6. Coalitions in the Chamber

Navigation: Previous PageContents | Next Page

By this point, there should be no doubt that there are two dominating facts of parliamentary life in Canberra: first, the government controls the House, always in numbers and invariably in votes; and second, the government rarely controls the Senate in either sense. In turn, there are two primary reasons for the second of these facts: first, both the ALP and the Coalition enjoy roughly comparable levels of public support, but neither enjoys majority support in the electorate; and second, the system of proportional representation for electing Senators ensures a fairly accurate translation of votes into seats, so that neither an ALP government nor a Coalition government can expect to have a dependable voting majority in the Senate. Complementing these reasons is a third. The staggered election of Senators, by which half the Senate is elected every three years (except following double dissolutions) means that even if a landslide election should bring a new government into office, that government may have to wait another three years before it can even hope to translate its public support into a solid Senate majority, assuming its public support survives that long, and it can win a second, successive landslide victory.

The need for Senate coalitions

The key implication of this situation is that the Commonwealth legislative process is not simply a process of translating the government’s policies into laws, as would be the case in a true ‘Westminster’ system in which (1) the government can depend on majority support in the lower house, as it can in Canberra, but (2) the government also controls the upper house, or the upper house can do no more than delay enactment of the government’s program, which is not the case in Canberra. In the Commonwealth Parliament, as we have seen, a non-government majority can block enactment of the government’s program by refusing to pass any bill, even, in extremis, the most essential money bills, as the events of 1974 and 1975 demonstrated. Consequently, the legislative process in Canberra inescapably is a process of coalition-building.[113] The government can depend on its disciplined party majority to pass its legislation in the House, but it must construct a coalition that extends beyond, even if not far beyond, its own party members to ensure that the same legislation also passes the Senate.

As of the end of 2002, the Coalition Government held 35 of 76 Senate seats, with 39 constituting an absolute majority. Therefore, the government needed to find at least four votes from outside its party ranks to pass each bill (and take most other actions in the Senate), though it needed only 38 votes to defeat any motion or block any other action that it opposed.[114] The 41 non-government Senators were distributed among the ALP Opposition (28), the Australian Democrats (7), the Greens (2), Pauline Hanson’s One Nation (1), and 3 Independents. From the government’s perspective, this situation had remained fundamentally unchanged since the Liberal-National Government under Prime Minister Howard took office in 1996. During those six years, two Senators had left Labor to become Independents and one Australian Democrat had done the same, developments which, as we shall see, made a difference. Otherwise, the Howard Government faced the same challenge in the Senate throughout the six-year period. In years to come, the numbers may change, and it is possible that some minor parties may come and go (as I write, the future of the Australian Democrats is a subject of frequent speculation), but the essential fact—the government’s need to find support on each vote from non-government Senators—is very unlikely to change.

Voting in the Senate

This and the following chapter provide glimpses into how the government has tried to cope with this challenge and how successful it has been. To explore these questions, we shall look at the record of votes that have taken place in the Senate chamber. But to understand which of those votes we shall examine, a brief summary of how Senators vote is in order.

The Senate’s voting procedures are summarised in Odgers’ Australian Senate Practice (2001: 244–245):

Every sitting day the Senate determines a very large number of questions, most of which are determined by votes on the voices, that is, votes which are taken by the President calling for the ayes and noes and declaring the result without a record of how each senator voted. Most questions are determined in this way because they are uncontested, but it is not unusual for contested questions to be so determined when senators know and accept the way in which the majority is voting. ...

After a question is put and senators have called aye or no, the President declares whether the ayes or noes are in the majority. Unless the President’s determination is contested by the senators declared by the President to be in the minority, the determination of the President is recorded as the result of the vote. Only senators determined by the President to be in the minority may contest that determination and require a formal recorded vote, that is, a division, to be taken. ...

A division is held only if two or more senators call for the division ...

Several points are noteworthy. First, most questions are decided ‘on the voices’ and without a formal record of how any party group or individual Senator voted.[115] Second, however, it takes only two Senators to call a division, which is just about as minimal a requirement as the Senate in its standing orders could impose. And third, even controversial questions may be decided without a division, often because Senators on the losing side of a voice vote[116] conclude that nothing would be accomplished by insisting on a division, except perhaps to inconvenience and annoy their colleagues.

In this chapter and the next, we shall be concerned only with votes taken by division.[117] In doing so, we are looking at only a small fraction of the votes that take place in the Senate each year. In 1997, the Senate passed a total of 224 bills and decided 280 questions by division.[118] Three years later, the number of Senate divisions (115) was substantially less than the number of bills the Senate passed (181). It also is fair to say that questions decided by division are unrepresentative of the whole in that, in the great majority of cases, they are questions that are important and often contentious or controversial, at least to the Senators calling each division. However, it would be unwise to assume the opposite: that questions decided on the voices are, for that reason, clearly not important, contentious or controversial; nor would it be correct to assume that the government is happy with the outcomes of all votes on the voices. As suggested in the preceding paragraph, for instance, when a government defeat is a foregone conclusion, its leaders in the Senate chamber may decide that it would not serve their purposes to have their defeat documented by a time-consuming division.

Still, there are two compelling reasons for looking at divisions in the Senate. First, and notwithstanding the arguments above, the most important and divisive questions are the ones most likely to be decided by divisions. And second, not incidentally, we have absolutely no way of knowing for sure who voted on which side of any question that was decided on the voices. Not only are divisions the best choice for analysis, they are the only available choice.

Winning and losing

If we now turn to what the Senate actually has done, an obvious, and ultimately the most important, question to ask is how often the government has won and lost when it has come time for the Senate to vote. The first rows of Table 6.1 offer answers with respect to divisions in the Senate during 1996–2001.

Table 6.1: Divisions won by government coalitions, 1996–2001

 

1996

1997

1998

1999

2000

2001

Number of divisions1

197

280

216

224

115

66

Number of winning government coalitions

130

165

152

164

88

51

Percentage of divisions won by government coalitions

68.4

58.9

70.4

73.2

76.5

77.3

Number of minimum winning government coalitions

124

137

143

143

81

44

Percentage of winning coalitions that were minimal

95.4

83.0

94.1

87.2

92.0

86.3

Percentage of all minimum winning
government coalitions comprising:

Government and Opposition

18.5

40.1

32.2

44.7

70.4

77.3

Government and Australian Democrats

37.9

26.3

7.0

37.1

29.6

22.7

Government and Independents

43.6

33.6

60.8

18.2

0.0

0.0

1 Excluded are one free vote in 1996 and nine in 1997, all of which pertained to the Euthanasia Laws Bill 1996.

As we see, the number of divisions has varied considerably from year to year, from as many as 280 in 1997 to as few as 66 in 2001, which was an election year. What has not varied nearly as much, however, has been the government’s record of successes and failures on these votes. The government was on the winning side on percentages of divisions that varied from a low of 58.9 per cent in 1997 to a high of 77.3 per cent in 2001. And if we set aside 1997, the government’s winning percentage varied within only a nine point range. If we were concerned with genetic differences among types of fruit flies, these year-to-year differences would be considered great. For a large, complex, and human institution like the Senate, what is striking is the relative consistency of the government’s success rate.

