Cleavage-based voting behavior in cross-national perspective: Evidence from six postwar democracies

Clem Brooks; Paul Nieuwbeerta; Jeff Manza

doi:10.1016/j.ssresearch.2004.06.005

Cleavage-based voting behavior in cross-national perspective: Evidence from six postwar democracies

Clem Brooks, Paul Nieuwbeerta, Jeff Manza

Research output: Contribution to journal › Article › peer-review

Abstract

We investigate trends and cross-national variation in the impact of class, religious, and gender cleavages on voting behavior in six advanced capitalist democracies in the postwar period. Earlier research on cleavage voting has been criticized for utilizing outdated "two-class" models of class structure, simplistic left/right party distinctions, flawed statistical approaches, and incompatible and/ or limited of cross-national empirical evidence. We take such criticisms seriously and seek to overcome them. Using multinomial logistic regression models, we analyze data from a new dataset, the International Social Cleavages and Politics (ISCP) file, which contains comparable, over-time cross-national data from 112 nationally representative election surveys of voters in six Western democracies in the period 1964-1998. The six countries examined in the paper (Australia, Austria, Germany, Great Britain, the Netherlands, and the United States) are unique because of the existence of time series data available for all three cleavages. Our analyses examine the changing magnitude of the class, religion, and gender cleavages for up to five distinct party families for each country. Unskilled workers have become less distinctive in their partisan alignments over time, but other classes have experienced offsetting changes, yielding little evidence of a universal decline in the class cleavage. Further analyses suggest an important degree of stability in the aggregated effects of all social cleavages, while also revealing significant cross-national differences and trends in the magnitude of specific cleavages. These results refine debates concerning the possible decline of social cleavages; implications for research are discussed in conclusion.

Original language	English (US)
Pages (from-to)	88-128
Number of pages	41
Journal	Social Science Research
Volume	35
Issue number	1
DOIs	https://doi.org/10.1016/j.ssresearch.2004.06.005
State	Published - Mar 2006

ASJC Scopus subject areas

Education
Sociology and Political Science

Access to Document

10.1016/j.ssresearch.2004.06.005

Cite this

@article{2536898d78fa40858cce8b1b03a06814,

title = "Cleavage-based voting behavior in cross-national perspective: Evidence from six postwar democracies",

abstract = "We investigate trends and cross-national variation in the impact of class, religious, and gender cleavages on voting behavior in six advanced capitalist democracies in the postwar period. Earlier research on cleavage voting has been criticized for utilizing outdated {"}two-class{"} models of class structure, simplistic left/right party distinctions, flawed statistical approaches, and incompatible and/ or limited of cross-national empirical evidence. We take such criticisms seriously and seek to overcome them. Using multinomial logistic regression models, we analyze data from a new dataset, the International Social Cleavages and Politics (ISCP) file, which contains comparable, over-time cross-national data from 112 nationally representative election surveys of voters in six Western democracies in the period 1964-1998. The six countries examined in the paper (Australia, Austria, Germany, Great Britain, the Netherlands, and the United States) are unique because of the existence of time series data available for all three cleavages. Our analyses examine the changing magnitude of the class, religion, and gender cleavages for up to five distinct party families for each country. Unskilled workers have become less distinctive in their partisan alignments over time, but other classes have experienced offsetting changes, yielding little evidence of a universal decline in the class cleavage. Further analyses suggest an important degree of stability in the aggregated effects of all social cleavages, while also revealing significant cross-national differences and trends in the magnitude of specific cleavages. These results refine debates concerning the possible decline of social cleavages; implications for research are discussed in conclusion.",

