odds with his political commitment to Senate reform, that the prime minister may choose to fill all Senate vacancies before he leaves office to reward his supporters and, in time-honoured fashion, to deny patronage to his successor.

The prime minister’s power over the appointment of senators is the source of the Senate’s greatest lack of legitimacy and the greatest barrier to Senate reform. But to remove the prime minister’s control over the composition of the Senate is not just to reduce the opportunities for patronage, but to threaten the government’s familiar role in the parliamentary process. Under the current system, the Senate’s power is effectively constrained by the partisan appointment process. This clearly works to the benefit of governments with majorities in the Senate, but even a government which faces a hostile Senate can use the chamber’s lack of legitimacy to denigrate its resistance and to impugn the opposition for using an undemocratic and unrepresentative institution to thwart the will of the majority of a popularly elected house. Nothing indicates this more forcefully than recent legislation nominally aimed at Senate reform.

Senate Reform as a Tactical Device

The bills introduced by the Harper government in 2006 and 2007 are examples of the prime minister’s power to set the terms of political debate, the tactical goals to which Senate reform can be put and the dominance of partisan concerns in the parliamentary process. Two bills affecting the composition of the Senate were before Parliament in 2008: C-19, An Act to Amend the Constitution Act, 1867 (Senate Tenure);14 and C-20, the Senate Appointment Consultations Act.15 The first of these bills proposed the introduction of eight-year nonrenewable terms for senators in place of the current appointment for life until retirement at age 75. The summary of the bill published by the Parliamentary Information and Research Service (Canada 2007a) included extensive commentary on the bill and rehearsed its previous introduction into the Senate in 2006, its treatment before a Senate committee and the Senate’s decision not to proceed with it. It also included justification for the bill and an opinion that changing the term of senators did not fall within the class of subjects requiring constitutional change but could be altered by legislation of Parliament acting alone.

Bill C-20 was much more adventurous. It maintained the current monopoly of prime ministerial nominations for Senate appointments, but provided for “consultations” which would permit the prime minister to recommend candidates who had won electoral contests in each of the provincial jurisdictions for which there were vacancies. The bill provided details about the way these elections would be run: they would achieve proportional representation using the single transferable vote, they could be held at the same time as provincial general elections and the components of electoral administration might be delegated to provincial administrative bodies. Again, the legislative summary provided details, explanations and justification, and suggested that, “In future years, it may be that an informal practice of appointing senators from a list of selected nominees will transform itself into a constitutional convention that would ‘constrain’ the prime minister in making his or her choice for Senate appointments” (Canada 2007b, 19).

At first glance, these bills could be seen as a legislative scheme to create an elected Senate without making formal changes to the Constitution. This route had been flagged by the Progressive Conservative government of Alberta when it held an election in the province in 1989 for a nominee to fill a Senate vacancy. The winner, Stan Waters, was subsequently appointed by the Mulroney Conservative government in 1990. When elections for Senate nominees were held again in Alberta in 1998, Liberal prime minister Jean Chrétien declined to appoint such “senators in waiting” (for background and commentary, see Smith 2003a, 103-4). The experiment was resumed in Alberta with Senate nominee elections in 2004;16 upon taking office in 2006, Harper, as a Conservative prime minister and former member of the Alberta wing of the party committed to extensive Senate reform, appointed the winner, Bert Brown, to the Senate. As soon as the prime minister indicated that he would appoint to the Senate only nominees who had been chosen by provincial elections, the grand plan was revealed: a legislative scheme backed by executive discretion to force the provinces to hold elections for Senate nominees, on pain of eventually losing their representation in the Senate (Flanagan 2007).17 The only exception to this policy to date has been the appointment of Michel Fortier to both the Senate and cabinet in 2006 to enhance the latter’s representation from Quebec and to give Montreal a senior representative in the government.

