The 'Unrepresentative Swill' Feel Their Oats
by Geoffrey BrennanÊ

The Rise of Senate Activism in Australia

The main object of this paper is to analyse the recent rise of Senate activism in Australian parliamentary politics. My claim is that this increase, together with the lower house response to it, represents an extremely significant change in the nature of Australian government – one that is likely to have substantial practical effect, and will (or should) radically alter the way in which we understand Australia political institutions.

It was Paul Keating, the then Prime Minister, who coined the term 'unrepresentative swill' for the Senate. In the rhetorical arena, Keating had a touch of true genius. His parliamentary speeches were generally thought to be undistinguished, but his quips were deadly – always well targeted and with a poisoned edge. In the Australian parliamentary context, this made him an effective, and for his party much valued, player. For all this, there was something rather surprising about this particular case. Typically, Keating's one-liners were reserved for members of the 'other' party (the Liberal-National Party coalition), and occasionally for that party as a whole. For him to focus his vitriol on an entire chamber – a chamber in which his own party had substantial, if not majority, representation – was uncharacteristic.

However, the circumstances were unusual. The Senate was beginning to 'feel its oats.' Keating, as leader of the recently victorious party in the House of Representatives, Prime Minister and Chairman/Head of Cabinet, had wanted to get on with the business of flexing his newly-acquired governmental powers. Until the next general election came round in three years' time, he could expect to 'run the country' as other Prime Ministers before him had done because, in the tight party structure that characterises Australian politics, he had the numbers. Or at least he did so in the lower house, the traditional seat of power. To be faced with an interventionist Senate, to be called on to share the power of government in a novel way with the Senate, must have struck him as an affront, both personally and to the distinguished office he occupied. In such circumstances, a certain irritation is probably understandable.

But was Keating right to feel that affront? And was his quip apt? Should the Senate leave the 'government of the day,' as Keating would understand it – that is, the party with the numbers in the lower house – to get on with the business of governing, more or less unimpeded? What are we to make of increased Senate activism? And what, if anything, does this have to do with the 'representativeness' (or otherwise) of the Senate? Is the Senate 'unrepresentative'? And if so, in what sense?

The Senate's Dubious (?) 'Representativeness'

The procedures for determining representation in the Senate differ in several respects from those in the House of Representatives – reflecting, for example, the fact that the Senate is nominally a 'states' house,' as well as the fact that, within states, the electoral system used is proportional representation rather than a system of single member electoral districts. Each state has an allocation of twelve Senators, six of whom are elected at each election. Thus, in the absence of a double dissolution, each Senator has a term of office of up to six years. In addition to the twelve Senators from each state there are two Senate seats for each of the territories, making 76 in all.

Which aspect of the Senate's electoral arrangements struck Paul Keating as particularly 'unrepresentative' remains unclear. He probably had in mind the fact that the Senate system of representativeness violates the principle of 'one person, one vote' often held to be a sine qua non of proper democratic regimes. Clearly, it does violate that principle: seats in Tasmania are allocated on the basis of one seat per 27,000 votes (roughly), whereas seats in New South Wales or Victoria are allocated on a basis of something like one seat per 330,000 votes. But this fact reflects precisely the principle of equal representation per state. It may have been Keating's view that the federalist basis of the Australian Senate is inappropriate in some way, or that federalism itself is an unfortunate political structure (indeed, this clearly was Keating's view). But the Senate was never intended to represent voters: it was intended to represent states. It would be coherent but hardly interesting to oppose the Senate because it was constructed on federalist principles.

On the other hand, most commentators would concede that any conception of the Senate as a primary defender of states' interests or as a distinctive arena for states' business has ceased to have any practical relevance: the Senate has, in this sense, long since been swallowed up by party politics. Whether this is sufficient to constitute a case for the abandonment of the bicameral arrangement is an entirely open question. It seems self-evident that the desirability of a bicameral structure at the Federal level ought to be assessed quite independently of the desirability of federalism: after all, the questions engaged are conceptually quite different, whatever the particular institutional history in Australia.

