The relationship between social class and electoral behaviour in Australia is changing. In the 1960s the majority of working-class voters supported the Labor Party; in 1996 more working-class voters (manual workers and people in routine clerical jobs) voted for the Coalition than for Labor. A perception among some members of this group that Labor no longer reflects either their cultural values or their economic interests may help explain this shift..

From the time of Federation in 1901 politics in Australia have been dominated by two major political groupings: the Australian Labor Party and a relatively stable non-Labor grouping. Since the last war this non-Labor group has consisted of a coalition between the Liberals and the National Party.^[1] Because of the stability of the Liberal-National Party Coalition it is reasonable to see post-war Australian politics as a two-party system. Minor parties do have influence because they can direct their supporters' second preferences to one or other of the two major groupings, and candidates from minor parties have played a key role from time to time in the Senate. But no candidate representing a minor party has won a seat in the House of Representatives during the post-war period (though, since the 1990 Federal election, a few independents have been successful).

Many political scientists have explained the stability of the two-party system in terms of class-based voting and high levels of party loyalty which transcend the issues and personalities of the day.^[2] Most Australians identify with a mainstream political party. Even if they do not vote for this party on every occasion they will usually return to it. Some analysts add that the stability produced by predictable class-based loyalty is reinforced by the ability of political elites to confine political debates to narrowly focussed economic issues.^[3] This combination of class-based loyalty and appeals to economic issues should mean that the working class support the Labor Party while the middle, upper middle and upper classes support the Coalition.

There is no single accepted way of describing these classes and I am here following the model described by Chris Chamberlain in his book Class Consciousness in Australia.^[4] His upper class is numerically small but very wealthy, but the upper-middle class effectively runs the system and includes senior managers, business people and upper level professionals. The middle class consists of people who have a degree of autonomy at work. It includes lower level professionals, white-collar workers in relatively autonomous positions, and blue-collar workers who are self employed or in supervisory positions. The working class includes skilled and unskilled blue-collar employees as well as people working in routine white collar jobs.

A decline in the relative size of the blue-collar workforce has meant that Labor has had to develop its appeal to white-collar workers but in Chamberlain's model many of these people are in fact 'working class'. It is generally held that this development has not diminished the role of social class as an explanation for voting patterns. For example, the argument offered by David Kemp in the late 1970s that the relationship between class (as measured by occupation) and voting was becoming much weaker^[5] has its adherents but has not been widely accepted.^[6]

Table 1.

[Table 1] sets out the results of the first-preference vote for the House of Representatives in the 20 Federal elections held between 1949 and 1996, as well as the two-party preferred vote which results after minor party preferences have been distributed. While the proportion of seats held by Labor and the Coalition has fluctuated widely, their proportions of the vote are fairly stable. Only twice does one or the other reach or exceed 50 per cent; and the Coalition never, in this time period, fell below 41 per cent. Labor, however, has had three bad years (1977, 1990 and 1996) when its primary vote fell below this. Of course below forty per cent does not have to mean 'catastrophic'; in 1990, despite gaining only 39.4 per cent of the primary vote, Labor still managed to hold on to Government. In 1990 third parties attracted a high proportion of the vote and an unusually large number of third-party preferences went to Labor and kept it in power. In 1931 and 1934 Labor's vote was badly affected by the depression, Lyons' defection to the Nationalists and the challenge from Lang in New South Wales; it fell to 27.1 per cent in 1931 and 26.8 per cent in 1934.^[7] But the result for Labor in March 1996, when it won 38.8 per cent of the primary vote, was its worst ever in the series illustrated in Table 1. This result led to a resounding defeat with the former Government now holding only 49 of the 148 seats in the lower house. (Independents hold five and the Coalition has the other 94.)

The accompanying article on patriotism, immigration and the 1996 election explores some of the more immediate reasons behind the erosion of the Labor vote and the defeat of the Keating Government. It argues that, while Labor remained fairly strong among university graduates, retaining its appeal among new-class cosmopolitans, it lost ground with the less well-educated voters and that pride in Australia's history and opposition to current levels of immigration were associated with the vote for the Coalition.

