Political Parties and Electoral Behavior Simon Jackman Department of Political Science Stanford University
[email protected] http://jackman.stanford.edu to appear in The Cambridge Handbook of the Social Sciences in Australia Draft: September 2002
Introduction: The Australian Party System Australian electoral politics is dominated by the two largest parties: the Australian Labor Party (the ALP) and the Liberal Party, the latter aided by a coalition with the National Party (formerly the Country Party). Since 1910, when anti-Labor forces united to form the first Liberal Party, Australian electoral politics is centered on the contest between the ALP and a series of anti-Labor parties (the first Liberal Party, the Nationalist Party, the United Australia Party, and from 1944, the Liberal Party). The durability of the Australian party system can be traced to institutional features of Australian politics. Australian national governments are formed in the House of Representatives: the use of single-member districts for elections to the House - decided by absolute majorities, before or after preferences - makes it extremely difficult for minor parties to gain political traction. On the other hand, the electoral system used for Senate elections (the states serving as multi-member districts with seats filled by a Hare-Clark or “quota preferential” method) supplies minor parties and independents with more fruitful avenues to representation (e.g., Papadakis and Bean 1995; see also “Electoral Systems” in this volume). Thus while never winning House of Representatives seats, minor parties such as the Democratic Labor Party (DLP), the Australian Democrats and the Australian Greens (as well as independents) have won sufficient Senate seats to usually hold the balance of power. Second, preference flows in House of Representatives elections provide another possible avenue of leverage for minor parties, say, by demanding policy concessions in return for their second preferences. Moreover, Australia’s system of compulsory and preferential voting means that at each election, every citizen is legally obliged to record a preference for the ALP over the Liberal-Nation coalition, or viceversa (Rydon 1968). Some scholars have speculated that this combination of compulsory and preferential voting bolsters party loyalty among the mass public, further enhancing the stability of the party system (e.g., Aitkin 1982). However, it could also be the case that the legal requirement to vote ensures that citizens who are dissatisfied with the major
parties must nonetheless turn out, ensuring a steady supply of voters for parties and candidates presenting themselves as alternatives to the major parties. Despite these avenues for minor party representation and influence, major party dominance remains the most compelling feature of Australian electoral politics. The only non-major party to regularly win House of Representatives seats is the National Party, the Liberal Party’s rural-based coalition partner. None of the other minor parties gracing the Australian political stage since 1949 have won House of Representatives seats (an enumeration of Australia’s minor parties appears in Jaensch and Mathieson 1998). Between 1941 and 1998, just seven of 703 members elected to the House of Representatives have not been members of a major party or the National Party (Johns 2000, 403; DPL 1999). Support for minor parties is almost always too geographically diffuse for them to win House seats, their electoral successes confined to the Senate. Consequently, the ALP and the Liberal Party have supplied all of Australia’s prime ministers since World War Two (with the exception of the Country Party’s John McEwen’s three week caretaker tenure after Harold Holt’s disappearance in December 1967). With the exception of Country and National Party governments in Queensland, the ALP and the Liberal Party have also dominated politics in the Australian states, supplying nearly all post-war state premiers. Further evidence of the primacy of the major parties is in Figure 1, showing the percentages of first preference votes in House of Representatives elections since 1910. Despite a slight decline in the vote shares received by the two major parties (from better than 96% in 1949 to around 80% in 1998 and 2001), this evidence confirms that Australian electoral politics remains a contest as to which of the two major party groupings – Labor or the Liberal/National coalition – forms the government of the day. While the Australian Democrats have sometimes outpolled the National Party in recent elections, they have never won a House of Representatives seat, since Australian Democrat support is geographically diffuse relative to the concentration of National Party votes in rural seats. Even as National Party support has fallen from about 12% of national vote share (as recently as 1987) to less than 6% in 1998 and 2001, the concentration of these votes in rural seats sees the National Party continuing to win House seats (14 of 150 seats in 2001, or just over 9%; and 16 of 148 in 1998, or just under 11%).
2
96 92 88 84 80 54 56
20 01
19 96 19 98
19 93
19 87 19 90
19 80 19 8 19 3 84
19 7 19 2 7 19 4 75 19 77
19 69
19 66
19 61 19 63
19 58
19 5 19 4 55
19 19 49 51
19 46
19 43
19 40
19 37
19 19 28 29 19 31 19 34
19 25
19 22
19 19 13 14 19 17 19 19
54 56
19 10
80
84
88
92
96
Labor and Anti-Labor Combined
46 48 50 52 36 38 40 42 44 32 34 10
12
16
20 01
19 96 19 98
19 93
19 87 19 90
19 80 19 8 19 3 84
19 7 19 2 7 19 4 75 19 77
19 69
19 66
19 61 19 63
19 58
19 5 19 4 55
19 19 49 51
19 46
19 43
19 40
19 37
19 19 28 29 19 31 19 34
19 25
19 22
19 19 13 14 19 17 19 19
10
12
16
19 10
32 34
36 38 40 42 44
46 48 50 52
ALP Liberal
8
8
DEM
PHON
6
6
NP
NP DEM GRN PHON 4
4
Country/National DLP Green Pauline Hanson One Nation Democrats Communist
DLP 2
2
COM GRN
0 20 01
19 96 19 98
19 93
19 87 19 90
19 80 19 8 19 3 84
19 7 19 2 7 19 4 75 19 77
19 69
19 66
19 61 19 63
19 58
19 5 19 4 55
19 19 49 51
19 46
19 43
19 40
DLP 19 37
19 19 28 29 19 31 19 34
19 25
19 22
19 19 13 14 19 17 19 19
0 19 10
COM
COM
COM
Figure 1: Percentage Share of Formal First Preference Votes, by party, House of Representatives Elections, 1910-2001. “Anti-Labor” in the top panel refers to the Liberal Party (or its predecessors) plus the Country/National Party (plus various other anti-Labor parties, e.g., Liberal-Country League, Country Liberal Party). “Liberal” in the middle panel refers to the first Liberal Party (1910-1914), its Nationalist (1917-1929) and United Australia Party successors (1931-1943), and the second Liberal Party (1946-2001). The lower panel is scaled so as to highlight differences among the minor parties. Data: Hughes and Graham (1968), as corrected in Hughes and Graham (1974), and Newman (2002).
