Can the Senate Claim a Mandate?


Murray Goot

In the course of the 1996 election campaign for the House of Representatives and half the Senate, the Liberal-National Coalition committed itself to the partial privatisation of Telstra and to putting part of the proceeds of the sale into an environmental fund. The then Labor government campaigned against this proposal; so, too, did the Australian Democrats, the Greens and an Independent who, between them, were to hold the balance of power in the new Senate. Having won the election, the Coalition claimed a mandate to legislate for the part-privatisation of Telstra. The Labor Party, however, insisted that the government had won no such mandate. The Australian Democrats, going one step further, claimed that the opposition parties in the Senate had secured a `counter-mandate'—an electoral mandate to oppose the privatisation of Telstra.

The government based its claim on the fact that, since its plans for Telstra had been made known in advance of the election, voters had been afforded an opportunity to pass judgment on them. Having won what the new Prime Minister, John Howard, called `a huge mandate' [1]—the majority (93 out of 148) of the seats in the House of Representatives, a plurality (46.9 per cent) of the first preference vote and the majority (53.6 per cent) of the `two-party preferred' [2]—the government was entitled to expect that the policies on which it had gone to the people would be allowed a safe passage not only through the lower house, but through the upper house as well.

After an early by-election saw the government strengthen its hold on the seat of Lindsay, won from Labor at the general election, the Prime Minister called again on the opposition forces in the Senate to let the Telstra legislation through. `One of the messages coming out of [the by-election] loudly and clearly,' Howard told the Liberal Party's Federal Council, `is “we didn't make a mistake, we wanted a change and for those who would endeavour, unreasonably, to frustrate that change get out of the way and let the new crowd have a fair go.” ' [3]

The Democrats, by contrast, focused on the fact that while the Coalition had won a clear victory in the House of Representatives, it had not done so in the Senate; and the parties in the Senate threatening to block the government's Telstra Bill had stated their opposition to such legislation every bit as clearly as the Coalition had stated its support. The difference in the balance of forces in the two chambers could best be understood, the Democrats argued, in terms of voters hedging their bets. Public opinion polls not only showed that most voters were opposed to the sale of Telstra, they suggested widespread support for the Senate's right to block the sale. According to Senator Cheryl Kernot, the Democrats' leader, the electorate had taken out `insurance'. It had `deliberately provided two competing mandates: one for the government to be changed and one for a balance of power check on that new government in the Senate.' [4]

The result of the Lindsay by-election, Kernot insisted, reflected nothing more than the electorate's desire to give the Liberal candidate (forced by Labor to re-contest her seat) a `fair go'; it didn't give the government a `blank cheque'. [5] Similarly, Senator Brian Harradine (the Tasmanian Independent whose vote on Telstra would be crucial) said: `Let me see any exit poll that says that the one-third sale of Telstra was the uppermost thought in the minds of the electors' at the by-election. [6] None of the polls, however, had made much attempt to find out what was `in the minds of the electors'. [7]

What can be said about these diametrically opposed positions? In what sense can a government be said to have a mandate? Under what conditions, if any, might the Senate legitimately impose its own policies on a government? And to what extent might the answers given to these questions today differ from those given in the past?

Mandates, bicameralism and the meaning of the vote

In coming to grips with these questions, this paper starts by arguing the need to consider: the distinction between a government's claiming a specific mandate and its claiming a general mandate; the peculiar difficulties that strong bicameralism creates for any theory of mandates; and the means of validating key assumptions about the meaning of the vote that mandate theories generally make.

Mandates: specific and general

A government which argues that its election signifies the electorate's endorsement of at least some of the policies on which it campaigned appeals to a specific mandate. In its most plausible form, such a claim revolves around one issue, and in these circumstances, elections are sometimes described as referendums. But appeals to specific mandates can involve claims about several policies, even entire programs. As Gough Whitlam recalled, `We took the view that we had been given a specific mandate to implement each part of the program set forth in the 1972 policy speech.' [8]

By contrast, a government which argues for the legitimacy of its actions simply on the grounds that it won the preceding election seeks to found its authority on the idea of a general mandate. Associated famously with Joseph Schumpeter, what elections determine on this view is not what policies are to be pursued, but who is to govern. [9] That Australian governments have a general rather than specific mandate is, Don Aitkin believes, precisely the view of the electoral process that voters themselves take: `their job', as they see it, `is to put governments in and let them get on with the job; if a government does its job badly they will eventually “turf it out” and put the other lot in.' [10]

Unicameralism and bicameralism

If the Australian parliament were unicameral, the question of whether the government had a specific mandate or a general one would be a relatively minor matter. Provided it had a majority in the House of Representatives (something all Australian governments have enjoyed since 1941), a government would dominate the parliament through sheer weight of numbers.

But the parliament is strongly bicameral and for much of the recent past the governing party of Coalition in the House of Representatives has found itself in the minority in the Senate. [11] Strong bicameralism—where the powers of both chambers to initiate or block legislation are roughly the same, where the compositions of the two chambers are constituted under different sets of rules and where the dominant party or parties in one is generally not the same as the dominant party or parties in the other—is likely to make a government's claim to any sort of mandate much more difficult to sustain.

Those who defend the pre-eminent position of the lower house in matters of legislation do, of course, have a case. Where upper houses are unelected, the case for giving them any sort of veto over legislation is likely to be especially weak. But it is misleading to claim, as Arend Lijphart does, that second chambers that are not directly elected necessarily `lack the democratic legitimacy, hence the real political influence, that popular election confers.' [12] In Australia, for much of this century, state upper houses elected on a property franchise—or, in the case of New South Wales pre-1978, not elected on an extra-parliamentary franchise of any kind—wielded powers largely identical to those enjoyed by lower houses. [13] Where public opinion itself supports, or at least acquiesces in such an arrangement, second chambers that are not subject to popular election can be as influential as any others.