Whether, in terms of party representation, the composition of the Senate is significantly different from that which would have prevailed if the number of Senate seats had been allocated among states and territories according to population, for example, is an empirical question that I will not explore here. The answer would clearly depend on the degree of variation in party vote across states, but whether there would be any systematic effect on either of the major parties' representation seems unlikely. In this sense, Australia is somewhat unusual. The party structure is such that the major parties in all states are Labor and the Liberal-National Party coalition. There are no regionally specific parties such as exist in other federal systems. In Brazil, for example, many parties represented in the upper house have a significant presence in only one or two of the country's twenty-two states.

As things stand, and given the rise of the Australian Democrats over the last 20 years, it seems clear that the chief difference between the basis of representation in the House of Representatives and the Senate lies in the fact that a modified form of proportional representation is used for the Senate. The lower house is almost exclusively composed of the two major parties. The Senate is now, quite robustly, a multi-party house: the Australian Democrats have a sizeable presence and the Greens (and now One Nation) hover on the edge of a seat in a number of states, particularly when there is a double dissolution.

It must be an open question as to whether single member districts or proportional representation provides, in principle, the more 'representative' assembly. Much depends on one's views of how parliament operates, how it ought to operate, the effective strength in parliament of party allegiance and a host of other complex issues (on which I shall offer a few comments below). It seems doubtful that these were the issues that Mr Keating had in mind.

There is, on the other hand, one important feature of proportional representation that Keating may have had vaguely in mind, and which has become an issue of some moment for the two major parties – the Liberal-National Party coalition; and the Labor Party. This issue concerns the extent to which, in multi-party systems, the level of popular support, as reflected roughly in the number of seats won, maps into effective political power. There is a technical literature on this matter falling under the rubric of 'power indices,' but the central point can be made by appeal to a simple example – one that has some relevance to the Australian case.

Specifically, consider the case in which there are three parties in the upper house – denoted A, B and C. Suppose A has 49 seats, B 48 seats and C 3 seats. Notwithstanding the fact that A and B both have very much more extensive electoral support than C, and a correspondingly larger number of seats, all parties have identical power: in the sense that no party can form a majority without the support of one of the parties, but in all cases either of the other parties will do. That is, C figures as often as A and B in the set of possible majority coalitions.

Indeed, in some cases and under some definitions of 'power', the relation between representation (number of seats) and power can be negative: a party with a larger number of seats can have less measured 'power' than a party with a smaller number. It is, for example, perfectly possible for coalitions (A, C) and (B, C) to be more likely than the coalition (A, B), if C lies ideologically between A and B or if there are independent considerations that make a coalition between A and B unlikely. It is worth noting this latter possibility here because, as I shall argue, there are grounds for thinking that this is the situation in the Australian Senate and that the Democrats exercise power very much in excess of their number of seats (and popular electoral support). As I have indicated, however, this possibility is a characteristic feature of proportional representation: and it is, in one sense, as much attributable to the presence of parties as it is to the method of securing their representation.

Let me attempt to summarise to this point. The charge of 'unrepresentativeness' that Keating levels at the Senate might have force in three different senses: first, Senate election rules violate the one-person-one-vote principle; second, the election of Senate seats on the basis of a modified proportional representation system means that seats are allocated across parties in ways different from that in the House of Representatives; and third, parties have decisionmaking 'power' in the Senate that does not reflect the number of seats each party has.

In order to assess which, if any, of these dimensions of possible 'unrepresentativeness' is normatively objectionable, one would need a further account of 'representation' and the purposes it is designed to secure. It would be a distraction here to attempt such a fuller account. It is, however, relevant to note that such issues become more significant as the importance of the Senate's decisionmaking role increases. Moreover, the relative significance of the three senses may alter according to how much decisionmaking power the Senate exercises. My own view is that, in the current Australian setting, with relatively tightly disciplined parliamentary parties and with the rise of Senate intervention in ordinary legislation, the third of these senses is the most important. However, to claim this is to get ahead of the argument. The prior step is to describe the rise of Senate activism, and its effects within the Australian context. The issues of decisionmaking power of different parties within the Senate bear on these matters, but can to some extent be separated from them.