3
The Major Parties: History and Organization Australian Labor Party. The ALP is Australia’s oldest political party, pre-dating Federation, its origins lying in the growth of trade unions in the 1890s. A federal ALP was formed after Federation and elections to the first federal parliament. Labor’s ties to the union movement are critical to understanding the ALP’s history, organization, and political and ideological trajectory. Historically, trade unions affiliating with the ALP have supplied the bulk of the ALP’s membership, and via affiliation fees, the lion’s share of its revenue (unions also make additional contributions during election campaigns). Union affiliation fees accounted for roughly half of the New South Wales ALP’s revenue through the 1980s, and in recent years membership via union affiliations outstrip rankand-file memberships by a three-to-one ratio (SMH, 2002), a ratio that was as high as 4.5 to one in the 1960s. In return, trade unions are guaranteed representation in the Labor Party’s forums, in particular the state and national conferences, where the party’s policies are debated and adopted, and whose resolutions are (technically) binding on the parliamentary party. Unions are also guaranteed a specific number of votes in preselecting candidates (although these arrangements vary from state to state). As of mid-2002, affiliated unions are entitled to sixty percent of delegates to ALP conferences (the so-called “60-40 rule”), although this provision is due to be revisited at a special party conference in October 2002, amid calls for a “reform” of the party. These recent proposals for diluting union influence within the Labor Party are directly tied to the party’s recent losses at the federal level and broader changes in Australian society. Tension between the industrial (union) and political (parliamentary) wings is an enduring issue for social democratic parties everywhere, and has a long and celebrated history in Australia. For instance, prior to the 1963 election, Labor’s parliamentary leaders were photographed outside a Canberra hotel waiting the party’s platform to be finalized by a special meeting of the party’s national conference (the so-called “36 faceless men”). Partly as a result of this politically embarrassing incident, the party’s parliamentary leaders are now guaranteed seats on the party’s national executive committee. Recent calls for “reform” within the ALP can be seen as part of an ongoing struggle between the industrial and political wings of the party. Labor’s electoral success is widely believed to rest on building a broad coalition of support centered on its trade union base. Historically, Labor has attempted to build this coalition by pursuing what is best characterized as a social-democratic/reformist agenda, coupling interventionist macro-economic and industrial relations policies with progressive social policy. During the 1980s and 1990s, under the Hawke and Keating governments, Labor instigated a set of neo-liberal policy reforms at odds with its historical commitment to socialist goals, privatizing numerous government owned enterprises and deregulating significant sectors of the Australian economy. This apparent policy shift was widely interpreted as a major reorientation of the ALP’s priorities (e.g., Jaensch 1989; Maddox and Battin 1991; Beilharz 1994), striving to build a broader coalition of electoral support, beyond its trade union base. Indeed, the long-standing core
4
of Labor’s electoral strength – blue collar, trade union members – has steadily diminished over recent decades, with just 24.5% of employees unionized in August 2001, with the private sector unionization rate just 19.2% (ABS 2002a). Thus, Labor’s recent internal struggles are prompted by and tied to its broader political dilemma; to what extent can Australia’s diminishing numbers of trade unionists serve as the “base” or “core” of a political party competing for national office via a majoritarian electoral system? The Liberal Party. Robert Menzies, Australia’s longest serving Prime Minister (193941, 1949-1966), formed the Liberal Party in 1944 out of various anti-Labor progenitors listed earlier. After winning office under Menzies in 1949, the Liberal Party and its coalition partner, the Country Party (later the National Party) held office for twenty-three years. Menzies’ vision was for a party to represent the then-nascent middle class of postwar Australia, the “forgotten people” as he then referred to them (see Brett 1992). A noteworthy organizational feature of the Liberal Party is the power it vests in its parliamentary party and, in particular, its parliamentary leader. Unlike the Labor Party conferences, the organizational wing does not produce policy dictates that are binding on the parliamentary wing; the role of the organizational wing is formally described as a consultative or advisory role. However, Jaensch (1983, 120) suggests that from time to time (particularly in the face of electoral threat or defeat), the organizational wing becomes much more assertive. Within the parliamentary party, the leader is reasonably unconstrained in making appointments and promotions to cabinet level posts, subject to the constraints of coalition arrangements with the National Party. In addition, the party has no formal connections with other organizations (contrast Labor’s ties with trade unions). Moreover, the Liberal Party is structured on a federal model, with strong, relatively autonomous state branches (contrast the primacy of national forums in the Labor Party). Hancock (2000) documents the organizational challenge of building and sustaining the party around Menzies and his vision of a suburban-based middle class party, so as to viably compete with their well-organized Labor opponents, all the while maintaining a coalition with the Country Party. Jaensch (1996) makes a convincing case that antiLabor interests have never been particularly well organized, and that a genuine crisis lurks beneath an appearance of strength and success on the anti-Labor side of Australian politics. Labor’s success in national and state elections through the 1980s and 1990s (and continuing success at the state level) exposed severe structural problems in the Liberal Party: (1) a lack of ideological coherence, coming after 90 years spent largely in government, presenting themselves as “anti-Labor”/“anti-socialist”, to then have Labor under Hawke and Keating shed almost all traces of “socialism”; (2) the difficulty of building and maintaining a party to represent a plethora of “anti-Labor” interests, given a formal divide between the organizational and parliamentary wings of the party, and a federal model of party structure. In short, “the structure of the Liberal Party is sound, efficient and effective – for a party in government, with a strong and popular leader… For a party in opposition … without a clear, strong and popular leader, the structure is inappropriate – even a potential for disaster” (Jaensch 1996, 9; emphasis in original).
5
The “crisis” of the major parties. Jaensch’s (1996) insights into the weaknesses of the Liberal Party seem less pressing today after three consecutive federal election victories. At the time of writing (mid-2002), it is the Labor Party for which talk of “crisis” seems more appropriate. Recent commentary on the Labor Party focuses on themes that would be familiar to observers of the Liberal Party of the early 1990s (e.g., Warby 1994): “sclerotic organization”, declining branch membership, the poor state of party finances, poor prospects for recruitment and generational replacement, a lack of policy direction, a painful and seemingly futile search for strong and effective leadership, and an inability to win seats outside the party’s urban base (e.g., Button 2002; Ramsey 2002). Talk of a Labor crisis is in spite of the fact that as of mid-2002, Labor holds office in all states and territories (since the March 1995 elections in NSW, June 1998 in Queensland, August 1998 in Tasmania, October 1999 in Victoria, February 2001 in Western Australia, August 2001 in the Northern Territory, October 2001 in the ACT, and February 2002 in South Australia). It would seem that party “crisis” is the handmaiden of more than a parliamentary term or two in opposition in Canberra. More generally, Marsh (1992) presents a book length critique of the Australian two party system, arguing that both major parties are privileged by the electoral system, are shrinking from mass-based organizations to professional election-winning enterprises (see also Ward 1991), and have largely surrendered policy-making to special interest groups. Whatever may be true about the abilities of the major parties to drive policy innovation, the Australian party system can hardly be said to be in decay. Declining branch memberships and vote share aside, by any reasonable standard Australian electoral politics is largely about competition between the Labor and Liberal parties, thanks largely to the institutional arrangements discussed earlier (single-member House seats and compulsory, preferential voting). Constraints of space ensure that this survey of Australian political parties is necessarily incomplete. Contemporary perspectives on Australia’s political parties appear in the collection of essays edited by Simms (1996). Johns (2000) provides a recent comparison of the parties, focusing on differences in candidate nomination and other internal procedures. The major parties have generated a substantial body of research on their respective histories and organization: for example, see Crisp (1955), McMullin (1991) and Faulkner and Macintyre (2001) for the ALP, and West (1965), Tiver (1978) and Henderson (1994) for the Liberals, in addition to the works cited above. Bean (1997) provides a comprehensive analysis of support for the Australian Democrats, part of a collection of essays reviewing the Democrats (Warhust 1997). The coalition between the Liberal and the National parties is discussed extensively in the collection of essays edited by Costar (1994). XXX: Manifesto data? Castles and McKinley? Huber and Inglehart? Australian parties in comparative perspective? Voter Turnout
6