There are two other conditions under which the claims of upper houses are also likely to be weakened. One is where malapportionment is greater in the upper house than in the lower house. The other is where elections for the two chambers are held at quite different times, so that the lower house finds itself confronted by an upper house elected several years earlier.

The Australian Senate, at first glance, appears to fail the first test: each state, whatever its number of electors, sends the same number of senators to Canberra; and the fact that a senator from New South Wales represents roughly 12 times as many voters as a senator from Tasmania not only violates the ideal of one vote, one value, but does so to a much greater extent than any present—or past—malapportionment in the House of Representatives. Those intent on upholding `sound democratic principles' and defending the primacy of a government's mandate have traditionally made much of this, especially from the Labor side. [14] But closer scrutiny shows something quite different. If one calculates election results on a national rather than state basis (so that all votes count equally regardless of where they were cast) and then compares the votes cast with the seats won, it is the Senate that better mirrors the nation's mind almost always—not the House of Representatives; in 1996, for example, Labor, the Democrats and the Greens won 50.2 per cent of the votes and 50 per cent of the seats in the Senate while in the House of Representatives they won 50.7 per cent of the vote but only 33 per cent of the seats.

It is on the second test that the Senate has, on occasion, truly been found wanting. The Whitlam government, for example, elected at the end of 1972, was rightly affronted by the fact that the senators who obstructed its early legislation and forced it to the polls in 1974 had been elected in 1970 or 1967. On that occasion the Senate was not, in E.T. Brown's words, `less than half fresh and more than half stale'; it was entirely stale. [15] After a double dissolution, the composition of the Senate is as fresh as the composition of the House; after a half-Senate election, held (as in 1996) in conjunction with an election for House, the in-coming Senate is at least half as fresh. And since there is no reason to suppose that, where a half-Senate election is held in conjunction with an election for the House of Representatives, the parties' share of the vote would be any different to the share they would gain at an election for the full Senate, the distribution of votes for the two chambers can be regarded as equally fresh and directly compared.

After the 1996 election, the Democrats were careful to ground their claim to a mandate on the share of the votes won by opposition parties in the Senate, not on their share of the seats; `where mandate [sic] is concerned', Cheryl Kernot cautioned, `count the votes not the seats.' [16] The caution was well advised. According to David Butler, had there been a full Senate election, the Coalition would have won at least 38 of the 76 seats, and `almost certainly' picked up a seventh seat in Queensland to give it a majority of 39. [17] In effect, the support of Harradine (Independent, Tasmania) and the decision by Senator Mal Colston (Queensland) to quit the Labor Party for the cross-benches gave the Coalition, in highly unusual circumstances, what the electoral system might have delivered as a matter of course.

The meaning of the vote

Neither a general nor a specific mandate, Richard Mulgan insists, depends `on any conscious intention on the part of the voters'. Just as the general mandate follows automatically from the government's control of the lower house, so the convention of a specific mandate `follows from the inclusion of a policy in the government's election program, regardless of whether any voters knew of it, let alone whether their votes were determined by it.' [18] But most academic commentators would disagree. As Stanley Kelley has put it: `To suggest that a mandate exists for a particular policy is to suggest that more than a bare majority of those voting are agreed upon it.' [19]

When the idea of a government winning a policy mandate first took hold, after the Reform Act of 1832, there was no clear way of knowing the electorate's state of mind. A century later, this began to change as first pollsters and then political scientists brought the methods of survey research to bear. On the question of whether issues determined election results, political scientists were damning. In Australia, a 1958 by-election study reported that issues played a `relatively small part in the choice of the majority of voters'; another study, finding voters `almost incredibly ignorant', suggested that `most people support a party not because it favours a particular policy but because they think it is made up of particular types of people ... people like themselves.' [20]

But surveys did not necessarily damage the idea that a party's policies might have majority support. Some policies appeared to enjoy this support; others did not. Increasingly, polls that point to a government's policies being widely supported have strengthened its legislative position. The power of the polls is evident not only when the policies in question are those on which the government has sought an electoral `mandate', but when the policies it proposes run counter to those on which it campaigned. Consider, for example, the Howard government's decision to introduce a `work-for-the-dole' scheme, after Howard himself appeared to have ruled it out during the campaign. A Morgan poll, which reported 72 per cent in favour of the dole scheme, suggested an electorate that either did not know or did not care about the original promise. And the legislation passed the Senate after Labor, influenced by this and other survey research, shifted its position from opposition to support. [21]

Equally, polls have been used to undermine a government's legislative position by suggesting that its parliamentary majority may have been achieved despite its commitment to certain policies. In 1996, the Opposition's `counter-mandate' on Telstra was said to have been made manifest not only by the election results for the Senate, which went against the government, but also by opinion polls.