If we were examining the outcomes of divisions in the House, these data would be evidence of disastrous failure by the government. It bears emphasizing, though, that the present government has been winning between roughly three-fifths and three-quarters of all divisions in the Senate even though it lacks a Senate majority. In all these cases, the government has found sufficient numbers of allies from outside its own ranks. These data tell us nothing, however, about how successful the government has been in winning the divisions that mattered most to it and to the other parties in the Senate. This is a typical limitation of such quantitative analysis. Also (and this is a point to which we will return), we cannot assume that the government’s record of success is entirely attributable to the soundness of its policies or the abilities and persuasiveness of its Senate leaders. An unknowable number of government victories undoubtedly were passive or accidental.

Nonetheless, these reservations should not mask the story that the top of Table 6.1 tells: that this government has won in the Senate, far more often than not, even though non-government Senators could have combined to defeat it on each and every division. All of these government victories have required coalitions of support extending beyond the party rooms of the Liberal and National parties.[119] How far have these coalitions extended, and how often have the other parties in the Senate participated in them? Before turning to these questions, we need to consider the government’s options for constructing winning coalitions.

The government’s coalition options

As I recounted at the beginning of this chapter, the party distribution in the Senate at the end of 2002 was as follows:

Government (Liberal and National parties) 35
Opposition (Australian Labor Party) 28
Australian Democrats 7
Australian Greens 2
One Nation 1
Independents 3

There are a total of 76 Senators, so 39 of them constitute a majority that is sufficient to win. Therefore, the government needed to find at least four votes from outside its party ranks in the Senate to pass each bill and take most other actions. If we put aside for the moment the One Nation Senator and the three Independents, the government had several coalition options: it could construct a large coalition with the Opposition (making 63 votes) or a smaller but sufficient coalition with the Australian Democrats (42 votes). Coalitions with both the ALP and the Democrats also were possible, of course, but the government could win without one or the other. The Australian Greens also could join a government coalition with Labor, the Democrats, or both, but the support of the Greens could not be decisive. In 2002, the Greens were too few to make a majority with the government, and they were not necessary to make a majority if the government had the support of either of the other parties in the Senate.

The government had two more options. First, it could win with the support of the Greens and at least two of the other four Senators who were independent decision-makers—the One Nation Senator and the three Independents. Or second, the government could win without the support of any other multi-member party if it secured the support of all four of the other Senators. In earlier years, as we shall see, the government won a notable number of divisions by relying on the votes of Independents and, since the 1998 election, the sole Senator representing Pauline Hanson’s One Nation (whom I shall treat from this point on as if he were an Independent, rather than treating him as if he constituted a party group all by himself). My interest is primarily in coalition arrangements among party groups in the Senate, so in what follows, the Independent Senators sometimes tend to disappear from the analysis, except to the extent that they have made it possible for the government to win without any of the other party groups voting with it. This approach greatly simplifies the presentation, and I do not believe that incorporating the Independents more fully in the analysis would significantly change the essential arguments or findings. However, I do not intend in any way to dismiss the importance of Independent Senators. Instead, think of what follows as a discussion primarily of coalition-building among parties in the Senate.

Table 6.2 presents data on party representation in the Senate since 1996, and documents both continuity and change in the government’s options for forming winning voting coalitions. Over the years, the size of each of the Coalition, Labor, and Democrat groups varied by no more than two; but in a closely divided Senate, even such variations can make a difference. Throughout the period, the government could form winning coalitions with either Labor or with the Democrats; and for roughly three years between mid-1996 and mid-1999, when Coalition strength peaked at 37, adding the votes of the two Green Senators could produce the smallest minimum winning coalition possible. During this period, however, there actually were two Green parties, each with one member in the Senate. As a close observer of the Senate has explained:[120]

The two Green senators were in fact from different Green parties, the Western Australian Greens (Dee Margetts) and the Australian Greens (Bob Brown). My observation is that this partnership was much less cohesive and productive than the earlier partnership between the two Western Australian Greens (Margetts and Christabel Chamarette) and the current Australian Greens duo (Brown and Kerry Nettle), although I doubt that disagreements will be obvious from the records of the Senate. It was not uncommon for Brown and Margetts to put forward their own amendments to the same bill ... although they would of course vote for each other’s amendments.

Of course, coalitions between the government and either or both of the other parties also would produce winning majorities; in any such case, the support of the Greens would be superfluous, at least numerically. During some periods, the government also could win divisions without any of the multi-party groups, but solely with the support of Independent Senators. As the table reveals, that option was available during the period from September 1996 through June 1999, and then again during August to December 2002. Earlier in 1996 and between July 1999 and July 2002, a voting combination of the government and however many Independent Senators there were fell short of the magic number of 39.[121] During much of the time, however, the votes of Independents were enough to give the Coalition the numbers it needed for a minimum blocking coalition: a majority of 38 that would suffice to defeat any proposal put forth by one or more of the other parties.

Table 6.2: Party representation in the Senate, 1996–2002

 

Coalition

Labor

Australian Democrats

Greens

One Nation

Indep-endents

Jan 1996–Jun 1996

36

30

7

2

1

Jul 19961–Aug 1996

37

29

7

2

1

Sep 19962–Jun 1999

37

28

7

2

2

Jul 19993–Sep 2001

35

29

9

1

1

1

Oct 20014–Jun 2002

35

28

9

1

1

2

Jul 20025

35

28

8

2

1

2

Aug 20026–Dec 2002

35

28

7

2

1

3

1 On 1 July 1996, the Senators elected at the half-Senate election of 2 March 1996 took their seats.

2 On 20 August 1996, Senator Mal Colston left the ALP and became an Independent.

3 On 1 July 1999, the Senators elected at the half-Senate election of 3 October 1998 took their seats.

4 On 2 October 2001, Senator Shayne Murphy left the ALP and became an Independent.

5 On 1 July 2002, the Senators elected at the half-Senate election of 10 November 2001 took their seats.

6 On 26 July 2002, Senator Meg Lees left the Australian Democrats and became an Independent.

The likelihood of the various possible party coalitions actually being created in practice depends, naturally enough, on the policies or ideology of each party group (and that of each Independent) as well as its numbers in the Senate. On some bills, the government and Opposition can find themselves voting the same way, by pre-agreement or otherwise, because those bills do not address matters that separate the parties, nor are they bills on which the Opposition thinks that it can gain some political advantage by opposing the government. On other bills, however, the prospects for the Coalition and the ALP reaching agreement (forming ‘grand coalitions,’ if you will) are slight because of sincere policy differences, calculations of political advantage, or both. With regard to the Australian Greens, their positions on the issues that the Parliament addresses would lead us to expect that they should join with either Labor or the Democrats, or both, much more often than with the Coalition government. As we shall see, however, this has not been uniformly true.