An alternative explanation of the Harper government’s actions is that this scheme had little to do with Senate reform but was simply a way of embarrassing the Liberal majority in the Senate, further reducing the legitimacy of the Senate while placating those in the Conservative party who wanted Senate reform. Evidence for this is not hard to find. There was no suggestion that, if the scheme were fully implemented, the Harper government would relish an upper house composed of senators with an added sense of their own legitimacy, whose partisan loyalties could be very different from those of the government of the day. The prime minister had even mused about the possibility of abolishing the Senate. The scheme made a lot more sense as a device to confuse and discredit the Liberals as an opposition party which controlled a majority in the Senate. For a government without a partisan majority either on the floor of the House or in the Senate, proposals for Senate reform that involved limited terms and an electoral component would challenge the Liberal Party’s democratic credentials and, if rejected, help to undermine Senate opposition to government legislation. Senate reform of this kind could be seen as designed explicitly to weaken one of the strongest weapons of the opposition Liberal Party, which was its control of the Senate. The demonstration of the Senate’s reluctance to sanction changes to its current partisan composition — amply demonstrated by the hostile treatment of both bills by the opposition in the House and the Liberal majority in the Senate — would further reduce the chamber’s standing in the eyes of the public.

There is also the possibility that the Harper government is playing a very long game. If several provinces agreed to hold Senate nominee elections and it began to look as though there would be sufficient partisan representation from these elections to affect the partisan balance in the Senate, alarm bells would ring in party headquarters and premiers’ offices across Canada. This may provide the background for concerted moves for substantial constitutional reform of the Senate to reduce its powers and, perhaps, to modify its composition. But this is a situation which requires time and many preconditions over which the national government has little control.

Notwithstanding these speculations, the most interesting long-term consequence of this scheme for Senate reform was not its rejection by Parliament — its short-term partisan success can be seen as dependent on its nonacceptance by the Liberals — but how it indicated the ways in which the existing appointment procedure for senators might be modified by a combination of new statutory machinery and the willingness of the prime minister to vary the customary appointment practices.

Modifying an Appointed Upper House: The Australian Experience

In Australia, bicameralism has always been tied more closely to the politics of resisting radical action by governments than it has in Canada. The early extension of the franchise for legislative assemblies in the Australian colonies created apprehension among local elites and in the Colonial Office in London that governments supported by adventurous lower house majorities might use their power to threaten propertied and commercial interests. As a consequence, the arrival of responsible government for the Australian colonies after 1856 was linked to the continuation of powerful legislative councils as upper houses in new bicameral parliaments. But there was debate over how these upper houses should best be constituted.

Two models emerged. The more popular method was to ensure that elected legislative councils would maintain a conservative bias by having a franchise based on substantial property qualifications. In this way, the veto power of the upper house on all legislation, including financial legislation, would be given the legitimacy of popular election, albeit on a restricted franchise. The rival model, initially adopted by New South Wales in 1856 and then by Queensland in 1859, was to have an appointed legislative council similar in composition to the Canadian Senate but with no limit on the number of councillors as long as the governor was willing to appoint them on the recommendation of the premier. The adoption of this system had much to do with the preconceptions of colonial politicians and the lack of familiarity with responsible government in New South Wales during the 1850s (Clune and Griffith 2006, 65-75).

Over the following 70 years, issues emerged which put governments supported by a majority in the New South Wales lower house, the Legislative Assembly, at odds with a majority in the appointed Legislative Council. Two consequences followed. The first was perennial debates of varying intensity over the legitimacy of a nonelected body thwarting the popularly elected government and demands for reform or abolition of the council.

The second consequence was the politicization of the governor’s role in appointing new members of the Legislative Council. When a new government took office and found that its legislation was being blocked by a hostile upper house, premiers could go to the governor and request the appointment of enough new upper house members to give the new government majority support in the council. Governors were reluctant to do this — blanket acceptance would reduce the role of the council to a cipher and destroy any claim to its being a house of review. But ignoring the wishes of a popular government and rejecting the advice of a premier was politically risky. With varying degrees of skill, governors aimed to restrict any increase in the size of the council unless the subject of the conflict with the council had been an explicit part of the new government’s electoral mandate. This was especially the case where contentious legislation concerned the amendment of the council’s composition and powers.

