¹ Some examples, largely from the Anglo-American realm, are: Alford, Robert R., Party and Society: The Anglo-American Democracies (Chicago: Rand McNally, 1963)Google Scholar; Butler, David and Stokes, Donald, Political Change in Britain: Forces Shaping Electoral Choice (London: Macmillan, 1969)Google Scholar; Rose, Richard, ed., Electoral Behavior: A Comparative Handbook (New York: The Free Press, 1974)Google Scholar; Kemp, David, Society and Electoral Behaviour in Australia: A Study of Three Decades (St Lucia: University of Queensland Press, 1978)Google Scholar; Aitkin, Don, Stability and Change in Australian Politics, 2nd edn (Canberra: Australian National University Press, 1982)Google Scholar; Särlvik, Bo and Crewe, Ivor, Decade of Dealianment: The Conservative Victory of 1979 and Electoral Trends in the 1970s (Cambridge: Cambridge University Press, 1983)Google Scholar; Franklin, Mark N., The Decline of Class Voting in Britain: Changes in the Basis of Electoral Choice, 1964–1983 (Oxford: Clarendon Press, 1985)Google Scholar; Rose, Richard and McAllister, Ian, Voters Begin to Choose: From Closed-Class to Open Elections in Britain (London: Sage Publications, 1986)Google Scholar. This is merely a small sample of the literature pertaining to the topic. A complete list would be very long indeed.

² Two nationwide surveys of electoral behaviour have been conducted by academics, one by mail and the other by telephone, but both were somewhat limited in scope. These studies are reported respectively in Levine, Stephen and Robinson, Alan, The New Zealand Voter: A Survey of Public Opinion and Electoral Behaviour (Wellington: Price Milburn for New Zealand University Press, 1976)Google Scholar, and Gold, Hyam, ‘The Social Bases of Party Choice in New Zealand’ (paper presented to Australasian Political Studies Association Conference, Melbourne, 1984)Google Scholar. For further information about these and other survey studies of voting behaviour in New Zealand, see Bean, Clive, ‘An Inventory of New Zealand Voting Surveys 1949–84’, Political Science, 38 (1986), 172–84CrossRefGoogle Scholar. Within certain defined criteria, this document lists all academic sample surveys conducted in New Zealand with a focus on voting behaviour, many of which are mentioned at appropriate points throughout this article. The option of using data from commercial public opinion polls measuring voting behaviour to bolster findings from academic studies is not available in New Zealand since such data are not normally released for secondary analysis.

³ Robinson, Alan D., ‘Class Voting in Zealand: A Comment on Alford's Comparison of Class Voting in the Anglo-American Political Systems’ in Lipset, Seymour M. and Rokkan, Stein, eds, Party Systems and Voter Alignments: Cross-National Perspectives (New York: The Free Press, 1967), pp. 95–114, at p. 99.Google Scholar

⁴ These include: Milne, R. S., ‘Voting in Wellington Central, 1957’, Political Science, 10 (1958), 31–64CrossRefGoogle Scholar; Mitchell, Austin, ‘Dunedin Central’, Political Science, 14 (1962), 27–80CrossRefGoogle Scholar; Mitchell, Austin, ‘The Voter and the Election: Dunedin Central’ in Chapman, R. M., Jackson, W. K. and Mitchell, A. V., New Zealand Politics in Action: The 1960 General Election (London: Oxford University Press, 1962), pp. 171–200Google Scholar; Chapman, Robert, ‘The General Result’Google Scholar in Chapman, , Jackson, and Mitchell, , New Zealand Politics in Action, pp. 235–96Google Scholar; Brookes, R. H. and Ashenden, A. H., ‘The Floating Vote in Wellington and Palmerston North 1960–1963’, Political Science, 19 (1967), 17–39CrossRefGoogle Scholar; Robinson, , ‘Class Voting in New Zealand’Google Scholar; Mitchell, Austin, ‘Dunedin Central: A Long-Term Study of Voting’, Political Science, 19 (1967), 3–12CrossRefGoogle Scholar; Mitchell, Austin, Politics and People in New Zealand (Christchurch: Whitcombe and Tombs, 1969), Chapter 8Google Scholar; Levine, Stephen, The New Zealand Political System: Politics in a Small Society (Sydney: George Allen & Unwin, 1979), Chapter 5Google Scholar; Clements, Kevin, ‘Politics’ in Spoonley, Paul, Pearson, David and Shirley, Ian, eds, New Zealand: Sociological Perspectives (Palmerston North: Dunmore Press, 1982), pp. 143–67Google Scholar. The term ‘New Zealand-based’ is used advisedly, since more than one of these researchers was a temporary resident from elsewhere.