Four Facts And A Conjecture

I want now to attempt to set Senate activism in its context. To do so, I offer four background facts and a conjecture.

Fact One: It is part of the logic of our bicameral system, at least as it currently operates, that the party that holds a majority in the lower house will not hold a majority in the upper. This may not have been an entirely foreseen or intended attribute of the electoral arrangements for the Senate, but the rise of the Australian Democrats has made it so. In any event, it is a characteristic feature of the current structure of party support that, provided that the voting system does not change, no party is likely to command a majority in the Senate for the foreseeable future.

Fact Two: As is well known, the Australian Constitution vests considerable power in the Senate. The design of the Senate was an amalgam of the U.S. Senate and the U.K. House of Lords (which of course enjoyed very much greater powers in 1900 than it possesses now). The Australian Senate has the power to reject virtually all legislation initiated in the House of Representatives, subject to the 'double dissolution' provision. It may seem that the double dissolution establishes a substantial disincentive to reject House legislation, because half the Senators will thereby be sacrificing up to six years of office. But since that half can be virtually guaranteed re-election by receiving a place in the top half of their party's list of candidates, it is not clear that this is a particularly effective deterrent. The double dissolution option is to the advantage of smaller parties (those that poll between 8% and 16% of the vote, after the distribution of preferences) and hence to the disadvantage of the two major parties – but since neither major party is likely under normal circumstances to command a majority in the Senate it is not clear how severe a disadvantage this is.

Fact Three: The bicameral divide is the locus of most of the genuine parliamentary contest in the Australian system. Clearly, given the very tight two-party system that exists in the lower house, there is not much scope for any constraint of government power there. Apart from any effects on subsequent electoral contests (which need not be negligible), neither the debates in the lower house nor the attendant votes can secure any impact on government policy: except for special pre-announced 'conscience' issues, virtually no-one ever votes against her/his own party, so whichever party holds the majority of seats is secure within the House of Representatives. Within the lower house, the opposition has a voice, but its influence is mediated primarily through election-time effects. There is no need for the government of the day to bargain or negotiate within the lower house.

But the Senate does have genuine teeth – and has traditionally been the major parliamentary constraint on the exercise of government power. Double dissolutions are not uncommon, and the famous/infamous 1975 'dismissal' was grounded in the powers of the Senate to obstruct the legislative activities of the lower house majority. However, as far as one can tell, Senate powers did not traditionally act as a consideration in the framing of legislation: the majority party in the lower house acted on the basis of the convention that it had the presumptive right to govern unencumbered by Senate 'intrusion.' If the Senate chose to bring on a confrontation with the House and risk a double dissolution, then this would be seen to be the Senate's responsibility – and the
'recalcitrant' opposition parties that activated this option would have to take the electoral consequences.

Fact Four: In the recent past, the nature of Senate-House interactions has changed. Instead of carrying any fights between House and Senate up to the brink of double dissolution, it has become much more common for the majority party in the House to negotiate with the majority parties in the Senate over details (sometimes significant details) in new legislation. The departure point for this new form of bicameral relationship seems to have been the 1993 budget. On this occasion, the Liberal-National/Democrat/Green opposition in the Senate refused to pass the budget unamended. The ALP government then proceeded – with a due demonstration of reluctance (of which Keating's quip is an example) – to negotiate with the Democrats and Greens, to ensure the passage of the Budget through the Senate without further delay.

Since then, the Senate has pursued a more piecewise activism on both procedural and substantive matters. On the procedural front, it secured the imposition of a 'double deadline' to prevent a huge backlog of legislation being forced through the Senate in haste: under the new provisions, to be considered by the Senate in a given year a bill must be passed in the House prior to October 29th, and introduced into the House four weeks earlier. And the Senate has sought to negotiate (so far unsuccessfully) the power to compel bureaucrats to reveal the content of advice given to Ministers within the context of Senate reviews. Substantively, the Senate played an important role in the processes surrounding the formulation of the 1993 Native Title Act. And the 1995 and 1996 Budgets and the 1996 Industrial Relations Bill were all objects of detailed negotiation and amendment in the House before finally being passed by the Senate.