Given Australia’s system of compulsory voting, it is understandable that analyses of Australian electoral behavior focus less on questions of voter turnout and more on how voters choose among parties and candidates. Unlike democracies where turnout is voluntary, there is essentially very little variation in voter turnout in Australia to be explained: with turnout rates consistently around 95% of enrollees, there is simply not a lot of cross-sectional variation to study (see Jackman 2001 on the over-time effects associated with the introduction of compulsory voting). Accordingly, while the question “who votes?” (and why so many don’t) is central to the study of electoral behavior in the United States (e.g., Wolfinger and Rosenstone 1980), compulsory voting renders “who votes?” a matter of secondary importance in the Australian context. Nonetheless, several politically interesting avenues of research related to turnout remain under-explored. For instance, Australia’s turnout figures are computed with enrollees as the relevant population, which may overstate the “true” turnout rate; that is, of the citizenry legally obliged to enroll, who actually does so? Note that until a citizen actually enrolls, they are for all practical purposes invisible to the federal and state electoral commissions that enforce compulsory turnout laws and compile turnout statistics (while Australia’s compulsory turnout laws are enforced, it is much more expensive and draconian to compel enrolment). Little is known about Australia’s “hidden electorate”, or the political consequences of bringing the “hidden” electorate onto the electoral roll. Goot (1985, 194-195) surveys the small body of literature on the subject, noting the results of a 1983 AEC study finding that the “hidden” electorate was poorer than the enrolled population, but not necessarily less politically interested. Learning more about this population is an interesting avenue for future research. Among the few recent works examining turnout in Australian electoral behavior are McAllister’s (1986, 1987) aggregate-level analyses, aggregating 1970s census data to the level of Commonwealth electoral divisions; see also Goot’s (1985, 195-197) review of the literature up through the mid-1980s. Jackman (1999) considers the counter-factual of what might happen if turnout was made non-compulsory, arguing that turnout would fall substantially below the 80-90% range reported in surveys, when respondents are asked if they would turn out voluntarily; using 1996 survey data, Mackerras and McAllister (1999) examine the partisan consequences of a shift to voluntary voting, concluding that Labor would be a bigger loser than the Liberals. Voluntary postal voting was used for the 1997 election of delegates to the Constitutional Convention. All citizens on the electoral roll were mailed ballots to be self-completed and returned; 46.6% of ballots were returned to the Australian Electoral Commission, with considerable variation by age, gender and region, the three variables used in the AEC’s (1998) tabulations. Just 26.6% of enrolled Northern Territory males aged 18-25 participated, while 73.0% of enrolled males aged 65 and over in the Australian Capital Territory participated. Among Australia’s established political parties, only in the Liberal Party does one encounter calls for a repeal of compulsory voting (e.g., Minchin 1996), although it is unclear that this is a position endorsed by a majority of Liberal Party members (e.g., Georgiou 1996); further, surveys consistently show the mass public to be overwhelmingly in favor of compulsory voting (e.g., Goot 1985, 198-200; Lovell et al. 1998, 279). Accordingly, for the foreseeable future, with a regime of compulsory turnout under no major threat,
7
consistently high rates of voter turnout seem a fixed characteristic of Australian electoral politics. Vote Choice and the Party System: Aggregate-level evidence. The material bases of Australian electoral politics – the links between electoral choices and how and where people “make a living” (their income, job security, power in the labor market, family situation, residential tenure and so on) – can be easily seen in aggregate data. As its name implies, the Labor Party’s traditional supporters are workers; the Liberal Party, on the other hand, is historically associated with the anti-Labor interests of employers, owners and managers (ranging from self-employed professionals, small business owners through to big business). The National Party is for all practical purposes the anti-Labor party of non-urban Australia, representing farmers but also successful in winning the support of non-urban workers. Indeed, the dominant consensus (now challenged, see below) has been that three main parties are “the agents of economic or sectional interests” (Brett 2002, 40; citing Crisp 1973, chs 8-9). These patterns are easily seen via aggregate data analysis, facilitated by the fact that Australian census data are routinely aggregated to the level of Commonwealth electoral division. For instance, Kopras (1998, 2000) and Newman and Kopras (1998, 2001) provide aggregations from the 1996 Census to 1998 and 2000 electoral boundaries; earlier aggregations are provided by Kopras (1993), DTLG (1984) and McAllister (1977). Figure 2 plots ALP vote share against a number of social-structural indicators from the 2001 census (and 1996 census aggregated to 2000 electoral boundaries) and the ALP’s share of formal first preferences in the 2001 House of Representatives elections (Labor vote share is chosen as the outcome variable of interest since it contested all 150 seats in the 2001 election). Each plotted point in Figure 2 represents a House of Representatives seat, and the solid line (a “smoothed” or “moving” average) shows the extent to which ALP vote share increases or decreases as a function of the corresponding social-structural indicator (the dotted lines are 95% confidence bounds). The top left panel shows Labor Party vote share to decrease fairly dramatically as the proportion of the labor force employed in professional occupations increases, from as high as 60% of the vote in electorates with barely 10% of the labor force in professional occupations (e.g., the House of Representatives seats Chifley, Prospect and Fowler, forming a band in Sydney’s western suburbs, or Calwell and Scullin in Melbourne’s outer suburbs) to roughly 30%, on average, in electorates with the highest concentrations of professionals (e.g., the contiguous blocs of North Sydney and Bradfield on Sydney’s lower North Shore, Kooyong and Higgins in Melbourne’s inner eastern suburbs, and Curtin in Perth). A similar pattern results when we consider the relationship between ALP vote share and the proportion of the labor force unemployed. Labor vote share generally increases with the unemployment rate, at least for the bulk of the electorates with unemployment rates up to roughly 11%. These two relationships are moderately strong, accounting for 33% and 20% of the variation in ALP vote share, respectively, sufficient to remind us that Australian electoral politics retains a solid grounding in real economic conditions:
8
whatever else is true about Australian politics, the fact remains that Labor support is at its highest where proportions of non-professional workers are at their highest. However, the relationship between electorate median income and ALP vote share produces a less clear picture; this relationship is highly non-linear and accounts for only 20% of the variation in ALP vote share. To understand this pattern we need to consider a third variable, the urban/non-urban distinction. Generally, as median income increases, ALP support falls; again, electorates on Sydney’s lower North Shore and in Melbourne’s inner east are both high-income and Liberal Party strongholds. Labor Party support reaches its maximum among the poorer of the urban electorates (again, in Sydney’s west and in outer-suburban Melbourne), but falls away among still poorer electorates, found outside metropolitan areas. In fact, Labor’s support tends to be urban-based, apparent in the relationship between an electorate’s population density and ALP vote share (see the lower right panel of Figure 2). And since the poorest electorates are non-urban and mostly held by the National Party (although several are also held by the Liberal Party), we obtain the abruptly non-linear relationship between median income and Labor Party support. If we were to set aside rural seats, the pattern between Labor vote share and median income is roughly a simple linear decline, consistent with the substantive story underlying the relationships between Labor vote share and professional occupations and unemployment rates.
9
40 20
30
ALP 1st Preference Vote (%)
50
60
60 50 40 30
ALP 1st Preference Vote (%)
20
10
10 10
20
30
40
5
10
Professionals, % Lab Force (1996)
40 20
30
ALP 1st Preference Vote (%)
50
60
60 50 40 30
ALP 1st Preference Vote (%)
20
10
10
600
800
1000
1200
1400
0
10
Median Family Weekly Income (1996)
20
30
40
50
60
50 40 10
10
20
30
30
40
ALP 1st Preference Vote (%)
50
60
60
Non−English Speaking H’holds (% Pop)
20
ALP 1st Preference Vote (%)
15
Unemployed, % Lab Force (1996)
28
30
32
34
36
38
0.1
Median Age (1996)
1
10
50
500
5000
Population Density, Pop/sq km, log scale
Figure 2: Scatterplots of ALP Share of Formal First Preferences (2001 House of Representatives Election), against various social-structural indicators, aggregated from the 1996 and 2001 censuses. Each plotted point is a House of Representatives seat (red for seats won by Labor in 2001, blue for Liberals, green for Nationals and black for Independents, 150 in all). Each solid line is a locally-weighted regression line (Loader 1999), and the dotted lines are 95% confidence intervals. Data: ABS (2002b), AES (2002), and Kopras (2000).
10
Finally, two social indicators have noteworthy relationships with electoral outcomes. Non-English speaking (NES) households have long been considered a Labor stronghold (McAllister 1988; Jupp et al. 1989; McAllister 1992, 142-5; Zappalà 1998); the reasons for this include Labor’s embrace of a multiculturalist agenda from the mid-1970s (Forrest and Stockley 1988, McAllister and Makkai 1991) or the fact that NES migrants’ socioeconomic location predispose them to support Labor support (Economou 1994). The aggregate-level evidence in Figure 2 confirms that NES households tend to support Labor. In Sydney’s western suburbs, Labor strongholds such as Fowler and Watson have over 50% of the population residing in NES households (according to the 2001 Census); along with Blaxland, Prospect and Reid, these seats form an arc of high-NES/safe-Labor seats through Sydney’s west, with other western Sydney seats such as Grayndler, Lowe and Werriwa not far behind in terms of NES households and Labor vote share. Another contiguous bloc of five high-NES/safe-Labor seats lies in Melbourne’s suburbs, comprising Scullin, Calwell, Batman, Wills, and Gellibrand. The typical House of Representatives electorate has just 10% of its population residing in NES households, and at this level there is considerable dispersion in ALP vote shares; that is, while Labor does extremely well in electorates with high concentrations of NES households, high levels of Labor support are not necessarily confined to high NES electorates. Figure 2 also presents the relationship between the electorate’s median age and ALP vote share, finding a modest decrease in ALP support as median age increases, accounting for 13% of the variation in ALP vote share. Like some of the other aggregate relationships examined here, this relationship is dominated by the extremes: the youngest electorates tend (on average) to be Labor bastions (e.g., Werriwa, in Sydney’s outer-metropolitan south-west; Chifley, in outer-metropolitan Melbourne), while the oldest electorates tend (on average) to be reasonably safe Liberal or National seats (e.g., Lyne, on the New South Wales central coast; Fisher and McPherson, on Queensland’s Sunshine and Gold Coasts, respectively). XXX: MULTIVARIATE? Vote choice: micro-level models and evidence. The aggregate data analysis presented above is useful for grasping the broad contours of Australian electoral politics. But most theories of voter choice are pitched at the microlevel, and scholarly investigations of electoral behavior usually rely on data from representative sample surveys of the national electorate. Work by sociologists and political scientists at the University of Michigan, The Voter Decides (Campbell, Gurin and Miller 1954) and The American Voter (Campbell et al. 1960), provided a template for systematic research on electoral behavior that would be replicated in many other democracies, including Australia. The Michigan approach is defined by not just the survey methodology; surveys had been used for some time to study electoral behavior, including Australian electoral behavior, but largely through the prism of social-structural explanations. Rather, the Michigan model is distinctive in focusing on party identification, defined as “an individual’s affective orientation” to a political party,