The Australian Democrats present what are perhaps the best opportunities and the greatest uncertainties for the government. There has been a tension among Democrat Senators, reflected in recent leadership challenges and departures from the party, between those who envision their party as a force for moderation and sensible compromise between the more extreme positions of both the Coalition and the ALP, and those who stress the need for their party to respond to the preferences and discontents of those on the political left as well as to the danger that, if the Democrats fail to do so, their Senate representation may shrink in favour of the Greens. Because of these philosophical and strategic differences, the Democrats should be more likely to join a voting coalition with the government on some issues, but with the Labor Opposition on others. Unlike the other three parties, as we shall see, the Democrats even have been known to split their votes, some voting with the government and others voting against it. According to Sugita (1997: 157), ‘A study of the Senate divisions between August 1981 and December 1996 reveals that there were only eighty divisions during this period when the Democrat Senators did not vote as a bloc.’ In comparison with the voting patterns of ALP, Liberal, and National Senators, however, ‘only eighty’ is almost infinitely large.

This last observation also raises another question: whether we can think of each party group as a single, unitary player. Critics of the House of Representatives as a legislative body sometimes describe it as a place where the government writes bills and its party members vote for them. Reality, however, becomes more complicated if the government finds that its preferred version of a bill faces determined opposition, perhaps led by members of one of its parliamentary party committees, behind the closed doors of its party room. While this would be unusual in Parliament House today, if such a situation were to arise, the government might well have to accept changes in its bill either before the bill is introduced or at some later stage of the process. It may not be helpful to think of the government having to build a coalition among its own party members. Nonetheless, in the process of trying to build a winning coalition in the Senate, the government (and, to a lesser extent, the Opposition) still may want to glance over its collective shoulder from time to time to gauge reactions within its own ranks to whatever concessions it is in the process of making to prospective coalition partners.

Similarly, it is tempting to assume that if either the government or the Opposition reaches a coalition agreement with another party group (or with each other, in the case of a potential grand coalition between government and Opposition), that the negotiator for that party can commit all the other members of his or her party group. In most cases, this is a reasonable assumption because of the strength of party discipline in the parliamentary parties. The Democrats are the only party that has split its votes on divisions during 1996–2001 (except on the handful of free, or ‘conscience’, votes). However, the strict party discipline on which the other parties insist when it comes time to vote is not enforced with equal rigor at other times and in other venues. In other words, intra-party differences can arise after two or more parties have reached a tentative coalition agreement, with the result that one or more of them then finds that opposition within its own party ranks requires the negotiations to be reopened.

All of this makes the analysis of Senate divisions more interesting than in many other national assemblies, but also much more manageable, methodologically and analytically, than in the US Congress. In most parliamentary chambers with significant powers—that is, in most lower chambers—there is little point in analyzing voting patterns. Unless there is a minority government or one that rests on a shaky majority foundation, the government can be expected to win all (or almost all) contested votes, whatever the procedural equivalent of divisions may be. How non-government parties vote usually does not affect the outcome, nor do occasional defections from the government’s ranks matter very much unless they are sufficient in number to cause the government to lose a key vote. Those are the votes worth examining—the rare votes that the government loses—because any one of them could bring about the government’s demise.

In the US House of Representatives and Senate, on the other hand, voting cohesion among Democrats and Republicans is quite high—and considerably higher than the most common perception abroad—but still, the outcome of many votes is determined by how large a minority of Democrats vote with the majority of Republicans and how large a minority of Republicans vote with the majority of Democrats. Most Representatives and Senators do not cross the aisle very often but almost all of them do so on occasion, some much more frequently than others. When neither party holds a large majority of seats, the numbers moving (figuratively) across the aisle in each direction is critical to determining the outcome of each vote. This means that the analysis of voting patterns is important and it is a complex undertaking because the unit of analysis is each of the 435 Representatives or each of the 100 Senators.

By contrast, the unit of analysis for Senate divisions in Canberra is, for most purposes, the party. There are 76 Senators, to be sure, but the outcomes of most divisions are determined by how each of the four multi-member parties vote as a bloc. Only if these four ‘votes’ do not produce a conclusive outcome do the votes of Independents become determinative.

Minimum winning coalitions

With this prologue, we can ask a question that has interested political scientists studying coalition formation in a variety of settings in which groups of participants decide questions by voting. When a winning coalition is formed, what is the likelihood that it will be a minimum winning coalition—in other words, a coalition that is no larger than necessary to win?

We might expect to find minimum winning coalitions whenever the person or party trying to construct the coalition must pay a price of some kind to attract each new member to it. That price may take the form of a compromise. In a legislative setting, for example, the compromise may require the coalition-builder to accept a weaker proposal than he or she personally prefers because that compromise is the strongest proposal that is acceptable to a prospective coalition partner whose support is necessary if the coalition is to be large enough to win. Alternatively, the price may take the form of a side-payment. The coalition-builder may be able to secure a prospective partner’s support for the coalition-builder’s preferred position if the latter agrees in turn to support the prospective partner’s position on a later decision about which the partner cares more intensely. Assuming that a coalition-builder does not want to pay a higher price for victory than is absolutely necessary, he or she will have an obvious preference for a minimum winning coalition. If the group is going to decide the question by majority vote, a coalition comprising 51 per cent of the participants is sufficient to win. Constructing a larger coalition requires paying prices, in compromises or side-payments or both, for votes that are not needed to achieve the coalition-builder’s purpose: winning.

Constructing a minimum winning coalition that includes only as many votes from outside the government’s party ranks as it needs to win is a parsimonious strategy in two senses. Not only does this strategy minimize the compromises, concessions, or side-payments that have to be made, it also minimizes the time and effort that have to be expended in building the coalition. In an active legislative body, the time, energy, and attention of members are limited, and they are even more limited for party leaders with multiple responsibilities. If coalition-builders are satisfied to construct minimum winning coalitions, they can invest their remaining personal resources (time, energy, attention) in building more such coalitions. Negotiating to secure superfluous additional votes for one bill may come at the expense of being able to secure essential votes for one or more other bills.

On the other hand, there are at least three general reasons why some coalitions are larger than minimal. One reason is uncertainty. A coalition-builder who constructs a minimum winning coalition must be absolutely sure that every member of that coalition will vote as the coalition-builder expects and wants. There is no margin for error. Defeat will result if even one anticipated member of the coalition reneges on his or her commitment or fails to participate in the vote. If anything significant is at stake, this can be too much of a risk to take. So coalition-builders often want some cushion to support their majority. A coalition of 55 per cent may suffice, for instance, depending on how much confidence the coalition-builder has in his or her information about the voting intentions of each of the other expected coalition partners, how much trust he or she has in each of those partners, and how much is at risk. This inevitably involves finding the best balance between costs and confidence. A coalition-builder can be absolutely confident of victory if every participant is incorporated in his or her coalition, but the cost of constructing a universal coalition probably will be much too high, and unnecessarily high, even assuming that it is possible. Perhaps a five per cent cushion is enough, or perhaps ten per cent or more. It all depends on the circumstances.

Second, the results of votes are not always the results of coalition-building efforts. Theories of political coalitions sometimes start from the mistaken assumption, and often an implicit one, that voting outcomes necessarily reflect the successes and failures of attempts by protagonists to build winning coalitions, minimum or otherwise. Yet that is not necessarily the case. In legislative settings, some proposals are assumed, and assumed correctly, to enjoy such widespread support that those responsible for ensuring their adoption need do nothing more than let nature take its course—let the participants vote as they choose, knowing that their free choices will yield the requisite majority. In other cases, there may be legitimate doubt about what the outcome of a forthcoming vote will be, but neither side invests much time and effort to ensuring that its position will prevail. Instead, both sides again allow the vote to take place without having engaged in deliberate coalition-building activities. They may be preoccupied with or distracted by other matters, or they may not care enough about the outcome to make it worth their while to try very hard, if at all, to affect it. The result may be a loss, a narrow victory, or a comfortable or overwhelming one. Those assembled on each side of the question still can be called winning and losing coalitions, but not coalitions that were the product of calculation or effort.