The rise of the Australian Labor Party in the 1890s and the increasing frequency of Labor Party majority governments across Australia after 1900 intensified debate over the powers of legislative councils, especially the undemocratic nature of appointed councils. The Labor Party was committed to abolition and, in Queensland, the Theodore Labor government, after considerable political turmoil and manoeuvring, contrived to abolish that state’s Legislative Council in 1922 (McMinn 1979, 150).

New South Wales proved to be a more intractable case. Labor premier Lang had tried on several occasions during the 1920s and 1930s to abolish the Legislative Council, but was finally thwarted by his dismissal in 1932 as premier by the governor in the turmoil of Depression politics. But the removal of Lang and the installation of a conservative government did not remove the problem of the council — there remained wide agreement that it needed to be reformed and its composition determined by some method other than appointment by the governor on the recommendation of the premier.

Many options were canvassed, but the one chosen to operate from 1934 until 1978 was a halfway house between modification of the process of appointment and direct election. Membership of the council was set at 60 (the lower house, the Legislative Assembly, had 90 members in 1934). Each member served a twelveyear term, and a quarter of them retired every three years. These 15 vacancies were filled by indirect election — not by the public, but by the 90 members of the Legislative Assembly and the 45 nonretiring members of the council. A system of proportional representation by the single transferable vote was used for the elections. The combined effect of staggered threeyear terms and proportional representation meant that the partisan composition of the Legislative Council changed only slowly, reflecting changes in the party composition of both the lower and upper houses. This approach had short-term benefits for the conservative party in power in 1934, but it also expressed the belief that the council should be a stabilizing factor in the political process (Turner 1969; Clune and Griffith 2006, 320-51).

The success of this system was mixed. The reforms removed the problem of appointments and, by taking away the Legislative Council’s power to veto financial bills, reduced its potential to precipitate a constitutional crisis. The ability of a future government to amend the composition or powers of the council was constrained by the requirement that the legislation gain popular endorsement through a referendum. The problem of the council’s legitimacy remained, however, because of both its lack of direct representation and, of greater importance, its domination by government-controlled majorities or, occasionally, majorities controlled by the opposition. In both circumstances, party discipline and institutional inertia greatly reduced the standing of the chamber in much the same way as the Australian Senate had been until its rejuvenation by a change to the electoral system and the enlivening effect of minor parties on the role of the chamber.

In 1978, the New South Wales Legislative Council became a directly elected chamber chosen by proportional representation,18 a model pioneered for legislative councils by South Australia in 1975 and now adopted, with variations, by four of the five remaining legislative councils in Australia.19 These chambers are now valued for their representativeness and their often-constructive contribution to the parliamentary process (Stone 2002). These transitions, all of which were undertaken by Labor Party governments, resulted from the realization that abolition was unlikely to gain popular support and the belief that an upper house with the balance of power held by minor parties was likely to be of more long-term advantage to Labor governments than one which might be controlled by a belligerent opposition.

But the period of indirect election was a useful experiment. It probably made the transition to a directly elected chamber more acceptable by demonstrating to governments that indirect election would, in the longer term, do little to enhance the public acceptance of the institution. And an indirect election which produced a chamber controlled by the governing or the opposition party could create as many difficulties as one chosen by a more democratic system. It is interesting to speculate about the changes that could have given the indirectly elected council a more diverse partisan composition and one that neither of the major party groupings controlled. One solution would have been to give an independent appointing authority the power to choose a small number of nonpartisan and third-party members. This approach was unthinkable in the highly charged partisan climate of the early 1930s, but the example of the House of Lords Appointments Commission makes it more of a possibility today.

Propositions for Successful Senate Reform

From the discussion and surveys of the experience of similar UK-derived parliamentary systems, it is now possible to set up a series of propositions about the direction for successful Senate reform. As this paper has argued, there must be congruence among the assumed purpose of the Senate, the goals for reform and the methods through which reform is to be achieved. Smith (2003b) uses much the same procedure, although his “principles” are focused more on protecting the dominant position of the House of Commons than are the propositions listed below.

