Australian Political Studies Association annual conference, Canberra 2015 Rural conservative, inner ‐ city elites, and suburban aspirationals: Geographic variation in income voting in Australia S HAUN R ATCLIFF Monash University AND S HAUN W ILSON Macquarie University
The demise of economic cleavages? Historically, political scientists Harold Lasswell (1950), Anthony Downs (1957: 100) and Douglas Hibbs (1977) asserted modern politics is largely concerned with economic issues, economic class and income redistribution. However, in recent years, scholars in Australia (Alford 1963: 178; Kemp 1978: 64 ‐ 68; Jones and McAllister 1989; Goot 1994; Weakliem and Western 1999) and elsewhere have claimed socio ‐ economic class has lost its salience in modern party politics. These claims go beyond the academy and have also been made by journalists (Pearson 2012; van Onselen 2012), political staffers (Robb 1997: 36 ‐ 40) and politicians themselves (Kelly 2001; Uren 2013). Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
Some problems with these claims. Observational studies continue to find evidence of “income voting”: higher income voters and asset ‐ owners consistently favour the Coalition (Leigh 2005; McAllister 2011: 163 ‐ 167) Given clear evidence of declining class voting, any persistence of income on voting behaviour raises important puzzles for political scientists and sociologists, and calls for more sophisticated ways of identifying how broadly ‐ understood “economic interests” continue to shape politics (see Weakliem 2013). Trying to understand economic cleavages through occupational class alone is not enough. Why does income matter? Measures resources; different life chances and outcomes; a different relationship with the state (net beneficiary of spending or payer of taxes). Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
Mechanisms involved in income and class voting . We are interested in the first mechanism of class voting: income voting that produces support for “economic liberalism” or socio ‐ economic interests that in turn shapes votes for centre ‐ left or centre ‐ right parties (Coalition and Labor in Australia’s case). Figure reproduced by van der Waal (2007: 408). Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
Income versus class voting. As has been found in the United States, class and income voting trends patterns are diverging in important ways. Gelman (2009) and Brady et al (2009) have shown traditional manual workers moving to the Republicans while low income voters remain loyally Democratic. We find a similar divergence here. Figure 1. Coalition two ‐ party vote by occupation, sector and household income, 2001 ‐ 13. Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
Income voting patterns over time. Figure 2. Coefficients and standard errors for income in logistic regressions estimating Pr ( � � � 1� � logit �� �� � �� fit to ANPAS and AES data, 1967 ‐ 2013. The positive coefficients indicate voters with higher household incomes have consistently been more likely to provide the Coalition with their first preference vote. This pattern becomes even stronger once age, birthplace, education and gender are controlled for, as in the second plot. Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
Geographic variations in these patterns. These likely understate the impact of household income on voting behaviour in Australia, due to regional voting patterns (ie, conservative low ‐ income rural voters and progressive high income urban voters). A Multilevel Regression and Poststratification (MRP) is fit to 2001 ‐ 13 AES data to estimate the relationship between income and voting; taking into account electorate ‐ level differences. ��� � � ���� ������,������ � � ���� ���� � � ���� ������ � � ���� ��� � � ���� Pr ( y � � 1� � logit �� �� � � � ���� �������� � Also includes linear variables of income age, education; intercepts for state � � ����� , and division ‐ level predictors for the average two ‐ party vote � � ���.���� , median age � � ������.��� and ������� . income � � ������.������ , proportion of the population born overseas � � �������� and density � � Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
Poststratification. Results from model are poststratified using 2011 census data, with a cell provided for each demographic category in the model. Division Income Age Education Sex Birthplace N 1 Adelaide Lowest quintile Age 18‐29 Some school Female Born in Australia 218 2 Aston Lowest quintile Age 18‐29 Some school Female Born in Australia 161 3 Ballarat Lowest quintile Age 18‐29 Some school Female Born in Australia 657 4 Banks Lowest quintile Age 18‐29 Some school Female Born in Australia 140 5 Barker Lowest quintile Age 18‐29 Some school Female Born in Australia 715 6 Barton Lowest quintile Age 18‐29 Some school Female Born in Australia 124 7 Bass Lowest quintile Age 18‐29 Some school Female Born in Australia 741 8 Batman Lowest quintile Age 18‐29 Some school Female Born in Australia 165 9 Bendigo Lowest quintile Age 18‐29 Some school Female Born in Australia 298 10 Bennelong Lowest quintile Age 18‐29 Some school Female Born in Australia 92 .... .... .... .... .... .... .... .... 48000 Wright Highest quintile Age 65 plus University Male Born in Australia 58 Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
Figure 3. Coalition two ‐ party vote by income in each of Australia's 150 electoral divisions. Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
Figure 4. A closer look at 13 divisions. Each curve is Coalition support as income increases, estimated from the MRP model described above. Open circles represent the voting patterns observed in the raw AES survey data. The area of each circle is proportional to the number of respondents it represents in the survey data. Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
Figure 5. Geographic patterns in voting by age and income. Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
A closer look: Figure 6. Two ‐ party voting in Sydney, low ‐ income voters aged 30 ‐ 44. Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
Figure 7. Two ‐ party voting in Sydney, high ‐ income voters aged 30 ‐ 44. Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
Why this matters. Income is an important and broader way to understand individuals’ relationship to the state, and their access to resources and opportunities. Individuals from higher income (or SES) households have: Greater access to healthcare (Schoen et al. 2000). Better dental health (Sanders and Spencer 2004). Live longer (Clarke and Leigh 2011). Better educational outcomes (Williams et al. 1991; Ainley et al. 1995). Pay more tax, and are less likely to receive direct cash payments from government (Australian Bureau of Statistics 2013: Government Benefits, Taxes and Household Income, 2011 ‐ 2012 ‐ Detailed tables). Those from high ‐ income households will have different views of government and different policy and political perspectives than those from low ‐ income households. Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
Why this matters: the possible impact on party politics. Figure 8. Income voting is higher in marginal electorates. Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
Findings and an interesting puzzle. Income voting is stable over time and, once electoral geography is accounted for, the relationship strengthens. This suggests that socio ‐ economic interests continue to shape vote choice and in turn party positions. These results suggest caution in claiming that class influences on voting are disappearing or weakening, if by that it is suggested that socio ‐ economic differences no longer shape vote choice. Why is economic voting stronger in Australia’s marginal seats? There are cultural/economic differences between divisions, and particular social geographies—different average incomes, ages, education: poor rural voters more likely to support Coalition, better ‐ off inner city voters Labor (and Greens). Safe electorates build up local partisan control that (sometimes) weakens income effects. Marginal electorates have more frequent changes in party control. Parties target marginal seats – with Labor targeting low ‐ to ‐ middle income voters, and the Coalition middle ‐ to ‐ high income voters (or political geography influences targeting). Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
Thank you. Ratcliff and Wilson | Rural conservatives, inner ‐ city elites, and suburban aspirationals
Recommend
More recommend