⁵ In particular: Alford, , Party and Society, p. 105Google Scholar; Rose, Richard and Urwin, Derek, ‘Social Cohesion, Political Parties and Strains in Regimes’, Comparative Political Studies, 2 (1969), 7–67CrossRefGoogle Scholar; Butler, and Stokes, , Political Change in Britain, p. 65Google Scholar; Korpi, Walter, ‘Some Problems in the Measurement of Class Voting’, American Journal of Sociology, 78 (1972), 627–42CrossRefGoogle Scholar; Powell, G. Bingham Jr, ‘Voting Turnout in Thirty Democracies: Partisan, Legal, and Socio-Economic Influences’, in Rose, Richard, ed., Electoral Participation: A Comparative Analysis (Beverly Hills, Calif.: Sage Publications, 1980), pp. 5–34Google Scholar; Crewe, Ivor, ‘Electoral Participation’ in Butler, David, Penniman, Howard R. and Ranney, Austin, eds, Democracy at the Polls: A Comparative Study of Competitive National Elections (Washington, DC: American Enterprise Institute for Public Research, 1981), pp. 216–63Google Scholar. Some of these works refer to New Zealand in passing; others include it in a cross-national study.

⁶ The sample sizes of these surveys were, respectively: 551, 163 and 607. The 1962 Dunedin Central survey, which is seldom cited by other writers presumably because of its small sample, was the first wave of a panel study which was completed in 1964. The 1966 Christchurch survey was conducted in two rounds, the first in August at which time 533 interviews were conducted and the second in December (after the general election of that year) when 74 new respondents were interviewed in addition to 479 of the original sample. Results reported from the survey appear to be based on the initial sample of 533. For more information see the works by Mitchell cited above and Bean, , ‘An Inventory of New Zealand Voting Surveys’.Google Scholar

⁷ Since the aim of this article is to help set the record straight with respect to the empirical evidence on the subject of class and party in New Zealand for those who work by it, occupation is used throughout as a measure of social class, in accordance with previous empirical research. It is none the less acknowledged that such a complex concept as class is captured by this measure only up to a point. Similarly, when occupation is presented herein as a manual/non-manual dichotomy this is done more out of a desire to be consistent with past research than out of a conviction that it is superior to other alternatives.

⁸ Alford, , Party and Society, p. 105Google Scholar. The source of Alford's calculation is Mitchell, , ‘Dunedin Central’, p. 48Google Scholar. The data were collected just after the general election of 1960 (although Alford reports the date as being 1961). For further details about the index of class voting see Party and Society, pp. 79–86.Google Scholar

⁹ Alford, , Party and Society, p. 102.Google Scholar

¹⁰ Mitchell, , ‘Dunedin Central: A Long-Term Study of Voting’, p. 6Google Scholar; Mitchell, , Politics and People in New Zealand, p. 213.Google Scholar

¹¹ Korpi, , ‘Some Problems in the Measurement of Class Voting’, p. 636Google Scholar; Powell, , ‘Voting Turnout in Thirty Democracies’, p. 15.Google Scholar

¹² Milne, , ‘Voting in Wellington Central, 1957’, pp. 32–6Google Scholar; Chapman, , ‘The General Result’, pp. 255–61Google Scholar; Robinson, , ‘Class Voting in New Zealand’, pp. 96–9Google Scholar. While providing broad support for the findings from Mitchell's surveys, the original data reported in each of these studies (bearing in mind that Robinson cites figures from each of the other three authors as well as presenting some analysis of his own) are not strictly comparable for a number of reasons, mainly to do with the measurement of social class. Milne uses subjective class, as does Robinson, plus a composite measure of socioeconomic status; Chapman constructs an intricate measure of socio-economic status (which he calls a ‘status and possessions index’) and applies it to aggregate statistics, producing high correlations with voting of an order that aggregate data often yield when the degree of association at the individual level is much weaker.

