How many parties are enough? The changing structure of Australian political attitudes

This paper examines whether Australia's majoritarian electoral system remains fit for purpose, given declining major-party support. It tests the “fragmentation hypothesis” - that new political cleavages have emerged which existing parties cannot resolve - against the alternative that voters have simply dealigned in protest. Using k-means clustering on Australian Election Study data from 1998 to 2025, the paper inductively maps political and social attitude groups at each federal election. It finds no evidence that the electorate has become more fragmented. Instead, substantial realignment has occurred: the Liberal-voting professional-class base has declined, and the conservative right has split into two issue-based groups, one clustering on social issues and one on economic issues. Without electoral system change or internal party reforms, the major party duopoly will continue struggling to breach these cleavages.