author = "Clem Brooks and Paul Nieuwbeerta and Jeff Manza",

note = "Funding Information: Comparative and historical analysis visibly broadens scholarly debates, but past research on the impact of social cleavages has been constrained by either significant methodological limitations or a tendency to limit hypothesis testing to single countries. Virtually no previous studies have attempted to test hypotheses about social cleavage impacts in comparative context using a fully differentiated party family scheme and a multi-category typology of cleavages. The relatively broad scope of our analyses, covering six advanced capitalist democracies for a 30 year period, provides a new perspective from which to address these questions. While future investigations will be needed to refine the understandings developed here, we have shown the possibilities for systematic comparative-historical investigation, and we welcome further work that spans additional countries and time periods as a means of advancing research on cleavage voting in capitalist democracies. It is critical to emphasize, in this context, the relevance of the time period covered by our analyses to debates and unresolved controversies over trends in social cleavage-based voting within capitalist democracies. In particular, the theoretical underpinnings of most predictions concerning a decline or displacement of cleavage voting identify historical factors and novel processes that are hypothesized as emerging primarily during the historical era since the 1960s. These would include the following: new left-libertarian (and neo-conservative or new right) movements; novel (or re-emergent) ideological conflicts concerning issues of rights, the environment, and the character of civil society; recent patterns of postindustrial change; partisan volatility affecting left and other previously-dominant parties during the 1980s; and perhaps more recently, the maturation of globalization and transnational processes (e.g., Franklin et al., 1992; Clark et al., 1993; Dalton and Wattenberg, 1993; Pakulski and Waters, 1996 ). Accordingly, while it could thus be hypothesized that social cleavage voting within specific countries was possibly higher prior to the decade of the 1960s, if true, that line of argumentation might also call into question much of the theoretical basis for anticipating contemporary patterns of cleavage decline. Further research, then, situating the post-1960s era within a longer time frame, while employing suitable measures and methods for the study of social cleavages, may thus help to refine scholarly understanding of the historical origins and magnitude of trends in cleavage-based voting. The substantive findings of this study help to advance recent debates concerning the continuing significance vs. declining relevance of social cleavages to voter alignments and political conflict within capitalist democracies (see Evans (2000) for a recent review). Widely proclaimed assertions regarding the existence of working-class dealignment and its significant consequences for declining cleavage voting within Western democracies are partially supported by these analyses. More specifically, in four of the six countries we have investigated, unskilled workers, traditionally the bulwark of Left party support, have evolved over time to more closely resemble the average voter, making their partisan political alignments far less distinctive, and potentially providing far less incentive for left politicians to develop policies designed to appeal (disproportionately) to unskilled