Table 2.

[Table 2] draws on the Australian Electoral Study (AES).^[8] It shows that, while 90 per cent of those who had voted for the Coalition in 1993 stayed with the Coalition parties in 1996, fewer than three quarters of the 1993 Labor voters stayed with Labor. Table 3 shows the vote in 1993 and 1996 by occupation. The 1993 data depend on respondents remembering how they voted three years previously and thus may be less reliable than their reports of the vote which they had just cast. Nevertheless, the data show that in 1993 Labor had been in front of the Coalition in five occupational groups: para-professionals, trades, plant and machine operators, elementary clerks and labourers. It had also been in front with the group labelled 'other', which consists of people who have never had a paid job. In 1996 Labor was ahead in no occupational group. The group labelled 'elementary clerks' consists of filing and sorting clerks, telephonists, messengers, sales assistants, cashiers, cleaners and kitchen hands. Many of them are white- collar workers but, in Chamberlain's definition, they are working class.

Table 3.

[Table 3] shows that voting is associated with occupation; the Labor vote in 1993 was almost twice as high among labourers as it was among administrators and managers. But Table 3 also shows that, while the Labor vote fell across the board between 1993 and 1996, it fell most sharply among four occupational groups: labourers, tradespeople, plant and machine operators, and elementary clerks. With the exception of some tradespeople who may have been self-employed or working in supervisory positions, all of the people in these groups would count as working class in the classification outlined above.

Table 4.

[Table 4] is restricted to people who voted Labor in 1993. It sets out the 1996 vote for all 1993 Labor voters and for the six occupational categories which had given Labor the strongest support in 1996. It shows that the tradespeople and the plant and machine operators who had voted Labor in 1993 and who changed their vote in 1996 disproportionately moved to the Coalition.

Table 5 and 6

[Tables 5] and [6] present voting patterns by type of occupation between 1967 and 1996 and provide a historical context for this shift. The earlier data are taken from Graetz and McAllister,^[9] and are based on national surveys which were similar to the AES. (The earlier data are in fact coded as non-manual and manual, but the terms 'middle and upper-middle class' and 'working class' have been used here because the elementary clerks are included in the working class data for 1993 and 1996.)^[10]

Tables 5 and 6 allow us to put the results of the 1996 AES results into a context but they do not provide very many data points. For example, the dramatic elections of 1972, 1975 and 1983 are missing. Nevertheless, Table 5 shows that support for Labor among middle- and upper-middle-class voters reached a peak in 1984, when nearly half the people in this category voted for the Hawke Government, and that this support then declined. A similar, but higher, peak is shown for working class voters in Table 6, but here the decline has been more precipitous.^[11] While the middle- and upper-middle-class Labor vote dropped by 15 per cent between 1984 and 1996, it fell from an unusually high point and in 1996 it was still higher than it had been in 1967. In contrast, the Labor vote among the working class in 1984 was only five or six per cent higher than in the 1960s and 1970s. But by 1996 it had fallen by 23 per cent and was then much lower than it was in 1967. Indeed, as both Table 3 and Table 6 show, in 1996 manual workers and people in routine clerical positions were all more likely to vote Coalition than Labor.

In 1988, on the basis of clear evidence on shifts in party identification, Graetz and McAllister reported that the Liberal-National Party Coalition had been in a state of long-term decline since the 1960s. They concluded that since the heyday of the Menzies era in the 1950s and early 1960s the Liberals had failed either 'to project a distinctive enough profile to the voters or to attract sufficiently able leaders to arrest the decline'.^[12] The general trend from 1967 to 1984 had been a shift of support away from the Liberals and towards the Labor party.^[13] The current evidence suggests that the trend is now swinging the other way. Labor's loss of support among the working class may have been accentuated by factors specific to the 1996 election but it seems to be part of a long-term shift.