It is not entirely clear whether the 'correct' way to see these developments is in terms of increased Senate power or as a new means of exercising powers that the Senate always possessed. Arguably, the process of negotiation and bargaining is preferable to an occasional double dissolution outcome for both the majority party in the House and the minor party (or parties) in the Senate: it is hard to see why either would have agreed to that process otherwise. The major opposition party in the Senate might well prefer the all-or-nothing character of the double dissolution option, because that party stands to gain most (at least in expected terms) from a new election. But the opposition cannot secure this all-or-nothing option: that option could only ever be secured by an implicit bargain between the major parties. The interesting explanatory question would be this: what happened in 1993 to upset that implicit bargain, and to cause its replacement by a series of substantive explicit negotiations on policy issues?

Conjecture: This conjecture can be construed either as a claim about the way the world is or as a methodological principle of inquiry. It says that formal powers under the constitution tend to be exercised up to the limits the constitution allows. This conjecture is not meant to rule out the possibility that political agents may impose on themselves self-denying ordinances, or adhere to conventions that involve not exercising the powers that they formally possess. But it is meant to cast doubt on the survival of such self-denial or self-enforcing conventions 'in the long run.' Those who self-select for political careers tend to be those who desire the exercise of power, who have huge moral self-confidence about their own particular perceptions of the 'public interest,' who desire to 'leave their mark.' Explaining political outcomes is therefore mostly a matter of following where the logic of the pursuit of power leads. This may not be all of what matters politically; but I believe it to be an important part, and in particular an important part of what is systematic about politics, and hence amenable to institutional explanation.

Will the Real Opposition Please Stand Up?

One way of thinking about the issue of Senate 'activism' is in terms of the predicament facing the opposition party in the lower house (currently the ALP). Within the Australian system (like most party-dominated Westminster systems), the opposition has a conventional role: the opposition is supposed both to act as monitor and critic of the Government and to maintain the picture of a viable alternative government constantly in the minds of the electorate. There are of course other institutions (the media most notably) which have the role of monitor and critic, but the institutionalised opposition has a special role in this regard.

The spirit of the arrangement is deliberately adversarial, and as we have noted it goes with a rhetoric and theatre that is adversarial in nature. Parties – especially in the two-party system characteristic of lower house politics in Australia – tend to define themselves by their opposition to each other, and so much political rhetoric focuses on the inadequacies of the other (major) party. In this way, Liberal-Labor animosity has become one of the habits of Australian political discourse, and an explicit Liberal-Labor compromise on a matter of policy would be implausible (and perhaps electorally costly to both sides) except in circumstances that were widely regarded as 'exceptional.'

The patterns of relationship in the lower house are more or less replicated in the upper. To be sure, bipartisan committees can operate amicably and often with common purpose – but once lodged back into identifiable party roles, the adversarial element takes over. The relevant implication is that explicit Coalition/Labor Party negotiations over detailed aspects of proposed legislation are difficult to imagine: the two major parties in the Senate are more or less locked into their assigned lower house roles. Moreover, the cost to a major party in developing a practice of Senate negotiation when in a minority position in the lower house is that that party will have to contend with such negotiation when it achieves power. Both major parties would, in principle, prefer a more, rather than less, acquiescent Senate and would probably prefer to hold the Senate to a more limited 'strategy set' – namely, double dissolution or nothing.