11
measured by asking survey respondents (as in Australia in 2001) “Generally speaking, do you usually think of yourself as Liberal, Labor, National or what?” The Michigan team concluded that party identification is a relatively stable characteristic of an individual, usually inherited from one’s parents in much the same way as one might be socialized into a particular religious self-identification; that for most voters, most of the time, given knowledge of a voter’s party identification, one can reliably infer their evaluations of candidates, their policy preferences, and, most importantly, their vote choice. In fact, party identification not only structures these more transient political attitudes (candidate evaluations, stances on issues of the day), but is quite resistant in the face of “short-term forces” at odds with one’s party identification. A partnership between Don Aitkin and Donald Stokes (one of the original Michigan team) produced the first study of Australian politics in the Michigan style, Aitkin’s (1977) Stability and Change in Australian Politics, using data from two large scale surveys of Australian political attitudes conducted in 1967 and 1969 (Aitkin, Kahan and Stokes, 1975), closely modeled on Stokes’ work with David Butler in England (Butler and Stokes 1969). Aitkin’s substantive conclusions largely accorded with those of the earlier Michigan work, finding an even stronger role for party identification in the Australian context. Like the United States of the 1950s, Australian politics in the mid1960s was characterized by stability, with the overwhelming majority of respondents possessing a nearly immutable party identification, in turn, leading to high rates of stability in voting patterns. After this 1967 survey, it would be twenty years until the study of Australian political behavior had a regular set of national surveys along the lines of those available in the United States (the National Election Studies have been in the field every two years since 1952) and Great Britain (the British Election Study has been administered regularly since 1964); Aitkin conducted a follow-up study in 1979, and some politically-relevant questions were administered as part of the 1984 National Social Science Survey. But since 1987 the study of Australian political behavior has greatly benefited from a regular series of national surveys, the Australian Election Studies (AES), each administered immediately following a Federal election (McAllister and Mughan 1987; McAllister et al. 1990; Jones et al. 1993; Jones, McAllister and Gow 1996; Bean, Gow and McAllister 1998 and 2002), and the 1999 constitutional referendum on the republic question (Gow, Bean and McAllister 2000). Party identification. Table 1 presents rates of party identification and voting in accord with party identification, as reported in the 1967 study and in the six AES studies since 1987. XXX - FIX WITH CLIVE’S SUGGESTIONS. Reported rates of party identification in 1987 and 1990 were at least as high as in 1967, but appear to have fallen by roughly 10 percentage points in the 11 years between 1990 and 2001, from the low 90s to the low 80s (percent). By any reasonable standard, this was and is a high rate of party identification; McAllister (1992, 38) reports that the rate of no party identification in Australia is roughly half to a third the corresponding rate in the United States and the Britain. In addition, as the rate of party identification has fallen, so too has the rate of reporting a vote choice consistent with one’s reported party identification. While in 1967, roughly 90% of Australians had a party identification and of those, 90% voted in accord with that party identification, by 2001 the corresponding percentages had fallen to
12
the low 80% range; put differently, the proportion of the electorate voting in accord with an expressed party identification had fallen from about 83% in 1967 to 70% in 2001.
Percent with party identification Percent of identifiers voting with party identificationa Percent identifying with major partyb Percent of major party identifiers voting with party identificationa,c
1967
1987
1990
1993
1996
1998
2001
91.7
91.6
91.0
87.3
80.7
84.3
82.7
90.9
85.0
83.5
89.3
86.0
82.0
85.0
88.4
88.7
85.0
84.4
75.9
77.3
74.5
93.4
88.4
84.8
90.8
87.3
83.8
85.2
Table 1. Party identification and reported House of Representatives first preference vote, Australian National Political Attitudes Study Wave 1 (1967) and Australian Election Studies, 1987-2001. Notes: (a) respondents not reporting vote are dropped from the denominator of the percentage; (b) “major parties” are Labor, Liberal and Country/National; (c) includes Country/National identifiers reporting voting Liberal, and vice-versa.
In addition, Table 1 shows that the rate of identification with a major party (here defined as Labor, Liberal and Country/National) has fallen from the high 80% range to the mid 70% range, with a similar fall in the reported rates of voting in accord with that party identification, from the low 90% mark to the mid 80% level. That is, not only are fewer Australians reporting identifying with the major parties, but of those that do, fewer are reporting voting in accord with that stated identification. This evidence suggests a weakening of party loyalties in the Australian electorate, and is consistent with the macro-level falls in vote shares for the major parties reported in Figure 1, coinciding with the rise of minor parties such as the Australian Democrats and, to a lesser extent, the Greens, One Nation and independent candidates. But, as is the case with many aspects of Australian political behavior, we need to be mindful of the distinctive institutional constraints imposed by Australia’s electoral system. Preferential voting ensures that while the parties may not command the loyalties of citizens at 1960s 13
levels, votes leaking away to the longer menu of minor parties now available often find their way back to the major parties via preferences (see Bowler and Farrell 1991): irrespective of a voter’s first preference, a valid House of Representatives ballot requires a Labor candidate to be ranked above or below a Liberal (or National) candidate. In short, the preferential voting system helps reduce the seat turnover that might otherwise result from weakening voter loyalties. It should also be remembered that compared to the United States or Britain, rates of party identification and voting in accord with party identification in Australia remain high, and the decline in these rates is small relative to the decline in other English-speaking democracies (Bean 1996; Charnock 1996): speculation as to why Australian rates of party identification remain so high largely dwells on institutional reasons such as compulsory voting and the higher frequency of elections in Australia relative to the United States and Britain (McAllister 1992, 39; Charnock 1996, 266; cf Smith 2001, 56-8). Social-structural accounts. While party identification is a powerful explanation of vote choice, it is often criticized on the grounds that it is an attitudinal construct, deflecting analytical attention away from the connections between vote choice and voters’ social and economic conditions. In the Michigan account, party identification interposes social structure and vote choice, more causally proximate to vote choice than social structure. A rival, sociological approach, pre-dating the Michigan model’s emphasis on attitudinal variables (e.g., Berelson, Lazersfeld and McPhee 1954) investigates the extent to which vote choice is determined by social-structural variables, and in particular, variables tapping social class. This approach dominated understandings of Australian electoral behavior, at least prior to Aitkin’s deployment of the Michigan model. Alford (1963, 173) saw Australian politics as having been “dominated by class cleavages before and since its foundation as a nation.” Encel (1970, 96) asserted that “the correspondence between class identity and voting is so close…that the two may be regarded as largely alternative measures of the same general outlook”. And even Aitkin himself, writing with Kahan, described “occupational class as the rock on which the party system rests and a major influence on electoral behavior” (Aitkin and Kahan 1974, 477; all quoted in McAllister 1992, 152). Yet just three years later, summarizing the results of his survey-based work, Aitkin concluded that “Australian politics is the politics of parties, not of classes” (1977, 142). Other critiques of the sociological approach followed. Kemp (1978) concluded that class was not as important a determinant of voting behavior as the earlier literature had concluded, presenting an embourgiousement thesis: the decline of traditional blue-collar occupations and the growth of white-collar/public-sector occupations had undermined class distinctions in Australian society, and their importance for understanding electoral politics, both in the 1970s, and in the decades to follow. Quite simply, as the Australian working class (variously defined) was thought to be shrinking, electoral competition would move elsewhere, and the task for political analysis would be to understand vote choices among the politically decisive middle class. A lively debate ensued as to what exactly “class” might mean and how it might be measured (e.g., Connell and Goot 1979; Goot 1994) and a variety of measures of class are often used (e.g., self-assessed social class, occupation classifications) to tap the various facets of social-structural location
14
(e.g., Kelley and McAllister 1985; Hayes 1995; Charnock 1997; Lovell et al. 1998, 4689). XXX: REDO WITH MANUAL/NON-MANUAL??? The preference for multiple measures of social-structural location notwithstanding, Table 2 presents individual-level evidence on the link between vote choice and a single indicator of social-structural location, self-reported social class (contrast the multiple indicators used in the aggregate level analyses in Figure 2, above). Each cell of Table 2 shows Labor vote share minus Liberal vote share for the corresponding survey and group, with the actual Labor and Liberal shares shown in parentheses. The most politically relevant feature of the table is the moderate degree of political polarization along class lines, consistent with the aggregate-level evidence presented in Figure 2. The (selfdescribed) working class has always favored Labor, occasionally by a narrow margin (1996); the self-described middle class has always favored the Liberal Party, again, sometimes by narrow margins (1987), while the self-described upper class, never more than 1% of respondents, has overwhelmingly preferred the Liberal Party. Respondents reporting not belonging to any social class or offering no response (about 14% in any given study) have a slight tendency to favor Labor, with a plurality of these respondents favoring Labor in five of the seven studies.