Evidence from the polls

The message from the Lindsay by-election, the Prime Minister had insisted, was that the Senate should not `unreasonably' frustrate the new government. But how was the `reasonableness' of the Senate's behaviour to be judged? Thirty years earlier, a student of the Senate had suggested that `[i]n the final analysis, the view of the public' was `the best measuring stick to determine the reasonableness of the Senate's record'. And the record clearly showed an electorate capable of siding with, or at least not punishing, opposition parties whose actions in the Senate precipitated another election. In 1913, for example, the Cook government met its demise after the Senate, controlled by Labor, forced it to the polls; in 1931, the defeat of the Scullin Labor government followed a long period of Senate obstruction; and in 1975, Labor was again swept away when a stalling Senate persuaded the Governor-General to demand a new election. [22]

The idea that `key' government legislation ought to be passed by the Senate, as a matter of principle, is not one to which most voters appear to subscribe. Polled in 1997 about whether the present method of electing senators, which `some believe ... makes it too easy for smaller parties to control the balance of power in the Senate', was `a bad thing because the elected government can't be sure of getting key policy changes approved in the Senate' or `a good thing because the Senate can keep a check on government policy', few of those interviewed were prepared to write a blank cheque for either side: only 10 per cent agreed that `when the government does not have a majority in the Senate' it is `always a bad thing'; nearly twice that number (18 per cent) thought it `always a good thing'; while two-thirds (67 per cent) said it was `sometimes good and sometimes bad depending on the circumstances'. Nearly three-quarters (72 per cent) thought the electoral system `should stay the way it is' rather than be changed `to make it easier for the coalition or the Labor Party to have a majority in the Senate.' [23] Asked, after the most recent election, whether it was `better' when the `federal government has a majority in both the House of Representatives and the Senate or when the federal government in the House of Representatives does not control the Senate', only four in every ten respondents thought it `much better' (20 per cent) or at least `better' (19 per cent) when the government controlled both; nearly half thought it `better' (38 per cent) or `much better' (10 per cent) when the government did not control both. [24]

In 1996, the willingness of voters to hedge their bets was quite marked. The outward sign of this was the gap between the Democrats vote in the House (6.8 per cent) and its vote in the Senate (10.8 per cent). A difference of four percentage points was higher than at any election since 1983—another election at which a widespread expectation of a change of government may have encouraged voters to pull on the reins. [25] According to one post-election poll, 17 per cent in 1996 voted for one party in the House and a different party in the Senate—two-thirds because they did not want one party to control both. The corresponding figure for 1990, as estimated by the Australian Election Study, was 11 per cent; and for 1987, 15 per cent. [26]

For Kernot and the Democrats, the polls introduced two distinct, if related, kinds of evidence that bolstered the Party's stand: first, evidence of support for the Democrats' substantive position—that is, evidence against the government's claim that it had secured a specific mandate for Telstra's partial privatisation; second, and more importantly, evidence of support for the Senate's right to block the government's Telstra Bill. [27]

Evidence of the first kind came from three polls conducted during the course of the campaign and one after it. In two of the campaign polls, undertaken by AGB McNair, respondents were told that `Mr Howard' had `announced that $1.1 billion of the proceeds from the sale of one-third of Telstra' would `be used to fund his environment protection package' and then asked whether they would `support the sale of one-third of Telstra if some of the proceeds' were `used to fund environmental protection'; 49 per cent, in the first, and 41 per cent, in the second, said they would. In the third, conducted in between by Newspoll, respondents were asked whether they `personally' favoured or opposed `the Coalition's proposal to spend $1 billion on the environment funded by the partial sale of Telstra'; no more than 39 per cent said they favoured it. Some months after the election, AGB McNair told respondents that the government, as it had `announced during the campaign', had `introduced legislation into parliament to enable the sale of one-third of Telstra'; but the Coalition's victory notwithstanding, the level of support (51 per cent) for `the sale of one-third of Telstra if some of the proceeds are used to fund environmental protection' hardly represented a clear-cut endorsement. [28]

It does not follow automatically that those opposed to the Telstra sale would have thought the Senate within its rights to block the move. Before the 1998 election, for example, the number opposed to a goods-and-services tax exceeded the number who thought the Senate should block it. [29] But this was not the case with Telstra. Told, after the 1996 election, that both `Labor and the Democrats in the Senate' had `said they will refuse to pass the legislation enabling the sale of one-third of Telstra', a clear majority (56 per cent) of those interviewed by AGB McNair agreed that `irrespective' of whether they supported the sale, the Senate should have `the right to block' it. During the campaign, a similar proportion (64 per cent) of those polled by Morgan agreed that if Labor and the Democrats held `the balance of power in the Senate', they should have the right `to prevent the coalition from selling part of Telstra.' [30] Most (52 per cent) of those polled after the election rejected the Coalition's threat of an early double dissolution `if the Senate fails to pass important laws such as the one allowing for the partial sale of Telstra.' [31] Asked about the Prime Minister's view that the Liberal's by-election victory `means that the Senate does not have the right to stop the passing of legislation to sell part of Telstra', only a third of those polled agreed; more than half (54 per cent) disagreed. [32]

Support for the Senate's right to block government legislation—even budgets—was hardly new. In 1975, at the height of the most (in)famous stand-off between the House and the Senate, the majority (52 per cent) of those polled by McNair Anderson disagreed that the Constitution `should be changed so as to remove from the Senate the right to reject money bills', though nearly two-thirds (64 per cent) agreed that the coalition parties were `wrong in attempting to block the money bills in the Senate at the present time'. [33] In 1993, two-thirds (63 per cent) of those polled by AGB McNair Anderson supported the Coalition's attempt to block the Budget in the Senate; two-thirds (67 per cent) of those interviewed by Newspoll wanted the Labor government to `make further changes to the Budget', to overcome a parliamentary `deadlock', rather than having the two Green senators who were `delaying the budget' agree `to support the Budget as it is currently proposed'; and only a third (36 per cent) of those interviewed by Morgan about how they would vote in a referendum to remove the Senate's power to block supply said they would vote Yes. [34] In 1993, three-quarters (74 per cent) of those polled by Saulwick thought the actions by the Democrats and minor parties to `force changes in the budget' were `proper' rather than `not proper' and in 1995, a similar proportion (71 per cent) of those interviewed by AGB McNair—at a time when the Coalition, the Democrats and the Greens had `blocked certain aspects of the government's Budget'—agreed that `the Senate should be allowed to block parts of the Budget as it sees fit.' [35]