The situation for the Australian Democrats could hardly be more different. They currently have negligible chance of being a decisive lower house presence (a situation that seems certain to prevail as long as the system of single member electorates is maintained). The Democrats therefore have no interest in maintaining a system that leaves the dominant lower house party (whichever it may be) to exercise power unencumbered by Senate intervention. The Democrats' attitude to the double dissolution possibility is more complex. It is true that a double dissolution may well increase the number of Senate seats accruing to the Democrats – but to the extent that it increases the number of small parties in the Senate, double dissolution is a negative outcome for them. In a situation where the proportion of seats is, say, (47%, 47%, 6%) with Coalition, Labor and Democrat proportions in that order, the Democrats hold the balance of power in the Senate. In a situation where the proportion of seats is, say (40%; 40%; 9%; 5%; 3%; 2%; 1%) with an array of smaller parties, the Democrats' position is much less strong: Democrat support is not needed to form a decisive majority.

Accordingly, for the Democrats piece-wise negotiation is superior to both acquiescence and double dissolution. The Democrats perform the role of 'critic with teeth.' While the oppositional traditions of Liberal and Labor parties prevail, the Democrats can act to specify the terms of agreement to legislation, reasonably confident that the opposition party would support a double dissolution if that ultimate threat were to be unleashed on the Government. The Democrats act as an effective opposition, raising criticisms that the Government is obliged to consider – while the opposition party faces the unenviable task of choosing between ineffective oppositional rhetoric and itself playing a rival negotiation game (i.e. establishing a Lib-Lab majority in the Senate against the Democrats, and against their own historical rhetorical posture). Press coverage understandably pursues the Democrats – where the policy action is – and the main opposition party has no viably attractive strategy to adopt.

Hence the question that frames this section. Just who is the real political opposition in Australia now, seems to be a newly opened question. As the nature of Senate-House relations changes, as the double dissolution extreme gives way to a more routine consistent pattern of negotiation on bills, the de facto power of the Australian Democrats is dramatically increased and the salience of the main 'opposition' party correspondingly diminished. Whether the main opposition party can live with this situation, whether this situation represents a stable equilibrium, I rather doubt.

Responsibility vs Debate: What Kind of Democracy Do We Want?

Attitudes to Senate activism (party politics aside) engage quite basic questions about how democracy works, and/or what it is that one looks to parliamentary institutions to do. In an earlier more general paper on parliamentary processes (Brennan and Hamlin 1993), we found it useful to distinguish three basic 'models' or 'pictures' of parliamentary process: the 'parliament as forum' model; the 'parliament as microcosm' model; and the 'parliament as prize' model.

In the first of these, parliament is conceived 'discursively': it is the arena for debate about political options, designed to uncover the ethically correct policies which will then be drafted into legislation. Parliamentarians are conceived essentially as moral inquirers; and parliamentary debate is a process of persuasion of other parliamentary representatives as to the course it would be most justifiable to pursue, all ethically relevant things considered.

In the 'parliament as microcosm,' the central idea is that parliament should be statistically representative of the population at large.
Although debate may take place here, what is most fundamental is the fact of independent representatives voting, so that parliamentary outcomes more or less simulate what ordinary citizens would decide if they had the time to become informed about issues in the way that full-time representatives can. Parliament here is a 'micro' version of the polity at large: and representatives vote on options put to them by an executive (cabinet) on whatever basis they are inclined (e.g. narrow self-interested voting would not be ruled out here, because each interest would reflect the interests of the group whom the 'representative' statistically represents).

The 'parliament as prize' picture, by contrast, sees the centrepiece of democratic process as electoral competition: the 'final say' rests with 'the people,' but there is a role for parliament in providing arguments and counter-arguments about the qualities of policies and candidates to the electorate at large. Within this picture, the issue of 'responsibility' is central. If parties/candidates are to be evaluated by the electorate at large, they must both be able to promise delivery of policy outcomes and be held accountable by the electorate if those outcomes do not eventuate. Here the central institutional constraint is the general electoral one: other institutional arrangements may support its operation, say by helping to make voters better informed, but arrangements that muddy lines of accountability are necessarily somewhat dubious.

On this reading, the role of an upper house may include that of helping to inform the public through inquiries, independent review and so on: but the difficulty posed by inter-party negotiations and trades is precisely that both parties can claim credit for the policies and outcomes that prove electorally popular and both disclaim responsibility for any policies that prove unsatisfactory. If policy is a matter of negotiation between government and Senate, then neither potential governments nor (of course) Senate parties can deliver to the electorate policies that they might offer. Senate activism, interpreted as a mechanism by which the Democrats secure an opportunity to put into effect their own policy platform, can only obscure electoral accountability.