The purpose of the Senate is effective scrutiny

The greatest contribution of an upper house to a contemporary parliament is scrutiny of legislation and inquiries into the conduct of the executive and its agencies. The effective discharge of this function is the foundation on which the current reputations of the Australian Senate and the House of Lords have been built. Effectiveness requires two components: a constitutional one, that the chamber has sufficient formal power to amend and delay legislation; and a political one, that the chamber is not controlled by a majority of members which acts on the instructions of the governing party in the lower house. Both of these components need elaboration.

Governments are not persuaded by sweet reasonableness; effectiveness requires the upper house to have enough formal power to force the government to consider the cost of overriding the upper chamber’s decisions, even if this is possible. Substantial delay of at least a year might be the minimum requirement if delay is the only option — it is clear from the experience of the House of Lords that the two-year delay on nonfinancial legislation introduced in 1911 was sufficient to act as a substantial check on the pursuit by government of controversial legislation (Dorey 2006). In the Canadian case, the Senate already has power to reject legislation without any override from the House of Commons, but the question of delay may become important if the Senate’s powers were part of a reform package. Whatever modifications there might be to the legislative powers of the Senate, its privileges and its power to inquire into the actions of government should not be reduced. Parliamentary upper houses already have political limits on the extent to which they can use their powers to investigate the executive; any reduction of the Senate’s formal powers over inquiries would further undermine its ability to compel witnesses and to extract information from governments.

The partisan requirement for effectiveness in an upper house is more elusive. It is clear what is not effective: government-controlled majorities are unlikely to challenge legislation in ways which act as an effective check, and are even less likely to pursue the investigation of executive actions which may embarrass the government. But government-controlled majorities are not always the same thing as upper house majorities of the same partisan colour as the government. Prime Minister Thatcher, for example, found the House of Lords a persistent and effective critic of her government’s legislation, notwithstanding the nominal majority of Conservatives in the Lords (Baldwin 1999, 42-4). If party members in the upper house have a degree of autonomy in the organization of an upper house caucus and the habit of making their own decisions about the conduct of business in the chamber, partisanship may be moderated by other considerations in both the scrutiny of legislation and committee proceedings. There have certainly been occasions when members of the governing party in the upper houses of Australian parliaments have resented — and occasionally resisted — attempts at government manipulation of their activities, even when the government had a majority in the upper house. As Norton (2003) points out for the House of Lords since 1999, if the government has few carrots or sticks to enforce discipline, party allegiance is a matter of temperament rather than necessity (note also Shell 1992, 64-98). But it is unclear whether such a view of party would operate among members of the governing party without structural change in the selection procedure for members of the upper house and the chamber’s acquisition of an established reputation as an independent actor in the parliamentary process.

Control of a second chamber by the largest opposition party in the lower house may not be much of an improvement over control by the governing party. Opposition control makes it difficult for the party leader in the lower house to resist the temptation to see the upper house as no more than a component of strategies to win the next election. The experience of Australian upper houses in this position is that the chambers’ review function is subordinated to headlinegrabbing attacks on the government and to using the inquiry process to investigate specific government misdeeds rather than to generate information for long-term reform. Partisan, short-term and sporadic best describe the direction of parliamentary activity in upper houses controlled by an opposition party majority.

As both the House of Lords and the Australian Senate have shown, the ability of minor party and independent members to hold the balance of power in the chamber is the key to the development of an effective review function (Sharman 1999b). This is not because they are more virtuous but because they are interested in legislative politics. The payoff for such members is not gaining executive office and the perks of government but demonstrating to the public the utility of their role in forcing the government to negotiate on controversial legislation and to disclose information on matters of public interest. This does not prevent grandstanding and partisan attempts to claim media attention, but it means that these members have a long-term interest in establishing and maintaining structures for parliamentary scrutiny of legislation and executive action. In particular, they stand to gain from a well-developed and wellresourced committee system.