How should this be explained? Of course, the phenomenon of the manual worker who does not vote for a labour party is not new; it has merely grown. After all, Table 6 shows that between 35 and 41 per cent of the working class were voting for the Coalition between 1967 and 1993. Indeed, much research in political sociology has been devoted to analysing the support which some workers give to conservative or right-wing parties. Chamberlain's book, for example, is devoted to the problem of working-class adaptation to inegalitarian politics. But there is less research on the corresponding riddle of why middle- class people and professionals should vote Labor, as growing numbers of them do.

The voting patterns of people in different occupational groups are mutating. Why? Does class matter less in Australia in the 1990s than it did in the 1960s? Given current levels of unemployment and economic anxiety this seems unlikely. Perhaps we need to ask a different question. Does Labor still represent the working class?

The Australian Labor Party has changed. It is no longer a party dominated by unions and unlettered workers. The branches have grown in influence and university-educated activists now play an increasingly important role not only in these branches, and the Federal party structure, but within the Australian Council of Trade Unions as well.^[14] In the 1960s and 1970s most of the new-class Labor activists upheld the economic interests of working-class Labor supporters (regulation of the economy, including regulation of wages and working conditions, and a decent welfare system) but they differed from many of them on non-economic issues (such as the White Australia policy, the Vietnam war, overseas aid and, later on, multiculturalism and the refugee intake). ^[15]

Labor intellectuals' positions on many of these questions were morally sound and made practical good sense. Their position on others, such as multiculturalism, may have begun with good intentions but ended less happily.^[16] But irrespective of the merits of these ideas, sceptics, including Australian workers, were not always converted to them; rather, they were outmanoeuvred and led to feel that any objections were inadmissible, unfashionable and wrong.

During the 1980s the Labor party élite not only promoted their cultural values enthusiastically, they also abandoned many of their older protectionist economic goals. Many workers who had suffered the imposition of alien cultural ideas may have experienced Labor's conversion to economic rationalism as a double betrayal. The workers' party which had been led by Curtin and Chifley and Calwell spoke for the battlers. The data in the accompanying article on patriotism, immigration and the 1996 election give some plausibility to the idea that many of these battlers do not now see Labor as their champion. In their eyes the Labor Party may have become, not a modernised defender of the working class, but a conglomeration of special interest groups, led by an intellectual élite with no especial affection for Australia or Australians.

If this analysis is correct we can say, not that class-based voting has changed in Australia, but that the nature of social class is changing. Class still matters in Australian politics but it is expressed through culture and ideas as much as through economic claims. Of course, economics is important but other questions should be seen as part of the political debate as well. If the two forces of political stability have in the past been class-based party loyalty and an élite strategy of keeping non-economic questions off the political agenda, the first has changed and the second has become a force, not for stability, but for instability.

Some parochials now sound harsh and ugly but this may be because their concerns have been ignored for too long. A political élite which has focussed the public debate on economic issues and dismissed popular feeling about demography and identity bears some of the responsibility for this, but so too do the new class. People who are unhappy with the way in which public issues are discussed have a duty, not to put a damper on others whom they regard as inferior or distasteful, but to set an example of full and open discussion. Leadership consists not in suppressing debate but in encouraging it. This is done by offering cogent arguments, credible evidence and a sound policy framework.

Acknowledgment

The author wishes to thank Virginia Rapson for her analysis of the data.

References

1 The Liberals are the dominant partner in the Coalition, normally gaining well over 75 of the two parties' combined vote. (The National Party represents rural interests and was known as the Country Party until 1982.) The Australian Labor Party dates from 1890. In contrast, the Liberal party was founded in 1944. But the Liberal Party is the more-or-less direct descendent of three earlier non-labour parties (the first Liberal Party, founded in 1909, then the Nationalist Party, and then the United Australia Party). The Country (National) Party dates from 1913.