workers. By the same token, however, we find no evidence for a universal decline in the partisan relevance of social cleavages, including class in the post-1960s era. While some countries in our analyses conform to the declining cleavage scenario (Britain, Germany, and to a somewhat lesser extent, the Netherlands), others show some tendency toward a growing magnitude of social cleavages as a whole (Australia and the US), and one country (Austria) has experienced a pattern of stability coupled with a very recent increase. This larger pattern is visible in part because we have simultaneously examined trends among three major social cleavages. This enables various refinements to the overall picture of social cleavages in democracies. For example, we find that cleavage decline in Britain and Germany is a product of decreases in the magnitude of a single cleavage (class); gender and religion-based differences in voter alignments have experienced little change within these two countries. As suggested in the introduction to this paper, the existing scholarly literatures on social cleavages and political behavior is characterized by a sharply polarized debate between scholars who consistently assert decline vs. those who categorically reject such a scenario. Taken as a whole, however, our results suggest an important limitation of these polarized interpretations. Indeed, the clear association between country context and the extent and type of social cleavage change establishes a pattern of cross-national complexity with respect to which past debates over cleavage decline appear ill-suited to understand. Following an emerging strain of commentary on political-sociological research on voting behavior ( Evans, 1999; Evans and Whitefield, 1999; Weakliem and Heath, 1999 ), our results suggest the utility of more detailed and systematic comparisons between countries as a means of advancing debates. In addition to developing such comparisons, further research may be able to shed productive light by considering the potential relevance of such factors as nation-specific electoral systems ( Powell, 2000 ), welfare state type (e.g., Esping-Andersen, 1990; Korpi and Palme, 1998 ; O{\textquoteright}Connor et al., 1999 ), other long-standing sources of party-based political cleavages (e.g., Przeworski and Sprague, 1986; Kitschelt, 1994 ), and public opinion (e.g., Page and Shapiro, 1992; Powell, 2000 ). Cross-national differences in trends affecting social cleavages also cast significant doubt on the other interpretive pole of recent debates, namely, that the institutionalization of social group-based conflict within capitalist societies renders them impervious to any process of change affecting voter alignments. While the class cleavage has not declined universally, four of the six countries investigated in this study show clear evidence of working-class dealignment, and the very large decline of class voting is inconsistent with expectations of persistence within all Western democracies. Generic expectations of stability also provide little basis for understanding countries such as Australia and the US in which the magnitude of specific social cleavages has increased. Our results also suggest a degree of similarity in the average size of overall social cleavages within the six democracies we have investigated. Indeed, we find that country-specific patterns of cleavage change have lead to growing comparability in the magnitude of the overall social cleavage: By the mid-1990s, the average difference between two social groups in the