The House of Lords and the Australian Senate also demonstrate the differing ways in which minor parties can affect the dynamics of an appointed chamber as opposed to an elected one. In the Lords, minor party members add further political diversity to an institution without strong party discipline for the large parties (Baldwin 1999, 32-44; Norton 2003). In such a context, whether the minor party members hold the numerical balance of power is not critical
for the discharge of an effective review function. This is not the case in the Australian Senate, where strong party discipline in the governing and major opposition parties and dependence on the goodwill of state party machines for their endorsement at elections have made senators very sensitive to party direction. It is only when minor party and independent senators hold the balance of power that the Senate uses its review powers effectively.20 Given the nature of the system of proportional representation used to elect the Australian Senate and the continuation of the trend for 20 percent of the electorate to favour Senate candidates from other than the two largest party groups, the chances are that minor party and independent senators will usually hold the balance of power in the chamber.

Partisan balance of some kind is as much a critical component of the process of effective parliamentary scrutiny as the powers of the chamber. Only by continuing to use its reviewing power is a second chamber able to find a place in the governmental process which is broadly accepted as legitimate by both the public and the government of the day.

Direct election of senators is not a plausible strategy for reform of the Senate

The proposition that direct election of senators is not a plausible strategy for Senate reform is shorthand for a number of assumptions about the role of the Senate. While direct election greatly enhances a chamber’s “input legitimacy” (Kelso 2006), it raises many difficulties in the Canadian context. The most serious is that Canada does not have a tradition of strong elective bicameralism involving the acceptance that the dominance of a government with a majority in the lower house may be constrained by a powerful upper house that claims an equal mandate. It has taken the Australian political system the better part of 150 years to come, grudgingly, to an acceptance of the benefits of such an arrangement for representative democracy (Stone 2002). From one point of view, this is a pity, because it denies Canada the benefits of the symmetry of a powerful upper house matched with a democratic legitimacy that cannot be acquired by any other means.

But this is not the Canadian way, and there are many other reasons a directly elected upper house is not appropriate. High on the list is the difficulty in determining what an elected Senate would represent. As previously discussed, party division would certainly be the main organizing principle for the vote, but an electoral system structured by province — whatever the electoral system used — would produce a pattern of party representation that would almost certainly differ from that of the lower house. Whether this would produce a Senate that represented regional interests or simply transformed the representation of existing parties is hard to know, but the likelihood of intense partisan discord seems likely. This is not a good basis for effective Senate review as set out in proposition 1.

There are many pragmatic considerations which do not make a directly elected Senate a viable option. No government wants a powerful upper house underpinned by a popular mandate. It is true that the Harper government has introduced legislation which looks as though it is happy to accept that possibility, but, as discussed above, the proposals embodied in Bills C-19 and C-20 appear to be tactical ploys to embarrass the Liberal Party and placate some members of the Conservative Party rather than considered attempts at reform. Even if the proposals were intended as catalysts for long-term constitutional change, the changes required to implement an elected Senate would be substantial, complicated and likely to generate strong provincial opposition, particularly if the amendments were to alter the weighting of provincial representation in the Senate. Provincial governments — premiers, in particular — would not be enthusiastic about a group of provincial senators who could claim to speak as representatives of provincial interests in Ottawa.

In sum, the difficulties, both of principle and practice, of designing a directly elected Senate are overwhelming. Moreover, it is an unnecessarily fundamental and contentious change to achieve the specific goal of a Senate that undertakes effective scrutiny of government legislation and activities.

Senate reform can best be achieved through amending the appointment process

In the cases of the United Kingdom and the two Australian state upper houses discussed above, appointment to the upper house follows the same procedure as that for the Canadian Senate: appointment is by the monarch on the recommendation of the government of the day in the person of the prime minister or premier. But the circumstances and trajectory of the use of the appointment procedure have differed widely among the chambers. For the House of Lords, outside the process of ennoblement, the appointment of life peers is only 50 years old and the current mixed procedure of appointment — where most life peers are nominated by the government but a small number are nominated by the House of Lords Appointments Commission — has been in place for less than ten years. The Australian experience of appointed upper houses was terminated either by the abolition of the chamber (Queensland in 1922) or by the rejection of appointment as a suitable method of selecting members (New South Wales in 1934). Canada has been unusual in the persistence of the appointment process since the original design of the Senate in 1867.

