On one level the 2010 Australian election was a remarkable one. But on another level it was an election in which very little changed: the patterns of 2007 persisted with a regional overlay. In 2007 class alignments shifted as Labor won back the support of many working-class voters who had supported the Coalition in 2004 (as they had done in 1996). For all the media’s preoccupation with One Nation in 1998 and suburban redneckery in 2001 these two elections saw Labor do OK among its traditional support base. In 2010 the swing against (or towards) Labor was regional shifting patterns of class allegiance were irrelevant Tony Abbott may have almost won but his coalition of support was different from that of Howard in 2004.
How do I know this? I did a logistic regression analysis of the data in the 2004, 2007 and 2010 Australian Election Studies. The variables I sought to explain were a House of Representatives vote either for Labor or for either of Labor and the Greens. My explanatory variables were gender, trade union membership, manual working-class status (trade, labourer or process worker), self-employment, employment in a family business or a farm, professional-managerial employment. This was a multivariate analysis I am not just considering how trade unionists or manual workers voted in the aggregate but the extent to which personal attributes such as union membership or manual occupation impact on voting choices after all other variables are taken into account. Trade union members are disportionately government employees so some of the observed higher levels of Labor support among trade unionist is actually a reflection of government employment. The technique I use enables these factors to be separated. The patterns of support are similar for the Labor vote and the combined Labor+Greens vote.
Encouraged left vote:
Trade union membership, government employment
Discouraged left vote:
Self-employment, professional-managerial employment, family-farm employment
In 2004 the Liberals moved out from their heartland being a manual worker had no impact on electoral choice. Labor was pushed back to its diminished heartland of unionists and government employees.
Encouraged left vote: union membership, government employee, manual occupation
Discouraged left vote: self-employment
Labor regained its traditional base but Kevin07 struck a chord with many affluent voters as well. Noteworthy that family-farm employment ceases to be significant, was this a WorkChoices effect?
Encouraged left vote: government employment, trade union membership, manual occupation (a stronger effect than in 2007)
Discouraged left vote: male gender, self-employment, family-farm employment
With WorkChoices off the agenda small business employees (disportionately non-manual also) return to their traditional allegiance. However Labor held its gains among manual workers (Victoria outweighed Qld?). There was a male vote for Abbott. The Coalition did not regain its professional-managerial base lost in 2007. A new regional pattern was overlaid on top of the class realignment of 2007. It explains why Labor held Robertson & Dobbell.