by Colin A. Hughes
Parliamentary elections are a prominent feature of Australian politics. They are frequent, close, decisive, and extensively reported by the mass media. Federal elections for the lower house, the House of Representatives, must be held every three years; in practice they are held on average at approximately two and a half year intervals. Senate elections are held usually, but not invariably, on the same day. State and local government elections are kept separate from federal elections. For most of the 20th century there were only two serious competitors in parliamentary elections. Each could count on at least 40% of the vote, and usually polled 45% or better. Thus most outcomes were in doubt until the polls had closed, and campaigns were important because they might influence outcomes. Other candidates were unable to win in single-member electoral districts which ensured that the outcome of the House of Representatives election immediately settled which of the two rivals would form the national government. Even when the winner had a tiny majority, strict party discipline in the Parliament allowed it to govern effectively.
Referendums, in contrast, have been marginal. Their scope is narrow: the federal referendums have been confined to relatively obscure constitutional questions, the state ones to a limited range of topics. Their initiation has been restricted to the government of the day passing legislation. Recent proposals to allow other possibilities, for example to empower state legislatures or electors' petitions to initiate federal constitutional referendums, have not progressed. The 1998 Constitutional Convention was the first occasion since federation in 1901 when electors chose delegates to prepare constitutional amendments, and then only half the members of the Convention. Their proposals still had to go through the usual cabinet and Parliament channels before the referendums could be held.