Evaluating the arguments

The first stone

One objection to the Opposition's attempt to block the Telstra legislation in the Senate was that it was hypocritical: Labor in office, it was observed, had railed against the Senate—the Prime Minister, Paul Keating, calling it `an unrepresentative swill'. [36] The point, though valid, was easily countered. Labor's leader in the Senate, John Faulkner, noted that after the 1987 election (triggered by the Senate's rejection of the Labor government's Australia Card Bill), John Howard himself, then Leader of the Opposition, had scorned the very idea of an electoral mandate. [37] According to Howard, any suggestion that Labor had won a specific mandate for its Australia Card legislation was `invalid not only in terms of the number of votes cast [fewer for the government than for the opposition] but also on the simple proposition that when people vote at an election they do not vote on only one issue.' Subjected to `proper analysis', he insisted, `the mandate theory of politics' had `always been absolutely phoney'. [38]

Expectations and obligations

Another objection hinges on the argument that there is a reciprocal relationship between the electorate's expectation that governments will keep their promises and the obligation of opposition parties to respect a government's mandate. Responding to the argument that a government could claim a mandate only for policies electors both knew about and voted for, Richard Mulgan insisted that on this reasoning governments could choose to `break any election promises'—something, he was sure, voters would not want them to do. [39]

But this line of argument overlooks the fact that voters sometimes do want promises broken—and not just (as mandate theorists readily concede) when circumstances facing governments change in unexpected and adverse ways. One reason why Mulgan sets so much store by the idea of a mandate is that he assumes that if governments `depart from their election policy, the political consequences in terms of unpopularity and damage to their credibility will be very great.' [40] But this is not necessarily so.

The argument also misconstrues the relationship between mandates and accountability. In politics, promise-keeping is a political act; it is not a question of morality or of logic. Charges of `inconsistency' or `hypocrisy' against a government for, say, keeping its promise on Medicare but not on Telstra are unlikely to carry much electoral weight. No opposition is ever likely to be blamed by voters for blocking an unpopular policy while badgering the government to deliver on a popular one.

And the argument ignores the reasons why parties make promises and why parties that lose often—but not always—try to ensure that promises are kept by the party that wins. Typically, parties promise things that they have reason to believe will win them support. If what is promised has wide support, opponents will not disparage the promise, but attempt to cast doubt on the likelihood of it being delivered—or at least on it being delivered without it exacting an unacceptably high price. Any party that wins on the basis of a popular promise comes under pressure to deliver from opponents wary lest the party has `cheated' its way into office by making pledges that it cannot or will not redeem.

Where a government attempts to legislate a policy that does not have widespread support, opposition parties have two options. Typically, they try to thwart the government, so as to keep faith with their current supporters and win the adherence of others; hence the pressure on the government not to deliver on Telstra. But oppositions can opt for a quite different line of attack. Before the 1993 election, the Coalition declared that if elected it would introduce a goods-and-services tax. The Labor government, led by Keating, was not particularly popular, and the coalition's promise of a GST—not a particularly popular promise—seemed to offer the government a lifeline. The lifeline, however, was also a rope with which Labor might hang itself. What Keating feared was voters trying to have it both ways—handing the Coalition a majority in the House while relying on Labor and the minor parties to block the GST in the Senate. Keating decided that Labor's chances of winning would be much enhanced if he deprived voters of this choice; he may also have reasoned that if Labor lost the election, its chances of winning next time would be greater if it were trying to defeat a government that had introduced a new tax rather than one that had been prevented from doing so. So he announced that if Labor did lose the election, it would not oppose the GST in the Senate.

The tactic was as audacious as it was unusual. Far from being anticipated as the normal price of defeat—the acknowledgement of a specific mandate—the decision not to oppose a GST, as one biographer put it, `stunned those on both the government and opposition benches.' [41] Keating assumed that voters might be prepared to elect a government whose main policy they did not support, and that they might do this precisely because they did not subscribe to the theory that governments could claim mandates for unpopular policies.

Four years later, under Kim Beazley, Labor repeated Keating's GST pledge. In effect the party was saying that the next election, too, would be different from earlier elections: it would be `a mandate election'. [42] In May 1998, however, during the Budget debate, the party resiled from Keating's position: if the Coalition ran on a GST and won, said Beazley, Labor would oppose the legislation in the Senate. The reason for the turnaround is as significant as the turnaround itself. Labor's research indicated that its initial decision to respect the Coalition's `mandate'—to oppose the tax on the hustings, but to allow it to pass through the parliament, if the Coalition were returned—had left the electorate in considerable doubt about whether Labor's opposition to the GST was genuine. [43]

Responsible parties

Ultimately, the argument about expectations and obligations feeds into a wider set of arguments about responsible party government. In the standard view, parties compete for office on the basis of distinctive sets of policy proposals. Whichever party wins office is expected to implement its policies. Parties that do not win not only respect the right of the government to implement the policies on which (as the saying goes) it was elected, but to make sure that these policies are implemented.

In a critique of the Democrats' attempt to frustrate the government on Telstra one proponent of the responsible party model, Hugh Emy, linked `the growing practice' among parties `of making promises they have no intention of keeping', with the declining esteem in which politicians appear to be held. Unless governments keep to the straight and narrow on `mandate doctrines', he argued, `it is difficult to see what can help to arrest the growth of cynicism'—not just in Australia, but throughout the capitalist world. [44] But why link political cynicism to the breaking of electoral promises only by governments? Might not respect for the system also be undermined by opposition parties promising to do one thing, but doing another? A situation in which every party sticks to its guns might do more to reduce electoral cynicism than a situation in which only the government does.