15
1967 Working
Middle
None or No Response
Upper
“Index”
1987
1990
1993
1996
1998
2001
24
34
21
25
5
21
10
(51-27)
(60 – 26)
(51 – 30)
(59 – 34)
(43 – 38)
(50 – 29)
(45 – 35)
-36
-3
-17
-16
-24
-17
-22
(23-59)
(40 – 43)
(30 – 47)
(38 – 54)
(31 – 55)
(31 – 48)
(29 – 51)
-17
18
7
6
-8
6
5
(28-45)
(53-31)
(41-34)
(47-41)
(34-42)
(42-36)
(38-33)
-67
-13
-40
-40
-46
-57
-48
(0 – 67)
(31 – 44)
(30 – 70)
(29 – 69)
(18 – 64)
(13 – 70)
(20 – 68)
28
20
21
21
12
19
16
(51-23)
(60-40)
(51-30)
(59-38)
(43-31)
(50-31)
(45-29)
Table 2: Self-Assessed Social Class and Reported House of Representatives First Preference Vote, Australian National Political Attitudes Study (1967) and Australian Election Studies, 1987-2001. Each cell entry shows the difference between the reported Labor and Liberal vote shares (reported Labor-Liberal percentages shown in parentheses). The row labeled “Index” presents the difference between the percentage of self-reported working class respondents voting Labor and the percentage of self-reported middle class respondents voting Labor (actual percentages in parentheses). The distribution of self-assessed social class is quite stable over the seven studies, and is approximately Working 43%, Middle 42%, None or No Response 14%, and Upper 1%.
The interesting longitudinal feature in Table 2 is the decline in polarization along subjective class lines, tapped by the quantity labeled “Index” (the difference in Labor vote shares among working class and middle class identifiers). For the last three elections, this quantity has been below 20 points, a far cry from the 28 points recorded in the 1967 survey. This decline in class-based polarization reflects two related trends. First, the Liberal Party has steadily improved its vote share among working class identifiers since 1967, averaging 34% in the 1996, 1998, and 2001 elections, compared to just 27% in the 1967 survey. Second, the Labor Party has improved on its 1967 result among middle class identifiers, averaging about 30% in this group over the 1996-2001 series of elections, but as high as 36% in the 1987-1993 series of elections, compared to just 23% in the 1967 study. Both major parties, it would seem, have gone some way towards expanding beyond a particular class base. In particular, Table 2 confirms that reaching out across (subjective) class lines is vital to winning elections. For the Labor Party, winning over a third of the first preferences of middle class identifiers seems 16
something of a prerequisite for winning elections; contrast its performance among middle class identifiers in 1987-1993 with its performance in 1996-2001. For the Liberal Party, victory in 1966 came via a massive 36 point margin among middle class identifiers. But since 1987, the Liberal Party has won or retained office by keeping its vote deficit among working class identifiers small; this deficit averages just 12 percentage points in the 1996-2001 series of Liberal Party victories, but averages 27 percentage points in 19871993. The tension between the Michigan social-psychological approach and sociological models of electoral behavior can be found everywhere the Michigan model has been applied, and Australia is no exception. Clear empirical demonstrations of the superiority of one approach over the other are difficult to come by, since party identification is so causally proximate to (and measured contemporaneously with) vote choice and is usually highly correlated with social-structural variables; e.g., given that working class identifiers tend to be Labor Party identifiers, and Labor Party voters, isolating the independent causal contribution of attitudinal dispositions (party identification) and social-structural conditions (class location) on vote choice is difficult. In addition, there is no real possibility of running a “critical experiment”, say, randomly distributing party identifications or social-structural attributes across a population. It is clear that the influence of both party identification and social structure have waned since the 1960s. Yet over this period of mutual decline, rates of voting across subjective class lines have always been higher than rates of “defecting” from party identification. And since each party enhances the probability of electoral victory by reaching out across class lines, social structure could well further diminish as a determinant of voting behavior because of changed party strategies, rather than any fundamental change in the electorate (thinking about the “supply” of political issues from parties and candidates, in addition to the usual focus on “demand” from the mass electorate). Nonetheless, it is safe to assert that class continues to play a considerable though not “pre-eminent” role (Charnock 1997, 293) in shaping Australian political behavior and attitudes, and most scholars are rightly wary of overstating the “demise of class” in accounts of Australian political behavior. Perhaps the more interesting question is where in the causal scheme to locate social structural variables; prior to, or alongside party identification and other attitudinal variables? It is important to note that class, particularly self-assessed class, does not exhaust the content of “social-structure”. Other social-structural variables that have been of political relevance include gender (see “Gender Politics” in this volume), ethnicity (as the aggregate data analysis suggested) and religion (of diminished relevance in recent decades, with Catholics and secularists less likely than they once were to support Labor); McAllister (1992, ch 6) considers these “non-class” social-structural bases of party support. Vote choice: recent developments and controversies.
17
Australian electoral politics (as in other advanced industrial democracies) has seen three inter-related changes in recent decades: (1) a slight to moderate decline in the role of party identification and social structure in shaping vote choice; (2) a proliferation of minor parties and independents; (3) a decline in the first preference vote shares received by the major parties. In arguing that the role of class was diminishing (and would further diminish), Kemp (1978) suggested that values and clusters of issue-specific attitudes would become increasingly important determinants of vote choice. In the 1980s and 1990s, motivated by “real world” political events, scholars of Australian politics investigated precisely what values and attitudes might account for the three trends described above. Post-materialism. Across the industrial democracies, there is a growing body of evidence of value change among politically consequential segments of the electorate; as modern societies increasingly satisfy the more basic goals of personal survival and safety, the priority of “post-materialist” values increases, emphasizing social equality, individual rights, environmentalism and an improved quality of life as political issues (Inglehart 1977, 1990), at least for those believing that their “materialist” needs are satisfied. As elsewhere, Australia’s post-materialists tend to be tertiary educated and secular, and the proportion of the electorate fitting the post-materialist profile (those reporting a higher priority for, say, “more say in government” or a “more humane society” over “fighting rising prices” or a “stable economy”) tends to wax and wane in response to the state of the economy. In the 1996 Australian Election Study, 19% of respondents fitted the postmaterialist profile, 19% fitted the materialist profile, and most respondents were mixed (reporting no clear priority of post-materialist values over materialist values); the proportion of post-materialists was as high as XX in 19XX (see Lovell et al, 19XX, XXX). An interesting question concerns the political relevance of postmaterialism in Australian politics. Gow (1990) found postmaterialist attitudes to have no impact on the decision to vote for or against Labor in the 1990 Federal election, if voters’ views on the government’s economic performance are also included as predictors; on the other hand, if one distinguishes among vote choices among the minor parties, it appears that postmaterialist attitudes are important predictors of Green and Democrat support (Papadakis 1990; McAllister and Bean 1990). Using data from 1990 through to 1998, Western and Tranter (2001) conclude that postmaterialist values distinguish minor party voters from major party voters, but they do not distinguish Coalition voters from Labor Party voters. However, there is “some evidence of postmaterialists favouring Labor over the Liberal/National parties in House of Representatives elections in every election except 1993” (2001, 457). The dimensions of the Australian party system. Interest in the role of postmaterialist values and issues in Australian politics has led researchers to consider the dimensionality of Australian electoral politics. Most stable party systems are uni-dimensional, at least over the long run, in that electoral competition can be largely understood in terms of one ideological dimension, in turn reflecting class interests: e.g., perennial political issues such as welfare, public investment, and industrial relations, with Labor to the left of
18