The first of these reasons is not generally applicable to divisions in the Australian Senate because of the strength of party discipline in voting. Except on the few free or ‘conscience’ votes, Liberal, National, Labor, and Green Senators stand united with their party colleagues on divisions. The Democrats are the only party in recent years to split their votes, and then only rarely. Uncertainties about Senators’ voting intentions generally should arise only with respect to Independents, and any such uncertainties should matter only when the votes of Independents are likely to be decisive—that is, when the government has failed to form a winning coalition with one or more other party groups. On the other hand, the second reason for larger-than-minimum coalitions certainly does apply to the Senate. Close observers warn that it is a mistake to look behind the outcome of each division for evidence of the calculations and efforts by party leaders that must be there. Sometimes those leaders can only wait to learn the outcome of a division, perhaps because, with other matters demanding their time and attention, they had done little or nothing to try to determine the outcome. On each vote there is a winning coalition, but the voting record itself reveals nothing about how actively involved anyone was in trying to construct that coalition. Inescapably, I fear, the discussion that follows may imply that more effort was devoted to coalition-building than we would find if it were possible in each case to learn who did what, and with what effect. This is a caveat that deserves to be kept in mind.

The third reason for larger-than-minimum winning coalitions is that winning may not be the only goal that coalition-builders have in mind. There may be other purposes to which the vote can contribute, and those purposes may provide incentives to assemble more than a bare 51 per cent majority. Imagine, for example, that a labor union is about to vote to strike unless management agrees to a ten per cent wage increase. Those supporting the proposal want to win, of course; that is their first and essential goal. But they probably want to do more. They want the vote to contribute to preserving or increasing solidarity in the union’s ranks so as many members as possible remain firmly committed to the same goal and the same course of action. At the same time, they want to weaken management’s resolve by sending a clear and strong signal that the union is united and determined to achieve its objective. In this situation, winning by a 51 per cent majority is the next best thing to losing. An overwhelming majority is not a luxury that is not worth paying for; it is a necessity.

In the Senate, the government may be willing to make additional concessions in return for additional support when, for example, it wants to demonstrate that it does not stand alone, or when it wishes to leave no doubt that the government—and people—of Australia are united in the face of a common challenge or danger—terrorism, for example. Or the government may be willing to pay a high price for support for a bill by another party, especially the Opposition, in order to immunize itself against partisan attacks for having brought the bill forward. Or the government may find that additional legislative compromises are the price it must pay to move its bill quickly through the stages of the legislative process. The government may be able to win without making those compromises, but only after a more elaborate and time-consuming process than it is willing to endure. The government may face a deadline that is externally imposed—for example, the beginning of the new financial (fiscal) year or the opening of some new international conference or negotiation—or one that is internally imposed—for example, the time pressure that the government’s other legislation imposes on the parliamentary schedule or the date the government has chosen for the next election. Under such circumstances, the content of the legislative agreement that gives rise to a larger-than-necessary winning coalition may be less important than the size (even the unanimity) of the coalition and how quickly it can be assembled.

So for any of these reasons, or others, a minimum winning coalition is not always either the desired outcome of a Senate division or the outcome that should be predicted.

Before examining the composition of winning government coalitions, we need to clarify that, in the Senate, a minimum winning coalition is unlikely to be one that is composed of only 39, or 51 per cent, of the 76 Senators. In the context of voting in the Senate, a minimum winning coalition is one that involves the minimum number of coalition partners, not the minimum number of Senators. The only time when the government could form a multi-party coalition of exactly 39 votes was between July 1996 and June 1999 when there were exactly that number of government and Green Senators combined. At all other times, a winning coalition that included the minimum number of parties necessarily included a larger-than-necessary number of Senators: at the extreme, a grand coalition of the government and the Opposition during the first half of 1996 included 66 of the 76 Senators. Still, we should think of this as a minimum winning coalition because forming it required the agreement of only two participants, even though in this case they were the two largest collective participants in the Senate.

Returning to Table 6.1 with this understanding in mind, we find data on the number and percentage of winning government coalitions that were minimum winning coalitions, and these data are striking indeed. During each of the six years, no fewer than 83 per cent and as many as 95.4 per cent of all the divisions on which the government was on the winning side produced coalitions that were minimum winning coalitions. We cannot know how many of these coalitions were the product of conscious effort—in other words, how often the government worked to secure the support it needed to win, but only that much support and no more—and how often these coalitions were the product of each party following its own inclinations without the government having made much or any effort to ensure that it would prevail. However, if no less than 83 per cent of all the government’s winning coalitions were minimal, that means that the percentages of larger-than-minimum coalitions were strikingly low.

It should be emphasized once more, however, that this table tells us about the frequency of minimum winning coalitions only on votes that were decided by divisions, not on all the votes that took place in the Senate. Relatively few divisions were what in Washington sometimes are called ‘hurrah’ votes, with almost everyone voting on the same side,[122] but it is perfectly reasonable to assume that divisions were less likely to occur on propositions that enjoyed more widespread support. Perhaps what the data are telling us is that when the government had more than minimum winning coalitions on its side, any parties (or Independents) who were not part of those coalitions were less inclined than otherwise to call divisions.

The second half of Table 6.1 addresses the composition of minimum winning government coalitions. How often did the government win by securing the support only of the Opposition, or the Democrats, or Independents? Notice first that the government frequently relied only on the votes of Independents when it was possible to do so and win (which it could not do in 2000 or 2001). In 1996 and 1998, the government won in partnership with only the Independents more often than with only the ALP or only the Democrats. More than half the time the government won divisions in 1998, it did so without the support of either of the parties with which it could form winning coalitions. However, that was not the pattern in either 1997 or 1999. Depending on what we make of the data for 1998, we can see a quite stable pattern in the frequency with which the government won in minimum winning coalitions with the Democrats, and a decline in the government’s reliance on the support of Independents only.

But perhaps most interesting are the data on ‘grand coalitions’ between the great putative parliamentary antagonists, the government and the Opposition. In 1999, Senator Helen Coonan, a Liberal minister, wrote (1999b: 14) that, ‘With the election of the Coalition in March 1996 the attitude of the Opposition and the minor parties in the Senate could be fairly described as a “culture of confrontation”, where the Senate routinely opposes most of the Government’s policy agenda.’ However true that observation may have been in 1996, Table 6.1 documents how much that situation changed in the years that followed. In four of the six years beginning in 1996, when the government was part of a minimum winning coalition, its coalition partner was the Opposition more than 40 per cent of the time. And most remarkably, whenever the Coalition government won a division in 2000 or 2001 with the support of one other party, more than seven times out of ten that party was the ALP.