¹³ See Butler, and Stokes, , Political Change in Britain, p. 65Google Scholar, and Robinson, , ‘Class Voting in New Zealand’, p. 99Google Scholar. As well as citing the index score calculated by Alford, Robinson also quotes figures directly from Mitchell in building his argument.

¹⁴ Crewe, , ‘Electoral Participation’, p. 254.Google Scholar

¹⁵ There is also a marginal discrepancy between Powell's figure and that cited by Crewe (+43) apparently because Crewe used an earlier version of Powell's calculation that was computed on a slightly different basis: see Crewe, , ‘Electoral Participation’, p. 255.Google Scholar

¹⁶ The principal investigators of the 1963 survey were Ralph Brookes and Alan Robinson of Victoria University of Wellington and the data were generously supplied to the author for secondary analysis by Stephen Levine of Victoria University. The 1981 survey was designed and co-ordinated by the author in conjunction with Robert Chapman and John Prince of the University of Auckland, Jack Vowles of Massey University, Nigel Roberts of Victoria University of Wellington, Theodore Anagnoson and James Lamare of the University of Canterbury, and Antony Wood of the University of Otago, all of whose participation in the project the author gratefully acknowledges. For further details of both of these surveys see Bean, Clive S., ‘A Comparative Study of Electoral Behaviour in Australia and New Zealand’ (doctoral dissertation, Australian National University, 1984)Google Scholar, and Bean, , ‘An Inventory of New Zealand Voting Surveys’.Google Scholar

¹⁷ See Robinson, , ‘Class Voting in New Zealand’, p. 99Google Scholar. For a list of other works employing data from the 1963 survey, including some secondary analyses, see Bean, , ‘An Inventory of New Zealand Voting Surveys’.Google Scholar

¹⁸ See, for example, Pitt, David, ed., Social Class in New Zealand (Auckland: Longman Paul, 1977)Google Scholar; Bedggood, David, Rich and Poor in New Zealand: A Critique of Class, Politics and Ideology (Auckland: George Allen & Unwin, 1980)Google Scholar; Pearson, David G. and Thorns, David C., Eclipse of Equality: Social Stratification in New Zealand (Sydney: George Allen & Unwin, 1983).Google Scholar

¹⁹ Furthermore the impression gained is that the 1963 study was also more thoroughly planned and executed than some of the other surveys. None the less, the 1963 data are not without weaknesses and one objection that might be made is that the different level of class voting might be due to regional variation. This is unlikely in the New Zealand context, however, and even if it were true to some extent, the area surveyed in 1963 would almost certainly tend to be more representative of New Zealand as a whole than the parts of Dunedin and Christchurch covered in the surveys by Mitchell, which are central city districts, more socially homogeneous and less geographically diverse than the 1963 sample whose coverage includes a provincial town as well as city suburbs

²⁰ The 1975 survey was conducted in four urban electoral districts (in Wellington, Christchurch and Dunedin) and two rural ones (near Christchurch and Dunedin) by Stephen Levine of Victoria University of Wellington, Nigel Roberts of the University of Canterbury and Antony Wood of the University of Otago. The sample size was 899. The author is grateful to the original collectors for access to the data. For further details see Bean, , ‘An Inventory of New Zealand Voting Surveys’Google Scholar. Other evidence attesting to the weakness of occupation in shaping voting behaviour in recent times comes from Gold, , ‘The Social Bases of Party Choice’Google Scholar, and Vowles, Jack, ‘Social Structure, Political Attitudes, and Trade Unionism: An Analysis of Aspects of the Fourth Labour Government's Electoral Coalition’ (paper presented to New Zealand Political Studies Association Conference, Auckland, 1985).Google Scholar