center, and the Liberals to the right. In the absence of major social or economic upheaval, the Senate’s method of election provides the only serious challenge to the primacy of the traditional “left-right” continuum. Two-party competition (as is the case in the House) is by definition, bi-polar and uni-dimensional. But Senate elections ensure parliamentary representation for minor parties, and, in turn, issues that cut across the established left-right find voice on the national political stage. How many issue dimensions are there in Australian politics? And how has the rise and fall of various minor parties altered the character of those dimensions? The evidence on these questions is mixed, given the absence of a long series of comprehensive survey data. Using Gallup polls from 1943-96, Weakliem and Western (1999) suggested that support for the Greens and Democrats tapped a second dimension, cutting across a traditional ALP-DLP-Liberal dimension. Postmaterialism is a natural candidate for the substantive underpinning of this second dimension, as suggested by Western and Tranter (2001); McAllister and Vowles (1994) suggest that Labor has moved to accommodate the post-materialist dimension, as evidenced by the fact (noted above) that Labor tends to win more of the post-materialist vote than the Liberal Party. In a major challenge to the postmaterialist thesis, analysis of the 1998 AES by Charnock and Ellis (2003) find One Nation Senate voters to be as postmaterialist as Democrat voters, and considerably more postmaterialist than major party voters (see also Western and Trantor 2001, Table 8), counter to the notion of One Nation voters being economically threatened (by globalization, deregulation, the rise of the information economy) and hence presumably materialist in outlook. Charnock and Ellis argue that the dimensional structure of Australian is more fissiparous than previously thought, with up to six dimensions required for a “fuller picture of the divergent values” of Australian voters, at least in voting for the Senate: the issue dimensions are traditional “left-right”, “permissiveness”, immigration, the environment, Aborigines, and equal opportunities for women, although Charnock and Ellis also contend that the latter five can be combined into a “post-modernism” index, following Inglehart (1997). Race and Immigration: The rise and fall of Pauline Hanson and her One Nation party (PHON) generated intense interest in the extent to which attitudes about race and immigration form a critical issue dimension in Australian politics. Hanson swept on to the Australian political stage in 1996 after winning the previously safe Labor seat of Oxley (QLD), after being unendorsed by the Liberal Party. In her maiden speech in the House of Representatives, Hanson attacked policies giving “opportunities, land, money and facilities available only to Aboriginals” as separatist and based on the false assumption that Aboriginals are the most disadvantaged people in Australia (CPD Reps 1996, 3860), and that Australia was “in danger of being swamped by Asians” (CPD Reps 1996, 3862). PHON’s success in the June 1998 Queensland election (averaging 25% in the 79 seats it contested, winning 11 seats in the 89 seat parliament) raised expectations that the party would be a national political force. In the October 1998 Federal election, PHON contested 139 House seats out of 148, winning an average of 9%, but garnering over twice that in seven seats, six of which were in Queensland. Although Hanson did not retain her House seat (unhelped by a redistribution of Queensland’s Commonwealth
19
electoral divisions, although it is unclear whether Hanson would have retained Oxley had it not been redistributed), PHON won a Senate seat from Queensland. Numerous studies of the sources of the One Nation vote and its portent for Australian politics followed. These studies fall into two camps – aggregate and individual-level – the former matching census data to election returns (e.g., Davis and Stimson 1998; Reynolds 2000; Grant and Sorenson 2000), the latter using commercial opinion polls (Goot 1998) or the 1998 AES (Bean 2000; Bean and McAllister 2000; Denemark and Bowler 2002). Goot and Watson (2001) review this research, finding many similarities between the aggregate and individual-level studies, at least in terms of the socialstructural correlates of PHON support: more male than female, likely to have a working class identification and/or blue collar occupation, non-metropolitian (with studies of the Queensland election finding One Nation support to peak in the urban fringe within one to four hours driving time from Brisbane), older rather younger, and less well educated than average. But the 1998 AES data are unique in enabling an analysis of the political attitudes underlying PHON support: in a multivariate analysis controlling for the demographic correlates of PHON support (listed above), Goot and Watson (2001) test the effects of “economic insecurity”, “anti-Aboriginal” and “anti-immigration” attitudes, and “political dissatisfaction”. They find the last three help explain One Nation support, but “economic insecurity” does not; see also McAllister and Bean (2000, 396-397). Goot and Watson also take issue with the idea that PHON support as a “protest vote” or an expression of economic insecurity; rather, they point to the distinctiveness of PHON supporters on attitudes related to race. Race had been widely thought to be a relic in Australian politics, with a bipartisan commitment to multi-culturalism and promotion of Australia’s links with Asia and Aboriginal reconciliation emphasized under successive Labor governments. The broader question posed by the rise of Hanson concerned the “place of race” in Australian political ideology and the implications for the major parties and the party system. Jackman’s (1998) analysis of the 1996 AES found sizeable portions of Labor supporters were reporting quite conservative attitudes on race (ties with Asia, immigration, Aborigines) while nonetheless reporting traditional Labor attitudes (unions over big business, workers over employers), such that they were closer to Liberal and National party candidates than to Labor’s. According to Jackman (1998), race had all the hallmarks of a realigning dimension in Australian politics, and “coalition leaders could well be tempted to `play the race card’”. Indeed, some have suggested that the Coalition played the race card prior in anticipation of the 1996 election, that while Leader of the Opposition, Howard had clearly signaled his intentions to focus on issues of concern to the “mainstream”, repudiating Keating’s “Big Picture” focus on Aboriginal reconciliation and ties with Asia (Goot and Watson 2001, 184; Johnson 2000; Brett 1997, 9). The November 2001 Federal election was contested against the backdrop of the September 11 terrorist attacks in the United States, the Tampa incident, and an aggressive tightening of Australia’s immigration laws, and the Liberal Party’s campaign unambiguously centered on the government’s tighter immigration laws (Warhurst 2002). PHON was already starting to wither as a political force in its stronghold, Queensland
20
(see Williams 2001); some months before Tampa and September 11, in the Queensland state elections of February 2001, PHON contested 49 of 89 seats, averaging 20% of the vote, and winning just 3 seats. In the November 2001 Federal election, PHON contested 125 of 150 House seats, averaging 5.2%, and winning 5.5% of the Senate vote, but no seats. With PHON fading as a political force, but the Coalition winning its third consecutive election (and with border protection a central issue), a compelling research question is whether there has been a net permanent shift in partisan loyalties from Labor to the Liberal Party, a realignment of the Australian electorate around the issue of race. Recall that analysis of the 1996 AES found Labor partisans vulnerable to partisan conversion on race-related issues (Jackman 1998). Was PHON and the Hanson phenomenon a political interstice, a “way point” for voters realigning from Labor to Liberal? Thus, was 2001 a consolidation for the Liberals, completing the partisan conversion of formerly Labor identifiers? Or was 2001 the “return” of Liberal and National identifiers from a brief flirtation with One Nation, with some short-term defections from Labor? Will distinct stances on race-related issues (Aborigines, immigration, border protection), become another of the issue stances that distinguish the major parties, joining perennials of Australian politics such as tax, welfare, public investment, and industrial relations? These questions and the broader implications for Australian electoral politics are likely to motivate analyses of the 2001 election and beyond.
21
References ABS [Australian Bureau of Statistics]. 2002a. “Employee Earnings, Benefits and Trade Union Membership”. Catalogue Number 6310.0. Canberra. ABS [Australian Bureau of Statistics]. 2002b. 2001 Census of Population and Housing, Basic Community Profile (Commonwealth Electoral Divisions). First release. Canberra. AEC [Australian Electoral Commission]. 1998. 1997 Constitutional Convention Election Report and Statistics. Canberra. AEC [Australian Electoral Commission]. 2002. Election 2001 Results. Canberra. AHR [Australia, House of Representatives]. 1996. Parliamentary Debates. Volume 208. Parliament House: Canberra. Aitkin, Don. 1977. Stability and Change in Australian Politics. 2nd edition published 1982. Australian National University Press: Canberra. Aitkin, Don and Michael Kahan. 1974. “Australia: Class Politics in the New World.” In Richard Rose (ed.) Electoral Behavior: A Comparative Handbook. Free Press: New York. Aitkin, Don, Michael Kahan and Donald E. Stokes. 1975. Australian National Political Attitudes, 1967. Machine-readable data file number 7282. Interuniversity Consortium for Political and Social Research: Ann Arbor, Michigan. Alford, Robert A. 1963. Party and Society: The Anglo-American Democracies. Rand McNally: Chicago. Bean, Clive. 1996. “Partisanship and electoral behaviour in comparative perspective.” In Marian Simms (ed.) The Paradox of Parties: Australian Political Parties in the 1990s. Allen and Unwin: Sydney. Bean, Clive. 1997. “The Australian Democrats After Twenty Years: Electoral Performance and Voting Support.” In John Warhust (ed.) Keeping the Bastards Honest: The Australian Democrats’ First Twenty Years. Allen and Unwin: St Leonards. Bean, Clive. 2000. “Nationwide Electoral Support for One Nation in the 1998 Federal Election.” In Michael Leach, Geoffrey Stokes and Ian Ward (eds.) The Rise and Fall of One Nation. University of Queensland Press: St Lucia. Bean, Clive and Ian McAllister. 2000. “Voting Behaviour”. In Marian Simms and John Warhurst (eds.). Howard’s Agenda: The 1998 Australian Election. University of Queensland Press: St Lucia. Bean, Clive and David Gow and Ian McAllister. 1998. Australian Election Study 1998 [machine-readable data file]. Social Science Data Archives, Australian National Univesity: Canberra. Bean, Clive and David Gow and Ian McAllister. 2002. Australian Election Study 2001 [machine-readable data file]. Social Science Data Archives, Australian National Univesity: Canberra. Beilharz, Peter (ed.). 1994. Transforming Labor: Labour Tradition and the Labor Decade in Australia. Cambridge University Press: Melbourne. Berelson, Bernard R., Paul F. Lazarsfeld and William N. McPhee. 1954. Voting. University of Chicago Press: Chicago.
