Australian Society of Quantitative Political Science conference, Melbourne 2015 Do economic cleavages still matter in Australian politics? S HAUN R ATCLIFF Monash University
Economic cleavages: still salient or a relic of the industrial revolution? Traditional view of politics as a function of deciding questions of how resources were distributed and societies were ordered. Common in classical Greece and Rome (Aristotle 1944; Plutarch 1970: Ch 3; Appian 1996), modern political science (Lasswell 1950; Downs 1957; Lipset and Rokkan 1967; Hibbs 1977), and in Australia (Bryce 1921: 172 ‐ 173; Alford 1963: 11; Aitkin 1977: 9 ‐ 11). ‘Death of class’ theories (Clark et al. 1993; Pakulski and Waters 1996; Inglehart 1997). General account being economic cleavages were part of a specific historical process and politically generated. Declined in advanced (or post) industrial society, with diminishing economic/class identities (working class), ideologies (socialism) and organisations (unions), replaced with cultural, self ‐ expression, social, and consumption and lifestyle driven status. Accepted by some scholars examining Australian politics (Alford 1963: 178; Kemp 1978: 64 ‐ 68; Jones and McAllister 1989; Goot 1994; Charnock 1997; Bean 1999; Weakliem and Western 1999; Jansen et al. 2013). Ratcliff | Do economic cleavages still matter in Australian politics?
Reasons to be skeptical of declining economic cleavages However, these claims of a decline in the importance of economic cleavages rest on questionable assumptions. They largely use a relatively untheoretical view of economic class that includes missing some of the most important aspects of class, including the importance of power, social organisation and the creation and control of capital, and the conflicts that decide these matters (Goot 1994: 5). They are largely based on studies of voting by occupation, and particularly the Alford Index. They also largely ignore ongoing concerns of redistribution and inequality (Goldin and Katz 2001; Atkinson and Leigh 2013; Piketty 2014), and social mobility (Clark 2014). Ratcliff | Do economic cleavages still matter in Australian politics?
Another way to measure economic cleavages This paper examines patterns in major ‐ party voting by occupation, sector of employment and income, from 1979 to 2013. More detailed geographic patterns in the five elections between 2001 and 2013 are also studied. Politicised economic cleavages = political differentiation emerging from distinct life ‐ chances from different opportunities to access the means of production and capital (Marx and Engels 1848), trade and consumption (Weber 1978), with ownership of capital and possession of valued skills influencing the chances they have of gaining access to valued outcomes (see also Giddens 1973). It assumed in this study that parties operate as interest aggregators: favouring policies supported by or beneficial to specific groups in society, which in return provide their party with a disproportionate share of the group’s vote. Ratcliff | Do economic cleavages still matter in Australian politics?
This is tested in two ways: First, trends in voting by economic backgrounds are estimated. This is done using multilevel models that include linear slopes for time, which are allowed to vary by intercepts for occupation ‐ sector. Time is also allowed to interact with a linear predictor for income, which is also allowed to vary by occupation ‐ sector. Multilevel Regression and Post ‐ stratification (MRP) is used to examine geographic patterns in the economic cleavages at the five federal elections between 2001 and 2013. Ratcliff | Do economic cleavages still matter in Australian politics?
Why do we care? The key contribution of this paper is the examination of both temporal trends and political geography in the partisan preferences of voters with different occupations, sectors of employment and income levels. Helps us better understand voter behaviour in context of import aspects of their socioeconomic status. Through this we can try and understand voter revealed preferences. As Andrew Leigh (2005: 266) suggested, if voters with different interests systematically support different political parties and they are quasi ‐ rational actors, then their partisan choice may indicate different impacts of party policies when they are in government. If we take this behaviour by voters to represent their revealed preferences, the findings outlined in this paper suggests Australia’s major parties continue to act as interest aggregators (Weeden and Kurzban 2014) Ratcliff | Do economic cleavages still matter in Australian politics?
The ongoing importance of economic cleavages in the Australian electorate A multilevel logistic regression is fit to a pooled data file of the nine national election surveys conducted in Australia from 1979 to 2013 that include the variables in which we are interested ( � =12,920). Intercepts are allowed to vary by 16 occupation ‐ sector categories (5 occupation x 3 sector + not active in the paid labour force). Varying slopes by household income (measured as the percentile rank of a respondent’s household income then standardised) and time (mean centred and measured in decades), which also interact. Ratcliff | Do economic cleavages still matter in Australian politics?
This model was fit using the glm() function in the open ‐ source statistical software package R (R Core Team 2013), and can be written as: Pr ( y � � 1� � logit �� �� � � � ��� ������∗���� � ������ � � ���� � � ������∗���� � � ������ � � ��� ���� � � ��� ���.������ � � ���� ������∗���.������ � � ������ � � ���� ����∗���.������ � � ���� � � ���� where � is the probability of respondent � providing the Coalition rather than the Labor Party with their vote in each survey, conditional on their household income, and occupation and sector of employment � , the effects of which are allowed to change over time. Controls for age, birthplace and education, as well as the year of the survey, were then included to ensure patterns observed were not the result of these exogenous and potentially confounding demographic factors. Ratcliff | Do economic cleavages still matter in Australian politics?
Figure 1. Trends in vote by occupation, sector of employment and income, 1979 ‐ 2013. Each plot shows trends in Coalition voting for voters in each occupation ‐ sector group. Dark curve those voters with household incomes at the 95 th percentile. Grey curve those at the 50 th percentile. Light curve those at the 5 th percentile. Ratcliff | Do economic cleavages still matter in Australian politics?
Measuring the geographic patterns in economic cleavages To understand how economic cleavages work in this context, the relationship between individual ‐ level socioeconomic predictors and vote choice is then examined in more detail over the five elections from 2001 ‐ 13 using MRP to model a pooled dataset ( � =8,919). This model can be written with the notation: ��� � � ���� ���� � � ���� Pr( y � � 1� � logit �� �� � � � ���� ��� � � ���� ���.������ � � ���� ������ � � ���� �������� � Also includes linear variables of income, age, education and binary for birthplace; 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 � � ������� . Results are then poststratified using 2011 census data. Ratcliff | Do economic cleavages still matter in Australian politics?
Figure 2. Coalition voting by occupation, sector of employment and income, 2001 ‐ 13. Ratcliff | Do economic cleavages still matter in Australian politics?
Figure 3. Major ‐ party vote share in south ‐ east Australia by division, household income percentile and select occupations . Ratcliff | Do economic cleavages still matter in Australian politics?
Figure 4. Major ‐ party vote share in Melbourne electoral divisions. Low income public sector manual workers High income self ‐ employed managers Ratcliff | Do economic cleavages still matter in Australian politics?
Figure 5. Gap in Coalition support between high income self ‐ employed and private sector managers (capital) and low income manual employees (labour) in each electorate, by Coalition vote share and median division income. Ratcliff | Do economic cleavages still matter in Australian politics?
Figure 6. Comparison of estimated and observed Coalition vote, by division. Ratcliff | Do economic cleavages still matter in Australian politics?
Recommend
More recommend