²¹ Although some writers have claimed that class voting is higher in urban than rural settings (for example, Alford, , Party and Society, pp. 105 and 112Google Scholar, and Mitchell, , Politics and People in New Zealand, p. 213)Google Scholar, the empirical evidence suggests the reverse is true in New Zealand as well as in Australia. See Bean, , ‘A Comparative Study of Electoral Behaviour’, pp. 292–3Google Scholar. See also Kemp, , Society and Electoral Behaviour, Chaps 3 and 4.Google Scholar

²² For example, Lipson, Leslie, The Politics of Equality: New Zealand's Adventures in Democracy (Chicago: The University of Chicago Press, 1948).Google Scholar

²³ See Pearson, and Thorns, , Eclipse of Equality, p. 46.Google Scholar

²⁴ For the Australian figures (plus those from a 1969 survey showing a class voting index of + 27 for the whole sample and + 26 for the urban section), see Bean, , ‘A Comparative Study of Electoral Behaviour’, pp. 292–3Google Scholar. See also Aitkin, , Stability and Change, pp. 319–20Google Scholar, for a similar but not exactly comparable analysis. The British figures come from Särlvik, and Crewe, , Decade of Dealianment, p. 87.Google Scholar

²⁵ Supporting evidence for Australia can be found in Kemp, , Society and Electoral Behaviour, pp. 63–8Google Scholar. Admittedly, the bivariate treatment of class and party masks to some extent the complexities involved in this relationship, in particular in changes over time in Britain. But multivariate controls do nothing to alter the essential argument presented here. On the decline of the class-party link in Britain see, for example, Franklin, , The Decline of Class Voting in Britain.Google Scholar

²⁶ The proportions in each of the other age groups in the 1963 and 1981 data are, respectively: 25–34, 18 per cent and 22 per cent; 35–44, 25 per cent and 20 per cent; 45–54, 24 per cent and 15 per cent; 55–64, 16 per cent and 14 per cent.

²⁷ For the Australian evidence see Bean, , ‘A Comparative Study of Electoral Behaviour’, pp. 295–7Google Scholar. On Britain see Crewe, Ivor, Särlvik, Bo and Alt, James, ‘Partisan Dealignment in Britain 1964–1974’, British Journal of Political Science, 1 (1977), 129–90, pp. 172–5CrossRefGoogle Scholar. Among other things, the consistency of the pattern revealed by the New Zealand evidence helps strengthen confidence in the reliability of the 1963 data more generally.

²⁸ The total vote for each party, from which the figures in Table 1 are calculated, is in the 1963 survey: Labour, 43 per cent; National, 52 per cent; Social Credit, 5 per cent; and in the 1981 survey: Labour, 50 per cent; National, 35 per cent; Social Credit, 14 per cent. Other parties and candidates, who together received less than 1 per cent of the vote in both surveys, are excluded from the analysis.

²⁹ See Bean, , ‘A Comparative Study of Electoral Behaviour’, p. 301.Google Scholar

³⁰ For evidence of the shrinkage of the manual working class over the period (by about 5 per cent) see Bean, , ‘A Comparative Study of Electoral Behaviour’, p. 289Google Scholar. Data on aggregate electoral support for the parties in 1963, 1981 and elections in between can be found in Bean, Clive, ‘From Confusion to Confusion: The 1981 General Election in New Zealand’, Politics, 17 (1982), 108–20, p. 109CrossRefGoogle Scholar. The survey data actually overstate the true electoral movement towards the Labour party over the period.