Emy points to research in Britain, the United States and Australia which suggests that governments `take their manifesto commitments seriously' and infers from this that `mandate' is a term that `describes the way parties behave in countries which contain a responsible party model.' Even if he is right about governments feeling obliged to carry out the policies on which they campaigned, this tells us nothing about what oppositions feel obliged to do. Respect for a government's mandate may be part of the explanation for the way oppositions behave. Support for the substance of what a government promises may be another. But a government's `mandate' may also prevail because the opposition lacks the power to prevent it. What is relevant to any test of mandate theory is not just the proportion of the promises made by governments are kept, but how opposition parties respond to legislation embodying such promises when they are both pledged to oppose them and in a position to block them. The main reason why the British Conservatives were more successful than the Republicans in fulfilling their `mandate' was that in Britain, Thatcher faced quite limited institutional constraints, whereas in the United States, `if another party controlled one or both Houses of Congress, it too claimed and acted upon a separate mandate.' [45]

Leaders of Australia's major political parties are under no illusion that winning office gives a prime minister total control of the parliament. During the 1996 campaign, Howard acknowledged that the Coalition might win government but not win control of the Senate. But, Howard assured voters, he had `no doubt' of his ability `to persuade those who hold the balance of power' that there was `a public benefit in selling one-third of Telstra'. [46] Presumably he said this to protect the credibility of his promise from environmentalists who were doubtful that the sale of Telstra and hence the environmental fund would ever be delivered. Since he realised he would almost certainly not control the Senate and that those likely to control it were promising to prevent the Telstra sale, Howard was gambling that his powers of persuasion would work. What if they did not? Asked why voters should not believe that the Greens and the Australian Democrats would block the sale, the Coalition, as one newspaper put it, had `no answers'. [47]

In the course of the campaign, only one of the serious dailies was prepared to concede the conundrum produced by a bicameral system. In an editorial published on the day of the election, the Age argued that if the Coalition were elected it could `reasonably claim to have its Telstra blueprint endorsed by the electorate.' But if the Democrats were `returned to the Senate', Kernot could `also claim a mandate against the disposal of Telstra.' Even so, the Age warned the Democrats about `overplaying their hand'—a warning which ignored the fact that without Labor, the Greens and at least one Independent, the Democrats would have no hand to `overplay'. [48]


Conjuring a `huge mandate' out of his 1996 victory, Howard sought to privilege seats over votes and the prerogatives of the House of Representatives over the very real powers of a popularly elected Senate. In terms of seats, the result in the House was a landslide, but like most governments before it, the Coalition did not win even half the votes cast. In the Senate, it fared worse in terms of seats and in terms of votes.

Howard, however, was hardly the first person to interpret the outcome of an election for two chambers as if it were an outcome for just one. Editorial writers and others who insisted that `the mandate should be seen as the linchpin of democratic practice', and that parties need to `obey the will of the people', or `respect the democratic process', [49] not only ignored the complex relationship between voters' party choice and their policy positions, but were oblivious to the fact that the election was for two chambers—each of them endowed with more or less equal powers—not one. In principle, voters are just as capable of bestowing their `mandate' on one chamber as on another and equally able to express their `will' vis-a-vis the Senate as they are in respect of the House. Or they may bestow their mandate on neither chamber, leaving it to the parties themselves to `enter into negotiations' [50]—a possibility the survey research did not explore.

Certainly, the Howard government could not claim a mandate for its Telstra legislation, if by mandate we mean clear evidence that it had won government on the strength of Telstra or that a partial privatisation of Telstra, with or without the proposed trade-offs, enjoyed wide electoral support. In terms of a specific mandate, the opposition parties in the Senate were on stronger ground than the government. They had campaigned against the Coalition's proposal and won a greater share of the vote; they could point to surveys that supported their position on the Telstra Bill; and there was evidence that backed their right to block it—evidence that suggested (pace Aitkin) that the electorate did not think that a government, once elected, should just `get on with the job.'

It may be objected that survey evidence ought to be discounted either because many respondents may have given the issues little thought or because small changes in the wording of items can lead to substantial changes in results. [51] That the data are fragile cannot be gainsaid. But the conclusion drawn from this is hardly decisive. First, because no one has produced a set of questions and answers on Telstra that are as defensible as those generated by the polls but better able to shore-up the government's case. Second, because (short of embracing Mulgan's position) one would need to defend some alternative measure of popular opinion—one that no one, in this context, has proposed. [52] Third, and most strikingly, because in the afterglow of the election victory no member of the government directly sought to dismiss the polls the Democrats cited or that Harradine sought. [53]

Acceptance of the data supporting the Senate's right to block the part-sale of Telstra goes some way to settling the debate about whether parties with a relatively small share of the vote `abuse' their power if they join with others to defeat government legislation. [54] It marginalises the question of whether those who voted for the opposition parties, especially for the Democrats, were or were not particularly concerned about the Telstra issue—a matter of keen dispute between the Democrats, on the one side, and the Liberals and political commentators like Malcolm Mackerras, on the other. [55] And it makes nonsense of the Australian's forecast that if the Democrats joined forces with Labor and opposed `the key bills of the Howard Government' they would be treated with `little relevance as far as voters are concerned'. [56] Voters usually find opposition attempts to block unpopular legislation more persuasive than governments appealing to their `mandate'. More abstractly: voters' judgments about the substance of legislation are likely to trump countervailing notions of correct parliamentary form.

