Following the past two federal elections, the government and its supporters have engaged in a debate with opposition parties over the nature of the government's mandate. Assertion of a mandate suggests a certain view about the role of elections and parliament. John Howard, in his 1996 victory speech, claimed 'an emphatic mandate' on the basis of his large majority in the House. In 1998, he claimed a similar mandate for tax reform despite losing seats at that election. Conventional mandate theory would support Mr Howard's right to govern and implement his previously stated policies but the limits to this authority are far from clear. To expect legislation foreshadowed in the election campaign to be passed without significant amendment by the Senate is to champion responsible government over federalism and limited government, something of a change for the Liberal party since Novemember, 1975, when Mr Howard was beginning his parliamentary career.

Similarly, after both elections, the Australian Democrats questioned the strength of the government's claims (the ALP opposition dealt themselves out of such negotiations by remaining implacably opposed to the government's policies on both occassions). In 1996, then leader Cheryl Kernot claimed the government had received only a 'mandate to form government', asserting her own parties' mandate to 'check on [the] new government.' (It would be interesting to hear from Ms Kernot on this subject now that she is one of the major party 'bastards' that the Democrats claim to be keeping honest.) Kernot's intervention represented an attempt to reduce the concept of a mandate to sheer numbers, with the government's claim of a moral right to have the sale of Telstra, for example, met with equal and opposite claims by opposition parties that their rejection of that policy during the election campaign gives them the right to reject the bill in the Senate. In 1998, new leader Meg Lees claimed the Democrat's mandate to be more specific because of their policy to accept a GST with food exempted. If, however, these purported mandates simply reflect numbers in the parliament, the concept is useless.

The idea of a mandate is based on the 'delegate' theory of representation - that members of parliament are compelled to respect the wishes of their constituents. The opposite to the delegate view is the 'trustee' view of representation, where members vote on issues based on their best judgement of ever-changing circumstances. Mandate theory is important in Britain, since a government there is expected to seek a new mandate if it wishes to radically depart from its manifesto, the mandate is thus one of the few limits on government power in the British system. There are, however, institutional restraints on the power of an executive government in Australia, questioning the value of mandate theory here. Despite the tendency of our political parties to periodically self-destruct, the Australian electorate has shown a remarkable tolerance towards (or resignation to) the principle of an election being a contest between the records and policies of the major parties. In both of the elections under consideration here, the Coalition released detailed policies shortly before polling day. While this is more desirable than having new policies outlined shortly after the election, which is more common, it is difficult to judge how much notice voters take of policies released during a campaign.

Parties do not publish masses of detail in their policy statements. Voters would not be expected to read and understand it even if they did. In assessing claims for the government's mandate to sell a third of Telstra, for instance, the fact that voters rated six other issues as more important than privatisation in their decision on how to vote need to be taken into account. A vote is a single answer to many, many questions. Given a limited choice, a vote cast represents the selection of a preferred - not necessarily liked - candidate. The result of the 1996 election was widely seen as a delayed reaction to thirteen years of ALP rule. The opposition successfully tapped into the electorate's tiredness of social and economic change - hardly grounds for a mandate for sweeping change. If the only way an electorate can pass judgement on a government's performance is to vote them out at the next election, what does this say of the mandate of the incoming government? The idea of a mandate suggests that a truly democratic system must incorporate some kind of link between the government and citizens outside elections and campaigns. The theory of responsible government is an attempt to formalise (if not codify) that tenuous contract. The obligation of the victorious party to respect the rights and views of its constituents is said to be reflected in the chain of responsibility from the executive to the parliament and to the people. However, with the executive in modern parliamentary systems in control of the lower house, rather than vice-versa (by virtue of party discipline), this link in the chain of responsibility is broken. The two-party system has ensured that this responsibility is now entirely limited to the bond between the executive and the electorate. The executive may not control the upper house but the prime minister is not responsible to the upper house.

Australian elections deliver decisive results as to which party sits on the treasury benches but executive power is divided by federalism and is limited by bi-cameralism, entrenched constitutions and judicial review. In this system, governments can choose to seek consensus for their policies by consulting interest groups and negotiating agreements with the Senate and state governments. Alternatively, governments can point to their mandate, threaten the Senate with a double dissolution election and joint sitting, over-rule the High Court on common law decisions, by-pass state governments through tied grants and put referenda to the people to change the constitution if necessary, to put their policies into practice. Claims of a mandate tend to be aasociated with the latter style of government.

In Australia, the idea of a mandate has been an important weapon in the government's dealings with the Senate, whose electoral system mitigates against governments controlling it. If the electorate really wished the Senate to act as a vigorous check on the government of the day, voters would prefer minor parties in the Senate in greater numbers than they do. The fact that in 1996 more than three-quarters of voters gave one of the two major parties - the two candidates for government - their primary vote in the upper house would suggest that, at the ballot box, the majority of electors wanted one of the two major parties to win a majority in both the House and the Senate as long as it was their preferred party that won government. After the election, those that voted for the major party which lost expect their representatives in the Senate to act as a check on the government. At the ballot box, most voters wanted their preferred party to control both houses of parliament; they voted for responsible government. With their preferred party having lost the election, many then become advocates of the Senate's role in limited government!