³¹ On this theme see, for example, Alford, , Party and Society, pp. 114–22Google Scholar; Kemp, , Society and Electoral Behaviour, especially pp. 352–7.Google Scholar

³² See Sartori, , ‘From the Sociology of Politics to Political Sociology’ in Lipset, Seymour Martin, ed., Politics and the Social Sciences (New York: Oxford University Press, 1969), pp. 65–100Google Scholar. The argument owes much to the work of Lipset, Seymour M. and Rokkan, Stein, ‘Cleavage Structures, Party Systems, and Voter Alignments: An Introduction’Google Scholar in Lipset, and Rokkan, , eds, Party Systems and Voter Alignments, pp. 1–64Google Scholar, and also has overtones of Downs, Anthony, An Economic Theory of Democracy (New York: Harper & Row, 1957), pp. 24–31Google Scholar. For an empirical treatment see Kelley, Jonathan, McAllister, Ian and Mughan, Anthony, ‘The Decline of Class Revisited: Class and Party in England, 1964–1979’, American Political Science Review, 79 (1985), 719–37CrossRefGoogle Scholar. New Zealand evidence in support of the argument is that whereas in 1963 the class distinctiveness of the Labour party was greater than that of the National party, by 1981 Labour's class distinctiveness was lower than the National party's (that of both parties having declined considerably, of course).

³³ See, for example, Gustafson, Barry, Social Change and Party Reorganization: The New Zealand Labour Party Since 1945 (Beverly Hills, Calif.: Sage Professional Papers in Contemporary Political Sociology, 1976), pp. 29–30.Google Scholar

³⁴ There is evidence that citizens in Western societies are increasingly expressing their political views through ‘unconventional’ forms of political behaviour, an indication perhaps that the established political institutions and channels of participation are perceived as unsatisfactory and increasingly irrelevant. See Barnes, Samuel H., Kaase, Max et al. , Political Action: Mass Participation in Five Western Democracies (Beverly Hills, Calif.: Sage Publications, 1979).Google Scholar

³⁵ A recent argument to this effect in the British context is contained in Heath, Anthony, Jowell, Roger and Curtice, John, How Britain Votes (Oxford: Pergamon Press, 1985), pp. 28–43.Google Scholar

³⁶ See, for example, Milne, , ‘Voting in Wellington Central, 1957’, pp. 32–8Google Scholar; Mitchell, , ‘Dunedin Central’, pp. 43–52Google Scholar; Levine, and Robinson, , The New Zealand Voter, pp. 131–41.Google Scholar

³⁷ For a similar analysis of the 1967 and 1979 Australian surveys producing very similar results see Bean, , ‘A Comparative Study of Electoral Behaviour’, p. 309Google Scholar. Indeed all of the Anglo-American democracies are at the lower end of a scale of the influence of social structure in determining electoral behaviour. See Rose, Richard, ‘Comparability in Electoral Studies’Google Scholar, in Rose, , ed., Electoral Behavior, pp. 3–25, p. 17.Google Scholar

³⁸ Evidence of the aggregate decline in religion can be found in Bean, , ‘A Comparative Study of Electoral Behaviour’, p. 272.Google Scholar

³⁹ See Powell, G. Bingham Jr, ‘Party Systems and Political System Performance: Voting Participation, Government Stability and Mass Violence in Contemporary Democracies’, American Political Science Review, 75 (1981), 861–79.CrossRefGoogle Scholar

⁴⁰ See Bean, , ‘A Comparative Study of Electoral Behaviour’Google Scholar, which also contains examples of historical, social and cultural similarities between the two countries.

⁴¹ For example, the proportion identifying with or ‘feeling closer to’ a political party in the 1981 survey was 93 per cent. Of party identifiers, 36 per cent were ‘very strong’ supporters of their chosen party and 88 per cent of major party supporters voted in accordance with their party identification. These are high figures by international standards and, again, very similar to those for Australia. See Bean, , ‘A Comparative Study of Electoral Behaviour’, pp. 315–33Google Scholar. By contrast, as far back as 1974 equivalent figures for Britain were quite a bit lower. See Crewe, , Särlvik, and Alt, , ‘Partisan Dealignment in Britain’, pp. 142–7.Google Scholar