In the end, the government carried its legislation by winning over the two independents—each of whom bargained long and hard before deferring (without the slightest sense of irony) to what they were then happy to call the government's `mandate'. [57] What was unusual about all this was not the fact that the government won; governments almost always get their legislation through in one form or another. [58] Nor was the hard bargaining or brinkmanship odd. What was unusual was the size of the stakes and the publicity that the conflict engendered.

Governments have long realised that winning a majority of seats in the House of Representatives is only the beginning of their task when they lack a majority in the Senate. To get their legislation through the Senate, governments have sometimes had to second-guess what opposition parties might do and trim their sails accordingly. They have frequently had to initiate amendments of their own or accept those forced on them by the Opposition. And they have had to horse-trade [59]—though to characterise the minimum winning coalitions that result from such bargaining as a victory for `consensus' over `majoritarian' politics [60] is surely to gild the lily.

Where all else has failed, governments have had to rely on public support. Nothing in its immediate response to Telstra contradicts the view, articulated by the Clerk of the Senate, that the Senate is `unlikely to resist legislation in respect of which a government can truly claim explicit electoral endorsement'. [61] On the contrary, it is at least arguable that the presence of Democrat senators has `saved' both Coalition and Labor governments at various times from a `massive voter backlash'. [62] Ultimately, minorities in the Senate may be more inhibited about flouting public opinion than are governments with large majorities on the floor of the House.


For comments on an earlier version of this paper, I am grateful to Bob Goodin and John Kilcullen and to the participants in an ANU seminar on the paper in the Research School of Social Sciences, especially Geoffrey Brennan and Richard Mulgan.

[1] Australian Financial Review, 5 March 1996.

[2] Electoral Pocket Book, Canberra, Australian Electoral Commission, 1996, p. 73.

[3] Sydney Morning Herald, 21 October 1996.

[4] Australian, 21 March 1996.

[5] Sydney Morning Herald, 22 October 1996; Age (Melbourne), 21 October 1996.

[6] Age (Melbourne), 22 October 1996.

[7] Weekend Australian, 12–13 October 1996, for Newspoll; Sydney Morning Herald, 21 October 1996, for AGB McNair.

[8] E.G. Whitlam, The Whitlam Government 1972–1975, Ringwood, Vic., Viking, 1985, p. 24. The most sustained discussion of Whitlam's ideas on mandates is in C.J. Lloyd and G.S. Reid, Out of the Wilderness: the Return of Labor, North Melbourne, Cassell,. 1974, pp. 197–210; but cf. the uncritical account in Jenny Hocking, Lionel Murphy: a Political Biography, Cambridge, Cambridge University Press, 1997, pp. 147–48. For descriptions of the mandates claimed by other post-war leaders, see Paul Corcoran and S.A. Rowland, `The People on Probation: Rhetorical Images of the Australian Electorate', in G. Crowder et al., eds, Australasian Political Studies 1997, Adelaide, Department of Politics, Flinders University of South Australia, 1997, vol. 1, p. 234ff. In New Zealand, in the 1970s, Labour's Norman Kirk and the National's Sir Robert Muldoon also displayed a Whitlam-like `obsession' with their election policies; see Richard Mulgan, Democracy and Power in New Zealand, 2nd edn, Auckland, Oxford University Press, 1989, pp. 64–65.

[9] J.A. Schumpeter, Capitalism, Socialism and Democracy, London, George Allen & Unwin, 1943, chapter 22; see also, S.I. Benn and R.S. Peters, Social Principles and the Democratic State, London, George Allen & Unwin, 1959, p. 345 and J.P. Plamenatz, Consent, Freedom and Political Obligation, 2nd edn, London, Oxford University Press, 1968 (first published 1938), p. 21. For a critique of Schumpeter's distinction between `men' and `measures', see Jack Lively, Democracy, Oxford, Basil Blackwell, 1975, pp. 39–40; but for the view that this distinction needs to be maintained if we are to understand English voters prior to the emergence of mandates, see C.S. Emden, The People and the Constitution, London, Oxford University Press, 1933, chapter 8.

[10] Don Aitkin, `Foreword', in David Hamer, Can Responsible Government Survive in Australia? Canberra, Centre for Research in Public Sector Management, University of Canberra, 1994, p. vi.

[11] See Rodney Smith, `Parliament', in J. Brett, J. Gillespie and M. Goot, eds, Developments in Australian Politics, South Melbourne, Macmillan, 1994, p. 112, for data on the federal and state parliaments since 1960; and G.S. Reid, `Parliamentary-Executive Relations: the Suppression of Politics', in Henry Mayer, ed., Australian Politics: A Second Reader, Melbourne, F.W. Cheshire, 1969, p. 518, for federal data covering the period 1901–1968.

[12] A. Lijphart, Democracies, New Haven, Yale University Press, 1984, p. 97.

[13] See, H.R. Anderson, `The Constitutional Framework', in S.R. Davis, ed., The Government of the Australian States, London, Longmans, Green, 1961, p. 7; Ken Turner, House of Review? The New South Wales Legislative Council 1934–1968, Sydney, University of Sydney Press, 1969, pp. 43–44, 125; and Murray Goot, `Electoral Systems', in D. Aitkin, ed., Surveys of Australian Political Science, North Sydney, George Allen & Unwin, 1985, pp. 187–88.

[14] See, for example, L.F. Crisp, Australian National Government, 5th edn, Melbourne, Longman, Cheshire, 1983, pp. 149, 349; but compare Geoffrey Sawer, Federation Under Strain: Australia 1971–1975, Carlton, Melbourne University Press, 1977, pp.110–11, who stresses, correctly, the varying rationales underlying the electoral systems for the two chambers.

[15] E.T. Brown, Bread and Power, London, William Hodge, 1940, p. 78.