probability of favoring a given party family was approximately .10. Although such estimates represent a non-trivial difference in political behavior (and some pairwise contrasts produce much larger differences), we would caution against over-estimating their magnitude. The quite large deviance remaining after estimating our preferred models suggests the further importance of other sources of the vote. Given these results, further research and debate, including with respect to related debates concerning political change ( Brooks et al., 2003 ), may profit by broadening its scope to consider other factors behind the partisan alignments of voters. Appendix A List of original datasets Study Type of study No. of cases Archive Archive No. AUS65 Other 1925 SSDA M168 AUS67 Other 2054 ICPSR 07282 AUS73 Other 4939 SSDA M169 AUS79 Other 2016 SSDA M170 AUS84E NES 3012 SSDA M171 AUS85I ISSP 1528 ZA 1490 AUS86I ISSP 1250 ZA 1620 AUS87E NES 1825 SSDA 445 AUS87I ISSP 1663 ZA 1680 AUS90E NES 2037 SSDA 570 AUS92I ISSP 2203 ZA 2310 AUS93E NES 3023 SSDA 763 AUS96E NES 1797 SSDA 943 AUS98E NES 1897 SSDA 1001 AUT74P Pol.Act. 1585 ICPSR 7777 AUT85I ISSP 987 ZA 1490 AUT88I ISSP 972 ZA 1700 AUT89I ISSP 1997 ZA 1840 AUT91I ISSP 984 ZA 2150 AUT94I ISSP 977 ZA 2620 AUT95I ISSP 1007 ZA 2880 ENG64E NES 1769 ICPSR 7250 ENG66E NES 1874 ICPSR 7250 ENG70E NES 1355 ICPSR 7004 ENG740 NES 2365 UK-DA 0666 ENG79E NES 1893 UK-DA 1533 ENG83E NES 3955 UK-DA 2005 ENG85I ISSP 1530 ZA 1490 ENG86I ISSP 1416 ZA 1620 ENG87E NES 3826 UK-DA 2568 ENG87I ISSP 1212 ZA 1680 ENG88I ISSP 2614 ZA 1700 ENG89I ISSP 2594 ZA 1840 ENG90I ISSP 1197 ZA 1950 ENG91I ISSP 1257 ZA 2150 ENG92I ISSP 1066 ZA 2310 ENG92E NES 3534 DA 2981 ENG93I ISSP 1261 ZA 2450 ENG94I ISSP 984 ZA 2620 ENG95I ISSP 1058 ZA 2880 ENG97I ISSP 1087 ZA 3090 GER69E NES 939 ZA 0525 GER69F NES 1158 ZA 7098 GER75P Pol.Act. 2307 ICPSR 07777 GER76Z ZUBUS 2036 ZA 1233 GER77Z ZUBUS 2002 ZA 1233 GER78C ZUBUS 2030 ZA 1233 GER78X ZUBUS 2012 ZA 1233 GER79X ZUBUS 2007 ZA 1233 GER79Z ZUBUS 2012 ZA 1233 GER80A ZUBUS 2955 ZA 1795 GER80C ZUBUS 1939 ZA 1233 GER80P Pol.Act. 2095 ZA 1188 GER80Z ZUBUS 1997 ZA 1233 GER82A GSS 2991 ZA 1795 GER84A GSS 3004 ZA 1795 GER86A GSS 3095 ZA 1795 GER88A GSS 3052 ZA 1795 GER90A GSS 3051 ZA 1800 GER91A GSS 3058 ZA 1990 GER92A GSS 3548 ZA 2140 GER94A GSS 3450 ZA 2400 GER96A GSS 2800 ZA 2800 GER98A GSS 3234 ZA 3000 NET70 NES 1838 STEIN P1036 NET71 NES 906 ICPSR 7768 NET72E NES 3175 STEIN P0353 NET74P Pol.Act 1201 ICPSR 7777 NET76 Other 755 STEIN P0653 NET77E NES 1856 STEIN P0354 NET77L Other 4159 STEIN P0328 NET79P Pol.Act 806 ZA 1188 NET81E NES 2305 STEIN P0350 NET82E NES 1541 STEIN P0633 NET85S Other 3003 STEIN P1012 NET86E NES 1630 STEIN P0866 NET87 Other 795 STEIN P1194 NET89E NES 1745 STEIN P1000 NET89M Other 956 STEIN P1098 NET90S Other 2384 STEIN P1100 NET94E NES 1812 STEIN P1208 NET96 Other 790 STEIN P1370 NET97I ISSP 2267 ZA 3090 NET98E NES 2101 STEIN P1415 USA64E NES 1834 ICPSR 07235 USA66E NES 1291 ICPSR 07259 USA68E NES 1673 ICPSR 07281 USA70E NES 1694 ICPSR 07298 USA72G GSS 1613 ICPSR 09505 USA73G GSS 1504 ICPSR 09505 USA74G GSS 1484 ICPSR 09505 USA74P Pol.Act 1719 ICPSR 07777 USA75G GSS 1490 ICPSR 09505 USA76G GSS 1499 ICPSR 09505 USA77G GSS 1530 ICPSR 09505 USA78G GSS 1532 ICPSR 09505 USA80G GSS 1468 ICPSR 09505 USA82G GSS 1860 ICPSR 09505 USA83G GSS 1599 ICPSR 09505 USA84G GSS 1473 ICPSR 09505 USA85G GSS 1534 ICPSR 09505 USA86G GSS 1470 ICPSR 09505 USA87G GSS 1819 ICPSR 09505 USA88G GSS 1481 ICPSR 09505 USA89G GSS 1537 ICPSR 09505 USA90G GSS 1372 ICPSR 09505 USA91G GSS 1517 ICPSR 2685 