During these two years, there were not enough Independents to give the government all the additional votes it needed, so the government could build a minimum winning coalition with either Labor or the Democrats. When a minimum winning coalition was the result, the coalition partnership that emerged far more often than not was between the Coalition and its historic rival and enemy, the ALP—not, as we might have expected, the Democrats. This is one of several indications we shall encounter that the popular image of parliamentary and political warfare defining the relationship between the government and the Opposition, between the Coalition and the ALP, needs to be tempered.

Government coalitions on divisions

Tables 6.3, 6.4, and 6.5 all document in different ways the coalitions that were formed between the government and other parties on divisions during 1996–2001. The first of them, Table 6.3, presents the percentage and number of times each year that the government voted with each of the other parties singly or in various combinations on all divisions. The first six rows of this table all represent winning coalitions, except for one 2001 division when the support of Democrat Senators was not enough for a government victory because the Democrats split their votes, some voting with the Opposition. Only in 1997 did the Liberal-National Government and the Greens combine to win more than one division.

In each of the years between 1996 and 1999, the most common voting pattern on divisions was for the government to be opposed by all three of the other multi-member parties. The government voted alone in divisions more than half the time in 1996, 1997, and 1998. This pattern was roughly three times as common as any other in 1997 and 1998. In 2000 and 2001, on the other hand, the government and Opposition voted together against the Democrats and Greens roughly half the time and more than twice as often as the government Senators voted without the support of any other party. However, two points need to be emphasized about divisions in which none of the other parties voted with the government. First, these are not necessarily instances in which the government found no support outside its own party room because this table does not take into account how Independents voted. Second, and precisely because of the Independents’ votes, the government won significant numbers of divisions even when it lacked support from other parties. This will become evident when we turn to Tables 6.4 and 6.5.

Table 6.3 also provides several other interesting insights into Senate voting patterns. In each of the six years except the first (1996), for instance, the government voted only with the Opposition more often than it voted only with the Democrats; in 2000 and 2002, government-Opposition coalitions were more than twice as frequent as government-Democrat coalitions. These data are consistent with those in Table 6.1, and they are inconsistent with the characterization of the Democrats as the small centrist party, akin to the FDP in Germany, which is closer ideologically to each of its major rivals than they are to each other. It is hard to think of the Democrats as holding the ‘balance of power’ in the Senate when they have voted either with the government or the Opposition less often than the government and the Opposition have voted with each other.[123] Second, there were few divisions, almost none in four of the six years, in which the government, Opposition, and Democrats voted together, leaving the Greens isolated in opposition to the other three parties. This may say something about the Greens’ bark and their bite; it also suggests that the Greens have not looked for opportunities to force divisions that would clearly differentiate them from the other parties.

Table 6.3: Voting patterns in Senate divisions, 1996–2001

 

1996

1997

1998

1999

2000

2001

Government, Opposition, Democrats and Greens

1
(0.5%)

0

0

0

1
(0.9%)

0

Government, Opposition and Democrats

2
(1.0%)

13
(4.6%)

2
(0.9%)

13
(5.8%)

1
(0.9%)

1
(1.5%)

Government, Opposition and Greens

0

1
(0.4%)

1
(0.5%)

2
(1.2%)

1
(0.9%)

0

Government, Democrats and Greens

3
(1.5%)

6
(2.1%)

6
(2.8%)

4
(1.9%)

4
(3.5%)

6
(9.1%)

Government and Opposition

23
(11.7%)

55
(19.6%)

46
(21.3%)

64
(28.6%)

571
(49.6%)

34
(51.5%)

Government and Democrats

472
(23.9%)

36
(12.9%)

10
(4.6%)

53
(23.7%)

24
(20.9%)

113
(16.7%)

Government and Greens

0

8
(2.9%)

0

34
(1.3%)

0

15
(1.5%)

Government alone

121
(61.4%)

161
(57.5%)

151
(69.9%)

84
(37.5%)

27
(23.5%)

13
(19.7%)

1 Includes all 15 divisions on the Defence Legislation Amendment (Aid to Civilian Authorities) Bill 2000.

2 Includes 38 of 41 divisions on the Workplace Relations and Other Legislation Amendment Bill 1996.

3 On one division, the Democrats’ vote split and the government did not win.

4 The government did not win two of the three votes.

5 The government did not win the vote.

Note: This table does not reflect the votes of Independent Senators. For purposes of this table, the single Senator representing Pauline Hanson’s One Nation is treated as if he were an Independent. The Greens are counted as having voted with the government in cases in which either the Australian Greens or the Greens (WA) did so. The Australian Democrats are counted in the same way in cases in which they split the votes.

What is perhaps most striking about the data in Table 6.3, though, are the apparent trends in the frequency with which the government has voted alone and the frequency with which it has voted with the Opposition. Admittedly, with data for only six years we cannot distinguish with certainty between ephemeral phenomena and long-term patterns. Nonetheless, it is interesting, to say the least, that there has been an increase each year in the frequency with which the government and Opposition have voted together. In 1996, the government and Opposition voted together and against the two smaller parties only 11.7 per cent of the time. During 1997–1999, the comparable percentages climbed toward 30 per cent but then jumped to roughly 50 per cent in 2000 and 2001. For those who perceive the two parties to be moving toward the policy centre and narrowing the differences between them, here is supporting evidence that is dramatic if only indicative.

An alternative explanation is that what we are witnessing here is evidence of the process of adjustment on the part of a new party (actually, coalition) in government and a new party in Opposition. Since 1996 was the first year in office for the Howard Government, it may not have been particularly anxious to find areas of agreement with the Labor Opposition. What is perhaps more likely is that the ALP, now adjusting to being in Opposition, may have been particularly reluctant to support the government that had just driven it from office. Perhaps with time, one or both sides became more willing to look toward the other as a coalition partner. None of these possible explanations are mutually exclusive, and it would be in the nature of complex human behaviour if there were some degree of truth in all of them. At the same time, as Table 6.3 reveals, there were fewer occasions on which the government found itself opposed by all the other parties. There is a hiccup in the percentage for 1998; otherwise, the frequency with which all three other parties voted against the government declined from more than 60 per cent in 1996 to less than 20 per cent in 2001.

Tables 6.4 and 6.5 focus on divisions that the government won. The first of these tables presents data on how often the government won in each of three ways: (1) with the support of the Opposition, without regard to how the other two parties voted; (2) with the support of one or both of the minor parties, but not the Opposition; and (3) without the support of any of the other three multi-member parties. The last possibility deserves a word of explanation. If we refer back to Table 6.3 and add together for each year all the times the government won by voting with one or more of the other parties, we find that, in 1996–1999, that total does not account for all the government’s victories (from Table 6.1 or 6.4). On the remaining occasions, therefore, the government won with the support of a sufficient number of the remaining Senators: the Independents and the sole Senator affiliated with Pauline Hanson’s One Nation.

When we examine Table 6.4, we find, as Table 6.3 would have led us to expect, a growth in the frequency with which the government and Opposition voted together, with or without one or both of the other parties. The Opposition was part of the government’s winning coalitions more than half the time in 2000 and 2001, compared with less than a quarter of the time in 1997 and 1998, and even less often in 1996. Although the trend line is not perfect, it is so clear that it would be remarkable if the pattern of these data was accidental. The table also documents a fairly stable percentage of government victories attributable to the support it received from (or gave to) one or both of the multi-member minor parties: the Democrats and the Greens. In five of the six years, the percentage of divisions that the government won with the support of one or both of these parties, but not the Opposition, varied between 17.9 and 25.9, which is a reasonably consistent record, as such things go. The exception was 1998, when the frequency of such victories fell to only 7.4 per cent.