22
Bowler, Shaun and David M. Farrell. 1991. “Party loyalties in complex settings: STV and party identification.” Political Studies. 39:350-362. Brett, Judith. 1992. Robert Menzies’ Forgotten People. Macmillan: Melbourne. Brett, Judith. 1997. “John Howard, Pauline Hanson and the Politics of Grievance.” In G. Gray and C. Winter (eds.) The Resurgence of Racism: Howard and the Race Debate. Monash Publications in History. Number 24. Department of History, Monash University: Clayton. Brett, Judith. 2002. “Class, Religion and the Foundation of the Australian Party System: A Revisionist Interpretation.” Australian Journal of Political Science. 37(1):39-56. Butler, David and Donald E. Stokes. 1969. Political Change in Britain. Macmillan: London. Button, John. 2002. “Beyond Belief: What Future for Labor?” Quarterly Essay 6. Black Inc: Melbourne. Campbell, Angus, Philip E. Converse, Warren E. Miller and Donald E. Stokes. 1960. The American Voter. Wiley: New York. Campbell, Angus, Gerald Gurin and Warren E. Miller. 1954. The Voter Decides. Row, Peterson: Evanston, Illinois. Capling, Ann and Brian Galligan. 1992. Beyond the Protective State: The Political Economy of Australia’s Manufacturing Industry Policy. Cambridge University Press: Cambridge. Charnock, David. 1996. “Question-Wording Effects on the Measurement of Nonpartisanship: Evidence from Australia.” Electoral Studies. 15:263-268. Charnock, David. 1997. “Class and Voting in the 1996 Australian Federal Election.” Electoral Studies. 16:281-300. Charnock, David and P.Ellis. 2003. “Postmaterialism and Postmodernization in Australian electoral politics.” Electoral Studies. In press. Connell, Robert W. and Murray Goot. 1979. “The End of Class, Re-Run.” Meanjin. 38: 3-25. Costar, Brian (ed.). 1994. For Better or For Worse: the Federal Coalition. Melbourne University Press: Melbourne. Crisp, Leslie F. 1973. Australian National Government. 3rd edition. Longman Cheshire: Melbourne. Crisp, Leslie F. 1955. The Australian Federal Labor Party 1901-1951. Longmans, Green: London. Davis, Rex and Robert Stimson. 1998. “Disillusionment and Disenchantment at the Fringe: Explaining the Geography of the One Nation vote at the Queensland Elections.” People and Place. 6(3):69-80. Denemark, David and Shawn Bowler. 2002. “Minor parties and protest votes in Australia and New Zealand: locating populist politics.” Electoral Studies. 21:47-67. DPL [Department of the Parliamentary Library] 1999. Parliamentary Handbook of the Commonwealth of Australia. 28th edition. Department of the Parliamentary Library, Parliament of Australia: Canberra.
23
DTLG [Department of Territories and Local Government] 1984. “Australian electoral divisions 1984 and census 1981 collectors district codes.” Machine-readable data collection D0331. Social Science Data Archives, Research School of the Social Sciences, Australian National University: Canberra. Economou, N. 1994. “An Overstated Electoral Importance?: a note on `Ethnic’ Voting and Federal Electoral Outcomes.” People and Place. 2(4):45-51. Encel, Sol. 1970. Equality and Authority: A Study of Class, Status and Power in Australia. Cheshire: Melbourne. Faulkner, John and Stuart Macintyre (eds.). 2001. True Believers: The Story of the Federal Parliamentary Labor Party. Allen and Unwin: Sydney. Forrest, L. and D. Stockley. 1988. Australian Multiculturalism: A Documentary History and Critique. Multilingual Matters: Clevedon and Philadelphia. Georgiou, Petro. 1996. “The Case for Compulsory Voting.” Address to the Inaugural Meeting of the John Stuart Mill Society. Typescript. Parliament of the Commonwealth of Australia: Canberra. Goot, Murray. 1985. “Electoral Systems.” In Don Aitkin (ed). Surveys of Australian Political Science. Allen and Unwin: Sydney. Goot, Murray. 1994. “Class voting, issue voting and electoral volatility.” In J. Brett, J. Gillespie and M. Goot (eds.), Developments in Australian Politics. Macmillan: South Melbourne. Goot, Murray. 1998. “Hanson’s Heartland: Who’s for One Nation and Why.” In Nadine Davidoff (ed.). Two Nations: The Causes and Effects of the Rise of the One Nation Party in Australia. Bookman: Melbourne. Goot, Murray and Ian Watson. 2001. “One Nation’s Electoral Support: Where Does It Come From, What Makes It Different and How Does It Fit?” Australian Journal of Politics and History. 47:159-191. Gow, David John. 1990. “Economic Voting and Postmaterialist Values.” In Clive Bean, Ian McAllister and John Warhurst (eds.) The Greening of Australian Politics: The 1990 Federal Election. Longman Cheshire: Melbourne. Gow, David John, Clive Bean and Ian McAllister. 2000. Australian Constitutional Referendum Study, 1999 [computer file]. Social Science Data Archives, Australian National University: Canberra. Grant, Bligh and Tony Sorenson. 2000. “Marginality, Regionalism and the One Nation Vote: Exploring Socio-Economic Correlations.” In Marian Simms and John Warhurst (eds.). Howard’s Agenda: The 1998 Australian Election. University of Queensland Press: St Lucia. Hancock, Ian. 2000. The Federal Organization of the Liberal Party of Australia, 1944-1965. Melbourne University Press: Melbourne. Hayes, B.C. 1995. “The impact of class on political attitudes: a comparative study of Great Britain, West Germany, Australia and the United States.” European Journal of Political Research. 27:69-91. Henderson, Gerard. 1994. Menzies’ child: the Liberal Party of Australia, 1944-1994. Allen and Unwin: St Leonards. Hughes, Colin A. and B.D. Graham. 1968. Handbook of Australian Government and Politics. Australian National University Press: Canberra.
24
Hughes, Colin A. and B.D. Graham. 1974. Voting for the Australian House of Representatives, 1901-1964. Australian National University Press: Canberra. Inglehart, Ronald. 1977. The Silent Revolution: changing values and political styles among Western publics. Princeton University Press: Princeton. Inglehart, Ronald. 1990. Culture shift in advanced industrial society. Princeton, New Jersey: Princeton University Press. Inglehart, Ronald. 1997. Modernization and Postmodernization: Cultural, Economic and Political Change in 43 Societies. Princeton University Press: Princeton. Jackman, Simon. 1998. “Pauline Hanson, the Mainstream, and Political Elites: the place of race in Australian political ideology.’’ Australian Journal of Political Science. 1998. 33:167-186. Jackman, Simon. 1999. “Non-Compulsory Voting in Australia?: what surveys can (and can’t) tell us.” Electoral Studies. 18:29-48. Jackman, Simon. 2001. “Compulsory Voting.” In International Encyclopedia of the Social and Behavioral Sciences. Elsevier: Amsterdam. Jaensch, Dean. 1983. The Australian Party System. Allen and Unwin: Sydney. Jaensch, Dean. 1989. The Hawke-Keating Hijack: the ALP in transition. Allen and Unwin: Sydney. Jaensch, Dean 1994. Power Politics: Australia’s Party System. 2nd edition. Allen and Unwin: Sydney. Jaensch, Dean. 1996. “The Liberal Party `crisis’: prospects for reform.” In Marian Simms (ed), The Paradox of Parties: Australian Political Parties in the 1990s. Allen and Unwin: Sydney. Jaensch, Dean and David Mathieson. 1998. A Plague on Both Your Houses. Allen and Unwin: Sydney. Johns, Gary. 2000. “An Audit of Australian Political Parties.” Australian Journal of Political Science. 35(3):401-425. Johnson, Carol. 2000. Governing Change: From Keating to Howard. University of Queensland Press: St Lucia. Jones, Roger, Ian McAllister, David Denemark and David Gow. 1993. Australian Election Study 1993 [machine readable data file]. Social Science Data Archives, Australian National University: Canberra. Jones, Roger, Ian McAllister and David Gow. 1996. Australian Election Study 1993 [machine readable data file]. Social Science Data Archives, Australian National University: Canberra. Jupp, J., B.York and A. McRobbie. 1989. The Political Participation of Ethnic Minorities in Australia. Australian Government Publishing Service: Canberra. Kelley, Jonathan and Ian McAllister. 1985. “Class and party in Australia: comparison with Britain and the USA.” British Journal of Sociology. 36:383-420. Kemp, David. 1978. Society and Electoral Behavior in Australia. University of Queensland Press: Brisbane.