USA93G GSS 1606 ICPSR 2685 USA94G GSS 2992 ICPSR 2685 USA96G GSS 2904 ICPSR 2685 USA98G GSS 2832 ICPSR 2685 Data references ICPSR:7004 Butler, D., & D.E. Stokes, Political Change in Britain, 1969–1970 ICPSR:7214 Campbell, A., P. Converse et al., American National Election Study, 1956 ICPSR:7215 Campbell, A., P. Converse et al., American National Election Study, 1958 ICPSR:7216 Campbell, A., P. Converse et al., American National Election Study, 1960 ICPSR:7235 SRC (Survey Research Center), American National Election Study, 1964 ICPSR:7250 Butler, D., & D.E. Stokes, Political Change in Britain, 1963–1970 ICPSR:7259 SRC (Survey Research Center), American National Election Study, 1966 ICPSR:7281 SRC (Survey Research Center), American National Election Study, 1968 ICPSR:7282 Aitkin, D., M. Kahan & D.E. Stokes, Australian National Political Attitudes, 1967 ICPSR:7298 CPS Center for Political Studies, American National Election Study, 1970 ICPSR:7768 Verba, S., N.H. Nie & J.-O. Kim, Political Participation and Equality in Seven Nations, 1966–1971 ICPSR:7777 Barnes, S.H., & M. Kaase et al., Political Action: An Eight Nation Study, 1973–1976 ICPSR:9505 Davis, J.A. & T.W. Smith, General Social Survey Cumulative File, 1972–1991 ICPSR:xxxx Davis, J.A. & T.W. Smith, General Social Survey Cumulative File, 1972–1998 NSD:NOR77e Names of principal investigators: Valen, H., B. Aardal, Institute of social research, Oslo Norwegian Election Study, 1977 [computer file]. Data collector: Statistics Norway, Oslo SSDA:M168 Broom, L., F.L. Jones & J. Zubrzycki, Social Stratification in Australia, 1965 SSDA:M169 Broom, L., P. Duncan-Jones, F.L. Jones, P. McDonnel & T. Williams, Social Mobility in Australia Project, 1973 SSDA:M170 Aitkin, D., MacQuarie University Australian Political Attitudes Survey, 1979 SSDA:M171 Kelley, J.L., R.G. Cushing & B. Headey, Australian National Social Science Survey, 1984 SSDA:0445 McAllister, I. & A. Mughan, Australian Election Survey, 1987 SSDA:0570 McAllister, I., R. Jones, E. Papodalis & D. Gow, Australian Election Survey, 1990 SSDA: 0763 Roger Jones, Ian McAllister, David Denemark, David Gow, Australian Election Study, 1993 SSDA: 0943 Roger Jones, Ian McAllister, David Gow, Australian Election Study, 1996 SSDA: 1001 Clive Bean, David Gow, Ian McAllister, Australian Election Study, 1998 STEIN:P0136 Heunks, F. M., M.K. Jennings, W.E. Miller, P.C. Stouthard & J. Thomassen, Dutch Election Study, 1970–1973 STEIN:P0328 Centraal Bureau voor Statistiek (CBS), Life Situation Survey, Netherlands 1977 STEIN:P0350 Werkgroep Nationaal Kiezersonderzoek, Dutch Parliamentary Election Study, 1981 STEIN:P0353 Werkgroep Nationaal Kiezersonderzoek, Dutch Parliamentary Election Study, 1972 STEIN:P0354 Werkgroep Nationaal Kiezersonderzoek, Dutch Parliamentary Election Study, 1977 STEIN:P0633 Werkgroep Nationaal Kiezersonderzoek, Dutch Parliamentary Election Study, 1982 STEIN:P0653 Hermkens, P.L.J. & P.J. van Wijngaarden, Criteria for Justification of Income Differences, Netherlands 1976 STEIN:P0866 Van der Eijk, C., G.A. Irwin & B. Niemoeller, Dutch Parliamentary Elections Study, 1986 STEIN:P1012 Felling, A.J.A., J. Peters & O. Schreuder, Social Relevance of Religion in the Netherlands, 1985 STEIN:P1098 Arts, K., E. Hollander, K. Renckstorf & P. Verschuren, Media-equipment, Media-exposure and Media-use in the Netherlands, 1989 STEIN:P1100 Felling, A., J. Peters, & O. Schreuder, Social Relevance of Religion in the Netherlands, 1990 STEIN:P1194 Hermkens, P.L.J. & P.J. van Wijngaarden, Criteria for Justification of Income