Table 6.4: Results of Senate divisions, 1996–2001

 

1996

1997

1998

1999

2000

2001

Number of divisions1

197

280

216

224

115

66

Divisions that the government won

130
(68.4%)

165
58.9%)

152
(70.4%)

164
(73.2%)

88
(76.5%)

51
(77.3%)

Divisions on which the government and the Opposition voted together

26
(13.2%)

69
(24.6%)

49
(22.7%)

80
(35.7%)

60
(52.2%)

35
(53.0%)

Divisions that the government won with the support of one or both minor parties, but not the Opposition

50
(25.4%)

50
(17.9%)

16
(7.4%)

58
(25.9%)

28
(24.3%)

16
(24.2%)

Divisions that the government won with the votes of Independents only2

543
(27.4%)

46
(17.7%)

874
(40.3%)

26
(11.6%)

0

0

1 Excluded are one free vote in 1996 and nine in 1997, all of which pertained to the Euthanasia Laws Bill 1996.

2 ‘Independents’ include the One Nation Senator.

3 Includes 11 divisions on the Telstra (Dilution of Public Ownership) Bill 1996 and eight divisions on the Higher Education Legislation Amendment Bill 1996.

4 Includes 12 divisions on the Telstra (Transition to Full Private Ownership) Bill 1998 and 16 divisions on procedural matters on the same or preceding days.

It is the bottom two rows of the table that are most striking. These present the number and percentage of government victories that cannot be explained by support the government received from other multi-member parties, and so are attributable to support by the Independent and One Nation Senators. We saw in the previous table that the government voted without the support of any of the other three parties almost 70 per cent of the time in 1998. Perhaps it is no coincidence that this was an election year, so all the non-government parties may have been more inclined than usual to oppose the government, and to demand divisions showing that the government lacked the support of any other party. We see in Table 6.4 that the dip in government-minor party winning coalitions in 1998 was offset by the remarkable spurt during that year of instances in which the government did not rely on any multi-member party for its victories. The frequency of such winning coalitions jumped by more than twenty percentage points, to reach 40 per cent of all successful government coalitions, before dropping back by almost 30 percentage points in the following year. Winning coalitions of government and Independent Senators then disappeared entirely in 2000 and 2001 for a simple and sufficient reason that emerges from an inspection of Table 6.2. In those years, even if the government had the support of both Independents and the One Nation Senator, that brought it only 38 votes, leaving it one crucial vote short of a majority. We will return to this point below.

Table 6.5: Winning government coalitions in Senate divisions, 1996–2001

 

1996

1997

1998

1999

2000

2001

Winning government coalitions including the Opposition

26
(20.0%)

69
(41.8%)

49
(32.2%)

80
(48.8%)

60
(68.2%)

35
(68.6%)

Winning government coalitions including one or both minor parties, but not the Opposition

50
(38.5%)

50
(30.3%)

16
(10.5%)

58
(35.4%)

28
(31.8%)

16
(31.4%)

Winning government coalitions including Independents only

54
(41.5%)

46
(27.9%)

87
(57.2%)

26
(15.9%)

0

0

Table 6.5 highlights some of the developments we have noted by presenting the number and percentage of winning coalitions that the government formed that included (1) the Opposition, regardless of whether they also included either or both minor parties; (2) one or both minor parties, but not the Opposition; and (3) Independents only (again, treating Senator Harris of One Nation as if he were an Independent). The table reveals an increasing frequency of winning government coalitions with the Opposition and a steady rate of winning government coalitions with minor parties; however, 1998 was the expected exception to both patterns. More generally, all these tables show a government and an Opposition that frequently voted together. These data are at odds with the simplistic notion that the role and responsibility of the Opposition is to oppose. What is more important for Australian politics, however, is that the way in which the Coalition and the ALP vote in the Senate chamber is quite different from the way in which the two parties portray themselves and each other to the electorate.[124] But what happens on those occasions when the Opposition not only opposes the government, but tries to construct winning coalitions of its own?

The Opposition’s winning coalitions

All of the discussion thus far in this chapter has proceeded from the perspective of the government. The primary burden for forming winning coalitions falls on the government because it is the government, not any of the non-government parties, that has the responsibility to initiate legislation and that also is expected to pass most of it. Any government that failed to do so, bemoaning the fact that it lacked ‘the numbers’ in the Senate, would not remain the government for long. It is also the government that exercises almost total control over the legislative agenda in the Senate as well as in the House of Representatives, and so it is the government that has the incentive to engage in successful coalition-building because it is the government’s own legislation and legislative record that depends on it.

There is another side to the story, however. The non-government parties in the Senate, and especially the Opposition, have their own incentives to construct winning coalitions in order to improve (from their point of view) the government’s legislation or, alternatively, to block that legislation or secure passage of motions that are critical of the government, and thereby thwart, embarrass, and discredit it. So this analysis would be incomplete if we failed to look at Senate voting coalitions from the perspective of the non-government side of the chamber. In practice, this means examining the Opposition’s options and record because it is sensible to assume that it is the Opposition, not the far smaller minor parties, that will take the lead most often in bringing other parties into coalition with it on individual bills and votes. Certainly the Senate Democrats or Greens must approach the Opposition to solicit its support for their motions and amendments, but I am prepared to assume that it is usually the Opposition that acts as prime mover in attempts to build winning coalitions against the government.

Because of its smaller numbers, the Opposition’s prospects for creating such winning coalitions have been more constrained than the government’s. The essential facts for the Opposition in the Senate throughout the six-year period, as a review of Table 6.2 reveals, is that it could not win without the votes of the Australian Democrats and it could not win with only the votes of the Australian Democrats. In fact, during the 84 months of 1996–2002, there were only 33 months when the ALP, Democrats, and Greens together could form a majority, and then without a single vote to spare. These numbers drive home how important it is that, in recent decades, there always has been a non-government majority in the Senate, but not an Opposition majority.

The first row of Table 6.6 presents data on the number and percentage of divisions that the government lost each year. These data cannot be compared on a year-by-year basis with the total numbers of Labor, Democrat, and Green Senators because, as Table 6.2 demonstrates, that number sometimes changed in mid-year when newly-elected Senators were sworn in or when sitting Senators left the parties to which they had belonged. But Table 6.2 also indicates that the only times during 1996–1999 when the three parties together held 39 seats (they never held more than that) were during January–June 1996 and during July 1999–September 2001. If we examine Table 6.6 with this fact in mind, it is striking that the Opposition was not most successful when its representation, combined with that of the Democrats and Greens, was greatest.

The government had its worst losing percentage in 1997 when the three non-government parties held a total of 37 seats and could win only with the support of both Independent Senators. And in 1999–2001, when there were a total of 39 Labor, Democrats, and Green Senators during most of that time, the three parties combined to defeat the government on less than one-quarter of all divisions. Table 6.6 also shows that, except for 1997, there was a steady decline each year in the percentage of divisions that the government lost, though the percentage changes from year to year are too small to ask this apparent trend line to carry too much analytical weight. In hindsight, it may become clear that what these data really are showing is a fairly consistent record of government losses, varying from 22.7 per cent to 31.6 per cent, which is a fairly narrow range for such phenomena, with 1997 being the obvious exception.

