The social bases of the party system have been measured in a number of ways in New Zealand. This post sets out the methodological basis of how this series of blog posts is measuring the social support bases of the political parties. [Read more below]
A common method of measuring the social bases of political party support is to compare the electoral performance of parties in certain localities (electorates or polling booths) with the known social characteristics of that location. Those characteristics can be taken from common knowledge or more sophisticated sources such as Department of Statistics census data. Correlations between parties and social structure can then be calculated. For a critique of this method, see: Mulgan (1997a: p.272). For example, Labour is said to be a party of the working class because it has traditionally performed strongly in those electorates which have few employers in them. This is a method used notably by Alan McRobie (1997).
A second method involves surveying voters. This allows more sophisticated, in-depth and reliable investigation into the relationship between social structure and political parties, as it deals directly with individuals. However, electoral survey research does not have a long history in New Zealand. Few voter surveys were carried out before the 1970s, and none were nationwide. The first useful survey was carried out in 1960 by Austin Mitchell of Dunedin Central voters. [This survey involved 551 random personal interviews in the two weeks following the 1960 election. Results were published by Mitchell (1962d). The survey was repeated in 1962-64, and published in Mitchell (1967).] Then in 1963 R H Brookes and Alan Robinson surveyed the Palmerston North, Karori, Miramar and Manawatu electorates. [This survey involved 1555 random personal interviews following the 1963 election.] The results were reported by Robinson (1967).
The first national survey appears to have been carried out following the 1975 general election by Stephen Levine and Robinson. The results from this were reported by Levine and Robinson (1976) and Levine (1979). Since then, Stephen Levine and Nigel S Roberts have regularly carried out and organised various multi-electorate and national voter surveys at election time (1989; 1991b; 1992a; 1997). The most comprehensive New Zealand voter surveys have been undertaken by the New Zealand Election Study (NZES). This project has carried out large postal surveys of voters after every election since 1987. This data has been analysed most thoroughly by Jack Vowles (1987b; 1992b; 1994c; 1998b; 2000; 2001;), who is New Zealand’s pre-eminent expert on the relationship between social structure and voting.
Other researchers using the NZES data include Peter Aimer (1992;; 1998), Hyam Gold (1985; 1989; 1992), and Marcus Ganley (1997), and their analyses are also referred to in this chapter. Another regular and extensive survey involving party choice and social structure is carried out by Massey University social scientists Paul Perry and Alan Webster (1999). [This survey involved 1555 random personal interviews following the 1963 election.] A number of surveys are also now undertaken for media outlets by polling companies. The results from these are also used in this series of posts to illustrate my argument.
The issue of the categorisation of class also requires some discussion. The main academic approach in New Zealand is to use the dichotomy of manual/non-manual workers to distinguish between working class and middle class voters, but as Vowles has pointed out, ‘such a purely manual/non-manual definition of class voting is open to serious question’ (Vowles, 1987b: p.25). However, Vowles has continued the manual/non-manual tradition because, ‘although members of the salariat [non-manual working class] share some of the interests of the traditional working class in that they are usually employees, their jobs are of a higher status and usually provide greater autonomy and reward’ (ibid: p.24). This method has been criticised by Shane Hanley:
This definition is both too narrow and too broad. On the one hand, it excludes all non-manual workers who are nonetheless wage-workers, and subject to essentially similar conditions and interests. On the other hand, the category includes a significant proportion of self-employed who do not share these inherent class characteristics. These definitions make it impossible to pinpoint how the whole working class votes (Hanley, 1988: p.11).
Some social and political scientists have also rejected class categories in favour of a socio-economic gradationalist approach, whereby voters are distinguished by income, thus producing groups such as ‘high-income earners’, ‘middle-income earners’, and ‘low-income earners’. For the purpose of this blog series it is not useful to ignore the data generated by either the manual/non-manual distinction or the gradationalist approach, but such material is reproduced here with reservation.
A more useful understanding of class in New Zealand is put forward by Brian Roper (1991: p.148; 1997), who uses a deliberately expansive and Marxist definition of the working class. In this, the working class is defined as those who are under the more or less continuous compulsion to sell their labour in order to purchase their livelihood. Such a definition includes the majority of white-collar employees, who arguably share similar class characteristics. Beyond the working class, according to Roper, are a number of other significant classes:
Within sociology, there is broad agreement that those who exercise effective control over the productive resources of society – capital, labour, and land – constitute the dominant class in New Zealand society. There is also broad agreement that those who are self-employed and own their own small businesses constitute a distinctive ‘old middle class’ or petite-bourgeoisie (Roper, 1997: p.94).
Using this general approach, political scientist Rob Steven has analysed census data from 1976, which showed the existence of a ‘ruling class’ in New Zealand numbering about 10% of the population, a ‘middle class’ of 12%, a ‘petite-bourgeoisie’ of 7%, and a ‘working class’ of about 72% (Steven, 1989). [For a variety of classifications of class in New Zealand, Bedggood (1980: p.72), Steven (1989) and Wilkes (1994) provide Marxist interpretations, while Pearson and Thorns (1983) provide Weberian interpretations.]