Differences, Netherlands 1987 STEIN:P1415 Aarts, K., H. van der Kolk, M. Kamp, J. Thomassen, Dutch Parliamentary Election Study, 1998 STEIN:P1208 Stichting Kiezersonderzoek Nederland, Dutch Parliamentary Election Study, 1994 UK-DA:0666 Crewe, I., B. Saerlvik & J. Alt, British Election Study: October 1974, Cross-section UK-DA:1533 Crewe, I., B. Saerlvik & D. Robertson, British Election Study: May 1979, Cross-section UK-DA:2005 Heath, A.F., R.M. Jowell, J.K. Curtice & E.J. Field, British Election Study: June 1983 UK-DA:2568 Heath, A.F., et al., British Election Study: 1987. ZA:0426 Klingemann, H.-D., & F.U. Pappi, German Pre- and Post-Election Study, 1969 ZA:0525 Kaase, M., U. Schleth, W. Adrian, M. Berger & R. Wildenmann, German Election Study: August–September 1969 ZA:1188 Allerbeck, K.R., M. Kaase, H.-D. Klingemann, Ph.C. Stouthard, F.J. Heunks, J.J.A. Thomassen, J.W. van Deth, S.H. Barnes, B.G. Farah, R. Inglehart, and M.K. Jennings, Political Action II, 1979–1980 ZA:1233 ZUMA Zentrum fuer Umfragen, Methoden, und Analysen, ZUMA-Standard Demographie (Zeitreihe), Germany 1976–1980 ZA:1490 International Social Survey Programme (ISSP), International Social Survey Programme: Role of Government, 1985–1986 ZA:1620 International Social Survey Programme (ISSP), International Social Survey Programme: Social Networks and Support Systems, 1986 ZA:1680 International Social Survey Programme (ISSP), International Social Survey Programme: Social Inequality, 1987 ZA:1700 International Social Survey Programme (ISSP), International Social Survey Programme: Family and Changing Sex Roles, 1988 ZA:1795 Allerbeck, K.R., M.R. Lepsius, K.U. Mayer, W. M{\"u}ller, K.-D. Opp, F.U. Pappi, E.K. Scheuch, and R. Ziegler, Allgemeine Bevolkerungsumfrage der Sozialwissenschaften Allbus Kumulierter Datensatz, 1980–1988 ZA:1800 ZUMA Zentrum fuer Umfragen, Methoden, und Analysen, German Social Survey (Allbus), 1990 ZA:1840 International Social Survey Programme (ISSP), International Social Survey Programme: Work Orientations, 1989 ZA:1950 International Social Survey Programme (ISSP), International Social Survey Programme: Role of Government II, 1990 ZA:2150 International Social Survey Programme (ISSP), International Social Survey Programme: Religion, 1991 ZA:2450 International Social Survey Programme (ISSP), International Social Survey Programme: Environment, 1993 ZA:2620 International Social Survey Programme (ISSP), International Social Survey Programme: Family and Changing Gender Roles II, 1994. ZA:2880 International Social Survey Programme (ISSP), International Social Survey Programme: National Identity, 1995 ZA:3090 International Social Survey Programme (ISSP), International Social Survey Programme: Work Orientations II, 1997 ZA:3000 ZUMA Zentrum fuer Umfragen, Methoden, und Analysen, German Social Survey (Allbus), 1990–1998 Appendix B Party families and political parties by country Party family Party name Australia a Left Communist Party, Australian Labor Party, Queensland Labor Party, Lang Labor Party Liberal Australia Party, Liberal Movement, National Alliance, Service Party of Australia, Australian Democrats, Country-Liberal Party Other The Greens, a Tasmanian Greens, a Western Australia Greens, a Queensland Greens, a Victorian Greens a National Party Call to Australia Party, Democratic Labor Party, Australian Liberal Party, Other Parties Austria b Left Communist Party Socialist Green United Greens of Austria, Alternative List, Green Alternative Liberal Liberal Forum Religious Austrian Peoples Party Other b Association of Members of the Social Security