We can gain some purchase on these phenomena by referring to the second half of Table 6.6, which shows the frequency with which the Opposition joined with neither, one, or both of the two minor parties to oppose the government on divisions. These data are the same data that are presented in Table 6.3, but from the reverse perspective. Table 6.3 shows, for instance, that on 61.4 per cent of the divisions in 1996, the government voted alone—that is, without the support of the Opposition or the Australian Democrats, or the Greens. This can only mean that on that same percentage of divisions, the Opposition, Democrats, and Greens voted together. Similarly, when the government, Democrats and Greens voted together (see Table 6.3), the Opposition voted without the support of any other party (see Table 6.6).

Table 6.6: Opposition coalitions on divisions, 1996–2001

 

1996

1997

1998

1999

2000

2001

Divisions that the government lost

67
(31.6%)

115
(41.1%)

64
(29.6%)

60
(26.8%)

27
(23.5%)

15
(22.7%)

Opposition voting coalitions against the government:

           

Opposition, Democrats and Greens

121
(61.4%)

161
(57.5%)

151
(69.9%)

84
(37.5%)

27
(23.5%)

13
(19.7%)

Opposition and Democrats

0

8
(2.9%)

0

3
(1.3%)

0

1
(1.5%)

Opposition and Greens

47
(23.9%)

36
(12.9%)

10
(4.6%)

53
(23.7%)

24
(20.9%)

11
(16.7%)

Opposition alone

3
(1.5%)

6
(2.1%)

6
(2.8%)

4
(1.9%)

4
(3.5%)

6
(9.1%)

Opposition voting with the government

26
(13.2%)

69
(24.6%)

49
(22.7%)

80
(35.7%)

60
(52.3%)

35
(53.0%)

We see from the last row of Table 6.6 an almost unbroken pattern of increases in the frequency with which the Opposition voted with the government to form winning grand coalitions (which may or may not also have had the support of other parties or Independents). We also find that the frequency of government-Opposition coalitions was considerably higher in 1999 than in 1998 and then much higher in 2000 and 2001, even though the Opposition could defeat the government during most of 1999–2001 by joining with the Democrats and the Greens, something that was not possible in 1997 and 1998. In other words, it was when the Opposition seemingly had the best chances to form winning coalitions against the government that it was voting more often with the government.

Why? Data cannot provide a conclusive answer to the question, but they do suggest several possibilities or subsidiary questions. Note, for instance, how infrequently the Labor Opposition voted with the Democrats but not also with either the Greens or the government. In itself, this is not surprising because a vote on which the Opposition voted only with the Democrats (among the multi-member parties) is, by definition, also a vote on which the government voted only with the Greens, a strange bedfellows combination under most circumstances. The frequency of Opposition-Democrat pairings was consistently low, while the frequency of Opposition-Green pairings was considerably higher and fairly stable, ranging from almost 13 to almost 24 per cent, with 1998 being the clear exception. So there are no marked changes over time in the frequency of either pairing, and there certainly is no marked direction of change in either case.

But now consider the frequency of three-party anti-government coalitions. In 1996–1998, these coalitions formed on more than half of all divisions, though in the latter two years, even the three parties voting together did not form majorities without the support of one or more Independent Senators. Then from 1998 to 1999, the frequency with which the three non-government parties voted together dropped abruptly from 69.9 per cent to 37.5 per cent and continued to fall in the next two years to 23.5 and then to only 19.7 per cent. These data suggest, though they certainly cannot prove, that the Labor Opposition has encountered a frustrating dilemma. At precisely the time that the support of both the Democrats and the Greens could give Labor the numbers for victory, it was becoming harder and harder to bring all three parties together in support of the same positions, even if those positions were nothing more than an agreement to vote against the government’s positions. One interpretation that fits the data is that Labor was caught between a pull toward the left from the Greens and a pull toward the center from the Democrats. Whatever policy changes made it easier for the Labor Opposition to reach agreement with one of the minor parties made it more difficult to reach agreement with the other.[125]

The data in Table 6.6, when viewed in light of what we know from Table 6.2 about party representation in the Senate, also point unquestionably to the important, even pivotal, role that Independents (including Senator Harris of One Nation ) have played in determining the outcomes of Senate votes in recent years. From mid-1996 through mid-1999 and then again from October 2001 to the present, the votes of one or more of these Senators have been required to win divisions that the government opposed. From the government’s perspective, its larger numbers always have enabled it to form majority coalitions without any Independent votes. It also is noteworthy, though, that from September 1996 to June 1999, the government could win without any of the other parties if it had the support of both Independents, Senators Colston and Harradine. Much the same situation has prevailed after August 2002, when the government could reach the magic number of 39 whenever it could combine its 35 votes with the votes of Senators Harradine, Lees, and Murphy (Independents) and Senator Harris (One Nation). Such things are much easier said than done, however, and especially so in light of the fact that two of the Independents (Harradine and Murphy) had been members of the ALP—though Harradine ceased being a member of the Labor Party before his election to the Senate—and the third (Lees) had been the Australian Democrats’ leader in the Senate. The numbers make coalitions possible; skill and good fortune are required to make them happen, and even the most adept coalition builder cannot build a coalition that bridges unbridgeable policy and philosophical differences.

Minor parties and the balance of power

There are two conventional understandings about the political situation in the Senate, as it has been in recent years and as most observers expect that it will continue to be. First, there is a non-government majority in the Senate; and second, the minor parties and Independents, or some combination of them, hold the balance of power in the Senate. The first assertion is unquestionably true. The second is not as obviously and completely true as it might seem at first blush and deserves some final words here, although this matter of the balance of power will arise again in the next chapter.

One non-trivial implication of our analysis of Senate coalitions is that the minor parties —individually or jointly, and with or without the votes of Independent Senators—only hold the balance of power in the Senate when the government and the Opposition oppose each other (Young 1997). As we have seen, they have not always done so. In fact, the government and the Opposition have voted together quite often, and when they have done so, the votes cast by the Democrats, the Greens, and the Independents have mattered not at all, at least for the purpose of determining the outcomes.[126] So in this important respect, the leverage of the minor parties in the Senate is more limited than it might seem if we were simply to look at the numbers in Table 6.2 and assume that the government will try to govern and the Opposition will oppose.

There is another respect, though, in which a minor party in the Senate can be constrained in how and when it uses the leverage that its voting power in the Senate gives it. This can be a self-imposed restraint reflecting that party’s view of what role is constitutionally appropriate for it to play and what role is politically advantageous for it to play. In the case of the Democratic Labor Party during the 1950s–1970s, its Senators probably lost little sleep over these matters. They usually (but not invariably) voted with the Coalition and had little interest in somehow maintaining a balance between the Coalition’s influence over policy and that of the ALP from which the DLP had split. In the case of the Greens, its Senators have been too few and its voting leverage in the Senate too limited for its views on these matters to be very consequential—at least not yet.