[16] Australian, 21 March 1996.

[17] David Butler, `Six Notes on Australian Psephology', in C. Bean, et al., eds, The Politics of Retribution: The 1996 Federal Election, St. Leonards, Allen & Unwin, 1997, pp. 237–38.

[18] Richard Mulgan, `The Australian Senate as a “House of Review” ', Australian Journal of Political Science, vol. 31, no. 1, 1996, p. 196; `The concept of mandate in New Zealand Politics', Political Science, vol. 30, 1978, p. 91. See also, A. Schedler, `Introduction: Antipolitics—Closing and Colonizing the Public Sphere', in A. Schedler, ed., The End of Politics?, London, Macmillan, 1997, p.12.

[19] S. Kelley, Interpreting Elections, Princeton, Princeton University Press, 1983, p. 128. See also, for example, G.M. Pomper, Elections in America, New York, Dodd Mead, 1968, p. 247; H.F. Pitkin, The Concept of Representation, Berkeley, University of California Press, 1967, esp. p. 146 and Turner, op. cit., pp. 81–2. It is precisely the possibility that politicians might fail to act as `instructed' that explains the demand, at various times, for citizen's initiative and representatives' recall.

[20] Creighton Burns, Parties and People: a Survey Based on the La Trobe Electorate, Parkville, Vic., Melbourne University Press, 1961, p. 126; D.W. Rawson, Australia Votes: the 1958 General Election, Parkville, Vic., Melbourne University Press, 1961, p. 165.

[21] The Bulletin (Sydney), 25 February 1997, for the Morgan poll; Sunday Age (Melbourne), 9November 1997, for Labor's research. Howard's defence of his decision—that the statement issued under his name during the campaign was actually written by the federal director of the Liberal Party, after consulting with the relevant shadow minister rather than with him—may raise questions about what should count as a promise, but it hardly deflects from the point at issue here. For Howard, see Australian, 12 February 1997, and Age, (Melbourne), 13 February 1997.

[22] Anthony Fusaro, `The Australian Senate as a house of review: another look', Australian Journal of Politics and History, vol. XII, 1966, reprinted in C.A. Hughes, ed., Readings in Australian Government, St. Lucia, Qld, University of Queensland Press, 1968, p. 128; C.A. Hughes, `The Machinery of Government', in H. Mayer and H. Nelson, eds, Australian Politics: a Fifth Reader, Melbourne, Longman Cheshire, 1980, p. 203.

[23] Morgan Poll Finding No.2991, The Bulletin (Sydney), 10 June 1997.

[24] Clive Bean, David Gow and Ian McAllister `Australian Election Study, 1998: User's Guide to the Machine-Readable Data File', SSDA Study No.1001, 1999, p. 21.

[25] Hiroya Sugita links the 1983 and 1996 results to special efforts by the Democrats' leaders to get electors to hedge, but in the absence of more information about the Democrats' strategies at these and during the intervening elections, one has to suspend judgment; `Conflicting mandates: the Australian Democrats and the Howard government', Policy, Organisation and Society, no. 13, 1997. For data on the Democrats' performance in the House and Senate, 1977–1996, see, Clive Bean, `Australian Democrats: Electoral Performance and Voting Support', in John Warhurst, ed., Keeping the Bastards Honest: the Australian Democrats' First Twenty Years, St. Leonards, NSW, Allen & Unwin, 1997, p. 70; Bean, however, does not seek to explain the differences between the Democrats' House and Senate votes.

[26] S. Bowler and D. Denemark, `Split ticket voting in Australia: dealignment and inconsistent votes reconsidered', Australian Journal of Political Science, vol. 28, 1993, p. 25.

[27] Age (Melbourne), 26 February 1996.

[28] Sydney Morning Herald and Age (Melbourne), 5 February, 27 February and 4 June 1996 for AGB McNair; Australian, 14 February 1996 for Newspoll. For a wider discussion of how the Coalition negotiated the shoals of public opinion on Telstra, see Murray Goot `Public opinion, privatisation and the electoral politics of Telstra', Australian Journal of Politics and History, vol. 45, no. 2, 1999, pp. 214–38.

[29] Morgan Poll Finding No. 3089, The Bulletin (Sydney), 2 June 1998; and AC Nielsen, `Election `98 Issues Report', 30 September 1998.

[30] Morgan Gallup Poll Finding No.2858, The Bulletin (Sydney), 20 February 1996.

[31] Morgan Poll Finding No.2912, The Bulletin (Sdyney), 18 June 1996.

[32] Morgan Poll Finding No.2942, The Bulletin (Sydney), 5 November 1996.

[33] Cited in T.W. Beed, `Opinion polling and the elections', in H.R. Penniman, ed., Australia at the Polls: the National Elections of 1975, Washington, D.C., American Enterprise Institute for Public Policy Research, 1977, p. 244.

[34] Sydney Morning Herald, 23 August 1993, for AGB McNair Anderson; Australian, 22 September 1993, for Newspoll; and Morgan Gallup Poll Finding No. 1072, The Bulletin (Sydney), 31 May 1983.

[35] Age (Melbourne), 7 September 1993 for Saulwick; Sydney Morning Herald, 5 July 1995 for AGB McNair.

[36] CPD (Commonwealth Parliamentary Debates), 2 November 1992, p. 2549.

[37] Australian, 4 December 1996.

[38] CPD, 15 September 1987, p. 2. See, John Warhurst `The ALP Campaign', in Ian McAllister and John Warhurst, eds, Australia Votes: The 1987 Federal Election, Melbourne, Longman Cheshire, 1988, pp. 37–38, 53, for the double dissolution campaign. Although Ewart Smith is correct in observing that during `the election campaign hardly a word was said about the ID card', Howard's argument did not rest on that; The Australia Card: the Story of its Defeat, South Melbourne, Sun Books, 1989, p. 95.