There have been good reasons for this, chief among them the fixed size of the Senate and the specification of regional quotas — none of the other appointed chambers reviewed have had such constraints. The regional components are not only constitutionally enshrined and politically sensitive; they play a role in calculating the minimum number of seats in the House of Commons allocated to some provinces.21 Altering the formula requires broad agreement and constitutional change — which the experience of the Charlottetown Accord of 1992 shows is very difficult to achieve.22 This means that the first question to be asked about amending the appointment procedure is whether the size and regional constraints on Senate membership are to be maintained. If the answer is yes, change must fit within the constraints of the current composition of the Senate. This does not mean that constitutional change is precluded — altering the formal process of appointment, for example, is likely to require such a change — but that arguments over provincial representation would be less central to the debate. If the answer is no, reform of the Senate is likely to be part of a broad-based constitutional debate about the institutions of the national government in a federation. Again, if the goal of the Senate reform is to achieve a chamber with effective reviewing power, inviting wholesale constitutional change is an unnecessarily contentious method.

What choices are on the table if the current Senate of 105 members is maintained with the existing pattern of regional apportionment? If we accept that partisan balance of some kind is to be the goal of the appointment process, two methods used by the House of Lords are available. One is an independent appointments commission to select nominees; this could apply to all appointments or, as with the Lords, some proportion, fixed or variable. The other option is partisan nominations derived directly from the parties represented in the House of Commons or through a nominating committee or an appointments commission. A critical question is whether there should be an attempt to insert an element of regional consultation in the appointment process, by an appointments commission or by a process involving the parties or by some other procedure (see Stilborn 2003).

There is a further option, that of indirect election. In the institutions examined in this paper, only the New South Wales Legislative Council between 1934 and 1978 employed this method as a substitute for an appointed house rather than as a supplement to it. For the Canadian Senate, several suggestions have involved indirect election, most notably the Trudeau government’s “House of the Federation” proposal set out in Bill C-60 in 1978 and the reconstitution of the Senate envisaged by the Charlottetown Accord in 1992.23 Both proposals considered legislatures as agencies for selecting senators: Bill C-60 provided for half the Senate to be chosen by the members of the House of Commons and half by provincial legislatures; initially, the Charlottetown Accord proposed that senators be chosen by direct election, but its final form included a provision for senators to be selected by provincial legislatures rather than direct election if the provincial government so chose (Stilborn 2003, 35).

The difference between these two proposals reflects different perspectives on the kind of partisan balance to be achieved by indirect election. Bill C-60 was intended to provide a balance between national and provincial partisanship, reinforced by its choice of proportional representation as the electoral formula, while the Charlottetown formula made the Senate a manifestation of provincial partisanship with no specification of the electoral system to be used for either direct or indirect elections. The Charlottetown Senate might have produced a measure of partisan balance in the chamber, but this was not a feature of its design.

Both the Bill C-60 and Charlottetown proposals intended to use indirect elections to involve provincial electorates in the selection of senators, but indirect elections can be used for other purposes. If the goal is to achieve a Senate with a partisan balance which tracks, even if slowly, the pattern of representation in the House of Commons, then members of the House can be the electorate, with elections based on proportional representation, as in New South Wales. Under one version of such a scheme, Senate vacancies would not be filled until there were three or more vacancies. Once this number had been reached, an election would be held by secret ballot with each party represented in the House of Commons nominating a candidate for each vacancy. A government with a majority in the House would likely gain two of the three vacancies, with the third gained by the largest opposition party. Over time, the partisan balance in the Senate would change to respond to changing party fortunes in House of Commons elections. The Senate could, of course, be made more balanced and responsive to other parties by aggregating a larger number of vacancies — the more vacancies to be filled, the more proportional the result. Such a system could be used to fill all vacancies or to work together with an independent appointments commission which could, for example, be responsible for filling every fifth vacancy. Such a mixed system of indirect election could work with the existing terms of senators, or with fixed terms as long as they were at least eight years in duration and staggered so as to have a continuing series of appointments.