The Australian Democrats are another matter, however. In some ways, they have been the quintessential minor party: ideologically moderate, positioned in policy terms between the two political behemoths, particularly interested in issues of process as well as policy, and sometimes able to determine outcomes by their choice of which of the major parties to support as well as by their own amendments to government bills.[127] Although the recent fissures within the ranks of the Democrats’ Party group have called this characterization into question and even raised questions about the party’s political viability, it still is worth examining how the Democrats as a party have conceived of their place in the Senate and in the constitutional order.

In calculating how to take advantage of a position of strength in the Senate, a minor party must weigh its natural desire to use its ‘balance of power’ leverage to promote its own policy agenda while also demonstrating the public value of the Senate seats it holds against two offsetting considerations: first, its acceptance of the principle that the responsible government elements of the Constitution ultimately do give the government a strong claim to establish the parameters of national policy, if not all its fine details; and second, the minor party’s concern that, should it fail to give such deference to the government, the voters could penalize it for exceeding what they think, in however inchoate a fashion, are the legitimate uses of its power. The Democrats’ recognition of the balance involved was encapsulated in a statement by Senator Cheryl Kernot, then Leader of the Democrats in the Senate, concerning a 1993 tax bill:

These [tax increases] are not our priorities; they are this government’s priorities within the parameters of a straightjacket of its own making. Nevertheless, the final shape of the budget is its prerogative, and it will live with the consequence. The Democrats derive some satisfaction from being able to inject fairness into the original budget proposals, and I think it shows that minor parties can make a significant contribution and can achieve significant change.

Beyond that, we respect the government’s right to govern and to make the tough decisions on how the budget shapes up. If it were up to us, we would have done it differently. (Commonwealth Parliamentary Debates (Senate), 7 October 1993: 1818)

As this statement suggests, the Democrats then felt that they were constrained by what they perceived to be the appropriate constitutional rules of the game. These rules, as they understood them, limited what uses and how much use they could legitimately make of their leverage in the Senate and were rooted in the premise that, notwithstanding the powers that the Constitution gives to the Senate, ultimately the exercise of those powers must not undermine the relationship of responsibility between the government and the House and, through the House, between the government and the electorate.[128] Young (1997) analyzes the changes in the 1993 budget that Keating’s ALP Government made in order to secure the votes it needed from the Democrats and the Greens because of an essential strategic fact: the understanding of all parties that the Coalition was committed to opposing the government’s tax proposals. This fact obviously strengthened the hands of both the Democrats and the Greens. The Greens pushed somewhat harder and further than the Democrats, and extracted more concessions from the government. However, neither pushed as hard or as far as it might have, given how important it was to the government to have the Senate pass bills that were an essential part of its budgetary scheme.

Young (1997: 70) identified reasons of principle and political prudence for self-restraint by the Australian Democrats:

Put simply, when it came to wielding the balance of power on policy issues, the Democrats were extremely cautious in their approach as they sought to ensure that their actions could not be interpreted as undermining governability. Had the Democrats been seen as obstructing the Budget, or forcing the government to a double dissolution election, the negative fall out could have been significant and potentially undermined their position as balance of power holders. The experience of the Democrats in the Senate had sensitised them to this possibility and thus they sought to accommodate their need to be viewed as playing an active role within the Senate without opening up the party to claims of obstructionism.

With memories of the events of 1975 still so vivid in so many minds, the fact that it was budget-related bills that were at stake both increased the Democrats’ leverage and made them particularly wary about using it too forcefully. In her statement quoted above, Kernot emphasized a motive of principle; in her analysis, Young emphasized the Democrats’ worry about how the public would perceive and accept their actions. Both concerns undoubtedly were real, and they combined to impose limits on the negotiating leverage of the Democrats (and the Greens), limits that would not have existed if the only things to be calculated were who had ‘the numbers’ in the Senate and who needed them the most.

Goot (2002: 42) shows that the average major party vote in Senate elections has declined steadily, decade by decade, since the 1940s: from 95.3 per cent to 92.0 per cent to 88.3 per cent to 86.7 per cent to 84.4 per cent to 80.5 per cent. As investment counselors remind us, past performance is no guarantee of future results. If this trend continues, however, the number of minor party or Independent Senators, or both, will almost have to increase, as might the number of different minor parties that secure representation in the Senate. The difficulty, though, is in predicting just what the consequences of this development would be for governments and Oppositions engaged in the task of trying to win Senate votes. On the one hand, a greater fragmentation of party representation in the Senate would complicate the task of coalition-building because there would be more party groups or Independents with whom to consult and perhaps negotiate. At present, coalition-building in the Senate is largely an exercise in wholesale politics; successful negotiations produce Senate votes in bulk. A decline of two-party representation in the Senate (always for our purposes treating the Liberal and National parties as one) would require more retail politics, negotiating compromises or securing support, one vote or a handful of votes at a time.

On the other hand, that same fragmentation could well create a greater array of majority coalition options. There have been times during the years covered by this study when it was difficult, if not practically impossible, for the government to win in the Senate without the support of either the Opposition or the Democrats. And at all times, it was mathematically impossible for the Opposition to win a Senate division unless the Democrats were allied with them. If, let us imagine, the Australian Democrats decline as a force in the Senate, as some have speculated they are likely to do, their voters would have several options: supporting one of the major parties; supporting another, existing minor party—i.e., the Greens; supporting one or more new minor parties to emerge from the ashes the Democrats leave behind; supporting independent Senate candidates; or, what is most likely, some complex combination of the above.

Which tendency is strongest could have a significant effect on the dynamics of decision-making in the Senate. If the Democrats’ vote fragments among Independents or new minor parties or both, both the government and the Opposition, depending on their numbers, of course, might be able to piece together winning majorities from among a larger number of smaller political forces in the Senate. Perhaps ironically, the greater the fragmentation in the Senate, the easier it might be for governments and Oppositions to build majorities because of the greater number of potential coalition partners, albeit small ones, among whom to choose to negotiate. The task certainly would be difficult and frustrating, however, if it required either major party to satisfy all or almost all of the Senate’s minor parties and Independents.

If the Greens were to be the primary beneficiary of a decline or disappearance of the Democrats in the Senate, that would likely have a differential effect on the two major parties, whichever is in government. In light of the Greens’ expressed policy views and their expressed opposition to what US analysts call logrolling—offering their support to another party on one issue or vote in exchange for an implicit or explicit assurance of that party’s support on another issue or vote—we could expect that alliances between the Greens and the Liberal-National Coalition would be more difficult to form and less frequent than alliances between the Greens and Labor. Depending as always on the numbers, this situation could leave a Liberal-National government dependent on the support of Independents and whatever other Senate votes remain to be found. At the same time, it could make it somewhat easier for the ALP, whether in government or in Opposition, to join with the Greens to form majorities on Senate votes, but only at the expense of being pulled to the political left. (This assumes that it is fair to place the Greens to the left of the ALP and both of them to the left of the Liberals and Nationals on the proverbial unidimensional left-right spectrum.) To put it bluntly, a Labor government might find itself hostage to the Greens, whereas a Liberal-National government might become hostage either to the ALP Opposition or to everyone else in the Senate except the Greens. In the Senate, the number of seats not held by the major parties is critically important, but so too is who sits in those seats.

Navigation: Previous PageContents | Next Page

Back to top


Facebook LinkedIn Twitter Add | Email Print