The 'responsibility' doctrine has most force when democratic electoral constraints are most effective. Within the rational actor analytic repertoire, for example, if the median voter model applies, electoral competition in a two-party contest will force contending parties to occupy the ideal point of the median voter: both parties will tend to an identical centrist position on policy issues. But the rational actor tradition has also been much concerned with the possibility that majority coalitions might exploit minorities. Almost exactly one hundred years ago Knut Wicksell formulated his influential analysis of a bargaining model of expenditure and tax determination. A central anxiety for him was that the governing party (Wicksell thought of parties as class-based) would vote itself special-interest public expenditures paid for by the other, and he conceived a system of effective unanimity within the representative assembly as a means of avoiding such exploitation.

The classic modern variant of these ideas is Buchanan and Tullock's The Calculus of Consent: shorn of the class orientation of Wicksell's treatment (for Buchanan and Tullock any majority coalition is liable to exploit the minority) and with the unanimity requirement reinterpreted to apply at the more abstract 'constitutional' level of decisionmaking about political processes themselves, Buchanan and Tullock develop an argument for super-majoritarian decision rules on essentially similar lines to Wicksell's. Interpreted through the Wicksell/Buchanan-Tullock lens, the rise of Senate activism amounts to an effective extension in the required size of the majority coalition.

To the extent that what is at stake is a perpetual effective bargain between the Democrats and the ruling party in the lower house, the protection of the interests/values of a wider group is secured. To the extent that the main opposition party can compete effectively with the Democrats to be the negotiating partner with the majority party in the lower house, then the spread of interests/values actually represented in political decisions extends to much of the electorate: Senate activism becomes the mechanism by which Australian political process is transformed, from an essentially majoritarian one towards a quasi-consensual one. Senate activism will, however, be the better described in such terms to the extent that the main opposition party comes to participate more effectively in negotiations across the bicameral divide.

On an idealised view of 'parliament as forum,' this move will not perhaps be seen as particularly significant: the Senate has always had a voice, and the prospect of minority parties influencing government policy through parliamentary debate has always been available in principle. But it is difficult to render Australian parliamentary practice (and not just Australian) in these 'forum' terms: voice without votes, and votes without power, do not cut any ice. Which, in my view, reveals just how extreme an idealisation, how remote from reality, the parliament-as-forum view is. This is not to deny that elements of 'discursive democracy' may not be present in political process more broadly (for example, within Cabinet, or the party rooms, or Senate committees). I merely claim that in the essential lineaments of party influence, the matters on which Senate powers and the mode of their execution bear, the specifically discursive element in democracy is relatively minor.

References

Brennan, Geoffrey and Alan Hamlin 1993, 'Rationalising Parliamentary Systems,' Australian Journal of Political Science 28, 443-457.

Brennan, Geoffrey and Alan Hamlin 1998, 'Representation: The Theory of Political Agency,' British Journal of Political Science (forthcoming).

Brennan, Geoffrey and Akan Hamlin 1999, Democratic Devices and Desires, Cambridge University Press, Cambridge (forthcoming)

Brennan, Geoffrey and Loren Lomasky 1993, Democracy and Decision, University of Cambridge Press, New York.

Buchanan, James M. and Gordon Tullock 1962, The Calculus of Consent, University of Michigan Press, Ann Arbor.

Geoffrey Brennan is Professor in the Research School of Social Sciences at the Australian National University. An earlier version of this paper was prepared as notes for a seminar at the Australia Centre, University of Potsdam.

Policy is the quarterly review of The Centre for Independent Studies. For more information on subscribing to Policy, click HERE

If you are interested in the Centre's activities and publications, why not subscribe to e-PreCIS, our regular email update on the latest news and events.

(e-PreCIS requires html capable email facilities, such as Microsoft Outlook Express or Netscape Messenger)