In the absence of a change to the Constitution, indirectly elected senators would still have to be recommended formally for appointment by the prime minister. It is possible that the actors might broadly agree on the desirability of such an amendment, but it is equally possible that regional or partisan issues would make such a change as controversial as previous proposals have been. Whether constitutional change is a component or not, there are many variations possible on the themes of appointment and indirect election, but the goal is the same: to ensure the maintenance of a rough partisan balance in the Senate; only with such a balance can the Senate pursue its review function effectively.

There remains the question of the number of senators, if any, that the prime minister should be able to recommend without the use of any additional procedure. This remains the source of most new appointments to the House of Lords, but there is no limit to the membership of the Lords. In contrast, the Canadian Senate’s fixed number of members combines different modes of selecting members and few vacancies. Generating batches of, say, 21 vacancies — a fifth of the Senate — with each batch to be filled by two or three appointment procedures, requires a substantial change to the terms of senators and a complicated formula for the transition period. But this might be required if the prime minister were anxious to retain some personal discretion to choose senators.24

The prime minister needs to be persuaded of the benefits of reform

The prime minister’s unfettered power over the appointment of senators, and the way it has been used for partisan purposes, has been the most important factor in undermining the legitimacy of the Senate. But that power also generates many benefits for the prime minister and the governing party. Executive officers do not give up power voluntarily unless there is a clear payoff. Unfortunately, the experience of reforming the appointment process for the House of Lords suggests that the payoff must be significant. The creation of life peers for the Lords was triggered not by any commitment to the creation of balanced partisanship in the chamber, but to avoid the political embarrassment of perpetuating a hereditary aristocracy. The Labour Party was also committed to reforming the Lords as part of a larger package of constitutional reform in response to political pressure from its rank and file, even though the topic led to disputes within the caucus and government (Dorey 2006). On the other hand, favourable assessments of the reform — despite continuing political pressure for some form of direct election — have brought praise to the Labour government while still leaving it discretion over appointments.

For Canadian Senate reform, however, there appear to be few immediate payoffs for a prime minister. In fact, as the Harper government has shown, an unreformed Senate can be used to considerable political advantage. A reformed Senate could cause problems for a government because of the enhanced public support of its scrutiny of government legislation and activities. The only clear benefit of a Senate in which minor parties and independents — that is, neither the governing nor major opposition parties — held the balance of power is that it would remove the tactical advantage of denying the ability of an opposition party acting alone to use the Senate as a partisan weapon against the government. This, however, may be no small benefit. It was certainly the major motivation for Australian state governments to introduce proportional representation for the election of upper house members — even when the Labor Party has had a chance to control both chambers: denying any party the ability to control the upper house alone is strategically more valuable than having periods when control varies between the government and the opposition.

Occupying the moral high ground may also be useful for a government which introduces reform, especially if it is popular, fulfills an election promise and catches an opposition unwilling or unable to mount a strong argument against change. In the Canadian context, the prime minister will require at least some symbolic payoff, and reduction of the Senate’s power might be an acceptable one. It would also show, if supported by the Senate itself, that the Senate was serious about reform.

The power of the Senate to veto financial legislation should be removed and its power to veto other legislation should be replaced by the ability to impose substantial delay

Successful Senate reform depends on the chamber’s attaining a congruent relationship between its powers and its legitimacy. Given Canada’s parliamentary tradition, the Senate’s use of its powers to block financial legislation and precipitate a constitutional confrontation is unlikely ever to be seen as acceptable by the public. This means that the Senate can never have the legitimacy to exercise its powers fully and, if its powers are unchanged, it will remain a threat, however distant, to the government of the day. It is not sufficient to argue that the Senate’s current power to veto financial legislation would never be used because of convention. This may well be true of other sections of the Constitution Act, 1867 — such as the power under section 90 to disallow provincial legislation — but such sections do not have the potential to threaten the existence of the government, unlike the Senate’s power to block financial legislation, which could force a government to the polls. The experience of the denial of supply to the Whitlam government by the Australian Senate in 1975 and of Whitlam’s dismissal by the governor general is a reminder that events which “never” happen, occasionally do.