25
Kopras, Andrew. 1993. “Comparisons of 1991 Census Characteristics: Commonwealth Electoral Divisions.” Background Paper No.20. Department of the Parliamentary Library, Parliament of Australia: Canberra. Kopras, Andrew. 1998. “Electorate Rankings: Census 1996.” Background Paper 14, 1997-98. Department of the Parliamentary Library, Parliament of Australia: Canberra. Kopras, Andrew. 2000. “Electorate Rankings: Census 1996 (2000 Electoral Boundaries).” Research Paper 11, 2000-01. Department of the Parliamentary Library, Parliament of Australia: Canberra. Loader, C. 1999. Local Regression and Likelihood. Springer: New York. Lovell, David W., Ian McAllister, William Maley and Chandran Kukathas. 1998. The Australian Political System. 2nd edition. Longman: Melbourne. Mackerras, Malcolm and Ian McAllister. 1999. “Compulsory Voting, Party Stability and Electoral Advantage in Australia.” Electoral Studies. 18: 217-33. Maddox, Graham and Tim Battin. 1991. “Australian Labor and the Socialist Tradition”. Australian Journal of Political Science. 26:181-96. Marks, Gary and Clive Bean. 1992. “Sources of Electoral Suport for Minor Parties: The Case of the Australian Democrats.” Electoral Studies. 11:311-333. Marsh, Ian. 1992. Beyond the Two-Party System: political representation, economic competitiveness and Australian politics. Cambridge University Press: Melbourne. McAllister, Ian. 1977. “Federal election returns 1977 and 1976 census summary file.” Machine-readable data collection D0104. Social Science Data Archives, Research School of the Social Sciences, Australian National University: Canberra. McAllister, Ian. 1986. “Compulsory Voting, Turnout, and Party Advantage in Australia.” Politics. 21(1):89-93. McAllister, Ian. 1987. “Social Context, Turnout, and the Vote: Australian and British Comparisons.” Political Geography Quarterly. 6(1):17-30. McAllister, Ian. 1988. “Ethnic Issues and Voting in the 1987 Federal Election.” Politics. 23(2):219-247. McAllister, Ian. 1992. Political Behavior: Citizens, Parties and Elites in Australia. Longman Cheshire: Melbourne. McAllister, Ian and Clive Bean. 1990. “Explaining Labor’s Victory.” In Clive Bean, Ian McAllister and John Warhurst (eds.) The Greening of Australian Politics: The 1990 Federal Election. Longman Cheshire: Melbourne. McAllister, Ian and Clive Bean. 2000. “The Electoral Politics of Economic Reform in Australia: The 1998 Election.” Australian Journal of Political Science. 35(3):383-400. McAllister, Ian, Roger Jones, Elim Papadakis and David Gow. 1990. Australian Election Study, 1990 [machine-readable data file]. Social Science Data Archives, Australian National University: Canberra. McAllister, Ian and Toni Makkai. 1991. “The Formation and Development of Party Loyalties: Patterns Among Australian Immigrants.” Australia and New Zealand Journal of Sociology. 27(2):195-217.
26
McAllister, Ian and Anthony Mughan. 1987. Australian Election Survey, 1987 [machine-readable data file]. Data collected by A. Ascui. Data file produced by Roger Jones. Social Science Data Archives, Australian National University: Canberra. McAllister, Ian and Jack Vowles. 1994. “The Rise of New Politics and Market Libreralism in Australia and New Zealand.” British Journal of Political Science. 24:381-402. McMullin, Ross. 1991. The Light on the Hill: the Australian Labor Party, 1891-1991. Oxford University Press: Melbourne. Minchin, Nicholas. 1996. Letter to the Sydney Morning Herald. November 16. p44. Nethercote, John (ed.). 2001. Liberalism and the Australian federation. Federation Press: Annandale. Newman, Gerard. 2002. “Federal Election Results 1949-2001.” Research Paper No 9, 2001-02. Department of the Parliamentary Library, Parliament of Australia: Canberra. Newman, Gerard and Andrew Kopras. 1998. “Socio-Economic Indexes for Electoral Divisions.” Current Issues Brief 4, 1998-99. Department of the Parliamentary Library, Parliament of Australia: Canberra. Newman, Gerard and Andrew Kopras. 2001. “Socioeconomic Indexes for Electoral Divisions (2000 Electoral Boundaries).” Current Issues Brief 13, 2000-01. Department of the Parliamentary Library, Parliament of Australia: Canberra. Papadakis, Elin. 1990. “Minor Parties, the Environment and the New Politics.” In Clive Bean, Ian McAllister and John Warhurst (eds.) The Greening of Australian Politics: The 1990 Federal Election. Longman Cheshire: Melbourne. Papadakis, Elin and Clive Bean. 1995. “Independents and Minor Parties: the Electoral System.” Australian Journal of Political Science. 30:97-110. Ramsey, Alan. 2002. “The best little Prime Minister we never had.” Sydney Morning Herald. July 10. Reynolds, Paul. 2000. “One Nation’s Electoral Support in Queensland: The 1998 State and Federal Elections Compared.” In Michael Leach, Geoffrey Stokes and Ian Ward (eds.) The Rise and Fall of One Nation. University of Queensland Press: St Lucia. Rydon, Joan. 1968. “Compulsory and Preferential: The Distinctive Features of Australian Voting Methods.” Journal of Commonwealth Political Studies. 6:183-201. Simms, Marian (ed.). 1996. The Paradox of Parties: Australian Political Parties in the 1990s. Allen and Unwin: Sydney. SMH. 2002. “Crean must push on with reform.” Sydney Morning Herald. March 21. Editorial. Smith, Rodney. 2001. Australian Political Culture. Longman: Melbourne. Tiver, P.G. 1978. The Liberal Party: Principles and Performance. Jacaranda: Brisbane. Warby, Michael. 1994. “The Liberal Party: Past, Present and Future.” Quadrant. May. pp25-31. Ward, Ian. 1991. “The changing organisational nature of Australia’s political parties.” Journal of Commonwealth and Comparative Politics. 29(2):153—174.
27
Warhurst, John (ed.). 1997. Keeping the Bastards Honest: The Australian Democrats’ First Twenty Years. Allen and Unwin: St Leonards. Warhurst, John. 2002. “The Australian Federal Election of 10 November 2001.” Australian Journal of Political Science. 37(1):153-160. Weakliem, D. L. and M. Western. 1999. “Class voting, social change, and the left in Australia, 1943-96.” British Journal of Sociology. 50(4):609-630. West, Katherine. 1965. Power in the Liberal Party: a study in Australian politics. Cheshire: Melbourne. Western, Mark and Bruce Tranter. 2001. “Postmaterialist and Economic Voting in Australia, 1990-98.” Australian Journal of Political Science. 36(3):439-458. Williams, Paul D. 2001. “The Queensland Election of 17 February 2001: Reforging the Electoral Landscape?” Australian Journal of Political Science. 36(2):363-371. Wolfinger, Ray and Rosenstone. 1980. Who Votes? Yale University Press: New Haven. Zappalà, Gianni. 1998. “The Influence of the Ethnic Composition of Australian Federal Electorates on the Parliamentary Responsiveness of MPs to their Ethnic Sub-constituencies.” Australian Journal of Political Science. 1998. 33(2):187-209.
28