[39] Australian, 15 April 1996, for the argument about the pre-conditions for a mandate, advanced by Donald Horne; Australian, 23 April 1996, for the counter-argument by Mulgan.

[40] Mulgan, Democracy and Power, op. cit., p. 64.

[41] Michael Gordon, Paul Keating, new edn, St. Lucia, Qld, University of Queensland Press, 1993, p. 193.

[42] Australian Financial Review, 27 August 1997.

[43] Sunday Telegraph (Sydney), 17 May 1998.

[44] Hugh Emy, `The mandate and responsible government', Australian Journal of Political Science, vol. 32, 1997, p. 73.

[45] T.J. Royed, `Testing the mandate model in Britain and the United States: evidence from the Reagan and Thatcher eras', British Journal of Political Science, vol. 26, 1996, pp. 45–80; Emy, op. cit., p. 69.

[46] Daily Telegraph (Sydney), 3 February 1996.

[47] Australian, 28 February 1996.

[48] Age (Melbourne), 2 March 1996; emphasis in the original.

[49] Age (Melbourne), 11 November 1997; Sunday Telegraph (Sydney), 17 May 1998; Age, 16 May 1998.

[50] Campbell Sharman, `The Senate and good government', Papers on Parliament, no. 33, May 1999, p.154.

[51] See, most recently, D. Walton, Appeal to Popular Opinion, University Park, Pennsylvania State University Press, 1999, chapter 1.

[52] But cf. Julianne Schultz, Reviving the Fourth Estate, Cambridge, Cambridge University Press, 1998, p. 152 for the views of Australian journalists, and Susan Herbst, Reading Public Opinion, Chicago, University of Chicago Press, 1998, for the views of American staffers, political activists and journalists.

[53] Compare J.R. Nethercote's assertion that `it is unlikely that (polls) could be invoked authoritatively to settle a constitutional or even a political point'; `Mandate: Australia's current debate in context', Research Paper, Department of the Parliamentary Library, Canberra, no. 19, 1988–89, p. 32.

[54] This view was espoused by, among others, the former Liberal minister and founder of the Democrats, Don Chipp, after the election; Sydney Morning Herald, 25 March 1996. Compare the view, expressed by only 39 per cent of respondents in a 1993 post-election study, that it `is better when the Federal Government has a majority in both the House of Representatives and the Senate'; Roger Jones, et al., `Australian Election Study, 1993: User's Guide for the Machine Readable Data File', Canberra, Social Science Data Archives, Australian National University, 1993, p. 7.

[55] See Australian, 27 March 1996, 20 May 1996, 16 December 1996; Sydney Morning Herald, 28 May 1996.

[56] Australian, 6 May 1996.

[57] Australian, 13 May 1996, for Harradine; Sydney Morning Herald, 21 August 1996, for Colston. Apart from extracting what was reported by two different papers to be $53 million or $250 million for his constituents in Queensland, Colston was said to have persuaded the Prime Minister to approve a substantial pay rise for a member of his staff; Age (Melbourne), 13 February 1997, Sydney Morning Herald, 13 February 1997, Age (Melbourne), 15 February 1997. Harradine was credited, variously, with winning $58 million, $100 million, $150–$170 million and $180 million in cash and jobs for Tasmania, the Sunday Examiner in Launceston declaring: `Harradine joins ranks of greats'; Sydney Morning Herald, 1 February 1997 and Age (Melbourne), 15 and 19 February 1997.

[58] In 1976, the Fraser government (which enjoyed a majority in the Senate) secured 94 per cent of its program, while in 1995 the Keating government (which lacked a Senate majority) got 96 per cent of its program through the parliament; C. Elliott, `Less than optimal outcomes. Fraser and Keating without the numbers', Legislative Studies, vol. 11, no. 2, 1997, pp. 61–82. See also the discussion in Gwynneth Singleton, `Independents in a multi-party system: the experience of the Australian Senate', Papers on Parliament, no. 28, November 1996, p. 161 and the critical evaluation of `obstructionism' in Liz Young, `Minor parties and the legislative process in the Australian Senate: a study of the 1993 Budget', Australian Journal of Political Science, vol. 34, no. 1, 1999, pp. 7–27.

[59] Harry Evans, `Government and Parliament', in G. Singleton, ed., The Second Keating Government: Australian Commonwealth Administration 19931996, Canberra, Centre for Research in Public Sector Management, University of Canberra and the Institute of Public Administration Australia, 1997, pp. 32–33. The best discussion of the Democrats' performance is in Hiroya Sugita, `Parliamentary Performance in the Senate', in J. Warhurst, op. cit., chapter 9.

[60] Campbell Sharman, `Parliamentary federations and limited government: constitutional design and redesign in Australia and Canada', Journal of Theoretical Politics, vol. 2, no. 2, 1990, p. 225. As Geoffrey Brennan points out, it would require the active involvement of the opposition party to turn the system into a quasi-consensual one; `The “unrepresentative swill” “feel their oats”', Policy, vol. 14, no. 4, 1998–99, p. 9.

[61] Harry Evans, ed., Odgers' Australian Senate Practice, 7th edn, Canberra, Australian Government Publishing Service, 1995, p. 13; see also R.B. Scotton and C.R. McDonald, The Making of Medibank, Kensington, NSW, School of Health Services Management, University of New South Wales, 1993, p.75.

[62] Sugita, `Parliamentary Performance', op. cit., pp. 174–75.