The loss of this power could be seen as part of a scheme to reform the Senate and set it on a trajectory to gain wider public legitimacy. Any constitutional change is difficult, but limiting the legislative power of the upper house by abolishing one of its least-used powers would seem a relatively uncontentious proposal among the provinces, unlike the vexed question of regional representation. Such a limitation of its powers might also make other Senate reforms more palatable to the executive, and could even be a prerequisite for the government to take Senate reform seriously. It is true that, as with the reform of upper houses in Australia and the House of Lords, some would see this as strengthening an institution they wish to see abolished, and would oppose such reform. But if the supporters of a reformed Senate are serious about their wish to achieve change, the attempt at constitutional amendment must be made.

The Senate must prepare the ground for reform and hope for partisan commitment

Preparing the ground and hoping for partisan commitment to reform is more a process than a model for change, but without it Senate reform will remain a topic of speculation rather than of action.25 Yet there is no way of knowing in advance what will trigger the circumstances favourable for reform, short of a major constitutional crisis. For the House of Lords, it was a process which emerged under the Labour government of the years immediately after the end of the Second World War and picked up momentum with a revolt against unpopular legislation proposed by the Thatcher government. Under Thatcher, the Lords could claim backing by public opinion in its criticism of contentious government legislation. What strengthened its position was that it was nominally a Conservative chamber complaining about the actions of a Conservative government. Similarly, the fiercest debates over proposals for reforming the Lords took place within the Labour Party. These intraparty disputes help to reduce the view of upper house reform as simply a partisan tactic to embarrass the opposition. The Harper government’s bills for Senate reform fall into this latter category. Whatever the motivation, the proposals were not the result of any bipartisan discussion, but were greeted by the opposition parties as an attack on the Senate Liberal majority rather than the basis for constructive change.

The Australian experience of changes to the composition of upper houses varies from the brutal partisan attack on the Queensland Legislative Council in 1921 (McMinn 1979, 150) to a reform achieved as an incidental consequence of electoral changes made to the Australian Senate in 1948 for other purposes (Sharman 1999a; see also Uhr 1995). More recent changes to Australian state upper houses have retained a partisan edge, but disputes have been over the design of a reformed chamber rather than the need for change (Stone 2002). The example of the increased status and effectiveness of the Australian Senate since the 1960s gave state upper houses good reason to support reform measures and to argue that such changes were good for the parliamentary system as a whole rather than simply an expression of partisan advantage.

This may be the route for the Canadian Senate. If the chamber itself can develop proposals for change which are seen as remedying a systemic defect in the Canadian parliamentary process, this could be a way to persuade the public that Senate reform is desirable, not just a partisan scheme to attack the prime minister.26 Unfortunately, a minority government is a major impediment to proposals for Senate reform. Such governments are concerned with the very short term and are already subject to a highly effective check on their activities by opposition parties in the House of Commons. The most conducive circumstances for Senate reform may be the period immediately after a rampant government, supported by a large House of Commons majority, has trampled over widely held public objections in a way that offends the governing party’s members in the Senate. But the events that trigger reform are unpredictable. The best that reformers can do is produce plausible schemes and hope they will be taken up by one of the major parties.

Conclusion

The six propositions presented above list the key components to be considered if Senate reform is to be achieved. They are more important for setting out the problems which must be confronted by any reform process than for the particular suggestions they make. In particular, they point to the interests which must be accommodated: partisanship and the power of the prime minister. Ignoring these issues and generating ambitious plans for large-scale constitutional change is a recipe for failure. If Senate reform can focus on dealing with the central problem of achieving good and responsive government through the effective scrutiny of the national executive, then the issue is one of institutional design to harmonize competing interests. For this reason, direct election of the Senate is too ambitious a goal. But there is just a chance that amending the selection procedures for senators might be congruent with a range of interests concerned with parliamentary reform, and might provide Canada with the benefits of an effective system of parliamentary bicameralism.