2 This work is summarised in D. Mayer, Democracy in Australia, Dellasta, Melbourne, 1988, pp. 90-94. While Goot is less inclined to see evidence of stability, he also argues that voting continues to be linked to occupation and class-based issues. See 'Class voting, issue voting and electoral volatility', in J. Brett, J. Gillespie and M. Goot (Eds), Developments in Australian Politics, Macmillan, Melbourne, 1994, pp. 179-180. See also B. Graetz and I. McAllister, Dimensions of Australian Society, Macmillan, Melbourne, 1988, pp. 262, 264, 287, 294.

3 See I. McAllister, Political Behaviour: Citizens, Parties and Elites in Australia, Longman Cheshire, Melbourne, 1992, p. 16. The idea that there are deeply divisive questions which, in the interests of stability, must be kept off the political agenda is derived from élite theory. See G. L. Field and J. Higley, Elitism, Routledge and Kegan Paul, London, 1980, pp. 37, 72, 117, 129.

4 See C. Chamberlain, Class Consciousness in Australia., Allen and Unwin, Sydney, 1983, pp. x-xi.

5 See D. Kemp, Society and Electoral Behaviour in Australia, University of Queensland Press, St. Lucia, 1978, pp. 87-88. McAllister also argues that class based voting has declined, 1992, op. cit., pp. 157-161.

6 Mayer summarises the opposition to this view, Mayer, op. cit. p. 94.Goot argues that the association between occupation and voting may have weakened but that the role played by class-based issues is still strong. He argues that we do not have to find either class-based or issue-based voting; the issues which affect voting may be linked to class. See 1994, op. cit., pp. 178-180

7 See Goot, op. cit., p. 177.

8 The study was conducted by R. Jones, I. McAllister and D. Gow. See Australian Election Study [computer file], Social Science Data Archives, The Australian National University, Canberra, 1996. The original researchers bear no responsibility for the analysis and interpretation of their data in this present article. Sampling details of the AES are set out in the accompanying article.

9 Graetz and McAllister, op. cit. p. 285.

10 The vote for the elementary clerks was analysed separately and was found to be almost identical with the aggregate figures for three manual groups in 1996.

11 Goot presents a more detailed time series based on opinion polls on voters' intentions taken before each election from 1943 to 1993. Between 1943 and 1963 the proportion of manual workers intending to vote Labor averaged 66 per cent. This dipped in 1966 (to 51 per cent) but reached a new peak of 63 per cent in 1983 (and was 60 per cent in 1984). Between 1984 and 1993 it averaged 57. 5 per cent. Goot also finds a peak of Labor support among non-manual workers in 1984, when 46 per cent said they would vote Labor, the highest proportion for non-manual workers in his time series. Goot, op. cit., p. 164.

12 Graetz and McAllister, op. cit., pp. 280-281

13 ibid., p. 290

14 Calwell's memoirs document some of these changes. See A. A. Calwell, Be Just and Fear Not, Lloyd O'Neil, Melbourne, 1972, especially pp. 257-258. See also M. Grattan, 'The Australian Labor Party', in H. Mayer and H. Nelson (eds), Australian Politics: A Third Reader, Cheshire, Melbourne, 1973, pp. 399-401, 403; H. Albinksi, Australian External Policy Under Labor, University of Queensland Press, Brisbane, 1977, p. 72; and J. Cairns, 'Intellectuals and the ALP', Meanjin , vol. xxv, 1966, p. 113.

15 See K. Betts, Ideology and Immigration: Australia 1976 to 1987, Melbourne University Press, Melbourne, 1988, pp. 49-52,74-81, 94-98, 102-119, 142-147, 160-168.

16 See for example accounts of ethnic branch stacking in the Labor Party. E. Healy, 'Ethnic ALP branches - the Balkanisation of Labor', People and Place, vol. 1, no. 4, 1993, pp. 37-43, and of ethnic quotas in B. Birrell, 'Our Nation: the vision and practice of multiculturalism under Labor', People and Place, vol. 4, no. 1, 1996, pp. 19-27.

Back to People and Place Home Page