System, No-Citizens{\textquoteright} Initiative against the Sale of Austria, Czechs, National Socialist, German Workers{\textquoteright} Party, Hitler movement, Fatherland Front, Democratic Progressive Party, Freedom Party, Other Parties Germany Left German Communist Party, German Peace Union, Action for Democratic Progress Party of Democratic Socialism, Social Democrats, All-German People{\textquoteright}s Party Green Greens, Ecological Democratic Party, Alliance 90-Greens Liberal Free Democrats Religious Christian Democratic Union, Christian Social Union Other Centre Party, All-German Party, German Social Union, The Greys, Bavarian Party, South Schleswig Voters{\textquoteright} League, Refugee Party, Federal Union Economic Reconstruction League, German Reich Party, National Democratic Party, Republicans, German Party, Other Parties The Netherlands c Left Communist Party, Netherlands Labour Party, Democratic Socialists {\textquoteright}70, Pacifist Socialist Party, Socialist Party, Green Left, c Greens c Liberal People{\textquoteright}s Party for Freedom and Democracy, Democrats 66 Religious Anti-Revolutionary Party, Catholic People{\textquoteright}s Party, Christian Historical Union, Political Reformed Party, Catholic National Party, Reformed Political Union, Radical Political Party, Roman Catholic Party, Christian Democratic Appeal, Reformed Political Federation, Evangelical People{\textquoteright}s Party Other Centre Party, Centre Democrats, People{\textquoteright}s Party of the Right, Middle Class Party, United Old Persons{\textquoteright} League, 55 + Union, Other Parties United Kingdom c Left Communist Party, Independent Labour Party, Labour Party, Social Democratic and Labour Party, Green Party c Liberal Liberal Party, Liberal Democrat Party Conservative Conservative Party, National Liberal Party Other National Front, Sinn F{\'e}in, Ulster Unionist Party, Democratic Unionist Party, Alliance Party of Northern Ireland, Plaid Cymru (Party of Wales), Scottish National Party, Other Party United States Liberal Democratic Party Conservative Republican Party Other Communist Party d Socialist Labor Party d Socialist Party d Progressive Party d Progressive Party (H. Wallace), d American Party, Independent (J. Anderson), Independent (Ross Perot), Libertarian Party, Other Party a The very small number of Green party family supporters in Australia are combined with Other. b The very small number of Other party family supporters in Austria are treated as “non-voters” and deleted from the analyses. c In the Netherlands and Great Britain the very small number of Green party family supporters are combined with Left party. d In the US, third and independent candidates are combined in the Other category. Appendix C Multinomial logistic regression coefficients = 17,944) a (SE in parentheses) for preferred model of social cleavages in Australia ( N Independent variables Other vs. Left party family Liberal vs. Left party family Constant −.54 ∗ (.20) 2.45 ∗ (.15) = 1979) Year (reference 1984 1.48 ∗ (.24) .43 ∗ (.19) 1985 .66 ∗ (.26) −.14 (.20) 1986 2.55 ∗ (.26) .17 (.22) 1987 1.52 ∗ (.01) −.52 ∗ (.16) 1990 −.33 (.22) −.1.49 ∗ (.16) 1992 2.33 ∗ (.22) −.33 (.18) 1993 2.80 ∗ (.22) .01 (.18) 1995 2.21 ∗ (.21) −.95 ∗ (.17) 1996 2.36 ∗ (.20) −.89 ∗ (.18) 1998 2.33 ∗ (.22) −.72 ∗ (.18) = no religion) Religious group (reference Catholic .52 ∗ (.09) .4 8 ∗ (.09) Protestant .76 ∗ (.08) .21 ∗ (.08) = female) Gender (reference Male .30 ∗ (.07) .40 ∗ (.07) = non-labor force) Class category (reference Un/semi-skilled manual workers −.44 ∗ (.10) −.06 (.14) Skilled manual workers −.33 ∗ (.15) .01 (.14) Routine non-manual workers −.12 ∗ (.10) −.16 ∗ (.10) Service class −.4 8 ∗ (.08) −.64 ∗ (.08) Self-employed and farmers 2.00 ∗ (.46) −.44 ∗ (.17) Interactions Self-employed × year −.06 ∗ (.02) ___ b a An asterisk next to a coefficient denotes significance at the .05 level (2-tailed test). b Coefficient estimate constrained to 0. Appendix D Multinomial logistic regression coefficients = 5754) a (SE in parentheses) for preferred model of social cleavages in Austria ( N Independent variables Left vs. Religious party family Green vs. Religious party family Liberal vs. Religious party family Constant −2.43 ∗ (.18) −23.14 ∗ (.41) −.63 ∗ (.25) = 1974) Year (reference 1985 −.34 ∗ (.10) 21.23 ∗ (.41) −.26 (.24) 1988 −.21 (.12) 22.64 ∗ (.40) .02 (.26) 1989 −.06 (.09) 23.23 ∗ (.36) 1.03 ∗ (.18) 1991 1.29 ∗ (.17) 23.03 ∗ (.38) .86 ∗ (.20) 1994 .00 (.1 1) 23.16 ∗ (.38) .48 ∗ (.22) 1995 −.06 (.12) 23.97 ∗ (.37) 1.92 ∗ (.18) = no religion) Religious group (reference Catholic −2.11 ∗ (.16) −2.04 ∗ (.22) −.1.88 ∗ (.20) Protestant −.1.43 ∗ (.22) −.1.12 ∗ (.32) −.50 ∗ (.27) = female) Gender (reference Male .02 (.07) .05 (.14) .32 ∗ (.1 1) = non-labor force) Class category (reference Un/semi-skilled manual workers .44 ∗ (.1 1) −.25 (.32) .10 (.20) Skilled manual workers .61 ∗ (.12) .03 (.30) .61 ∗ (.18) Routine non-manual workers .08 (.10) .83 ∗ (.19) .27 (.17) Service class −.32 ∗ (.10) .97 ∗ (.17) −.02 (.16) Self-employed and farmers −.1.82 ∗ (.13) −.1.13 ∗ (.33) −.30 (.17) a An asterisk next to a coefficient denotes significance at the .05 level (2-tailed test). Appendix E Multinomial logistic regression coefficients = 27,758) a (SE in parentheses) for preferred model of social cleavages in Britain ( N Independent variables Cons. vs. Left party family Other vs. Left party family Liberal vs. Left party family Constant −.70 ∗ (.07) −4.29 ∗ (.35) −2.17 ∗ (.11) = 1966) Year (reference 1970 .05 (.08) .67 (.43) −.10 (.15) 1974 .22 ∗ (.08) 1.92 ∗ (.36) 1.12 ∗ (.10) 1979 .67 ∗ (.08) 1.26 ∗ (.39) 1.01 ∗ (.13) 1983 .67 ∗ (.07) 1.35 ∗ (.37) 1.57 ∗ (.1 1) 1985 .04 (.09) 1.02 ∗ (.40) .63 ∗ (.13) 1986 .19 ∗ (.09) 1.66 ∗ (.38) .66 ∗ (.13) 1987 .61 ∗ (.07) 1.30 ∗ (.36) 1.53 ∗ (.1 1) 1988 .33 ∗ (.08) 2.12 ∗ (.35) .20 (.13) 1989 .52 ∗ (.08) 3.18 ∗ (.34) .60 ∗ (.12) 1990 .13 ∗ (.10) 2.77 ∗ (.36) .04 (.16) 1991 .41 ∗ (.09) 2.12 ∗ (.37) 1.00 ∗ (.14) 1992 .48 ∗ (.10) 2.42 ∗ (.37) 1.11 ∗ (.14) 1993 .29 ∗ (.09) 2.15 ∗ (.37) .86 ∗ (.14) 1994 −.08 (.10) 1.94 ∗ (.38) .82 ∗ (.14) 1995 −.19 (.10) 1.74 ∗ (.38) .84 ∗ (.14) 1997 −.07 (.10) 1.97 ∗ (.38) .57 ∗ (.15) = no religion) Religious group (reference Catholic −.22 ∗ (.05) −.78 ∗ (.13) −.35 ∗ (.07) Protestant .57 ∗ (.03) −.16 ∗ (.07) .35 ∗ (.04) = female) Gender (reference Male −.04 (.03) −.22 ∗ (.07) −.14 ∗ (.04) = non-labor force) Class category (reference Un/semi-skilled manual workers −.85 ∗ (.12) −.01 (.12) −.15 ∗ (.07) Skilled manual workers −.62 ∗ (.05) −.11 (.11) −.34 ∗ (.06) Routine non-manual workers .18 (.11) .36 ∗ (.11) .49 ∗ (.06) Service class .70 ∗ (.04) .43 ∗ (.10) .81 ∗ (.05) Self-employed and farmers 1.52 ∗ (.18) .61 ∗ (.18) .53 ∗ (.11) Interactions Un/semi-skilled manual worker × year .02 ∗ (.01) __ b __ b Routine non-manual workers × year .01 ∗ (<.01) __ b __ b Self-employed × year −.02 ∗ (.01) __ b __ b a An asterisk next to a coefficient denotes significance at the .05 level (2-tailed test). b Coefficient estimate constrained to 0. Appendix F Multinomial logistic regression coefficients = 39,874) a (SE in parentheses) for preferred model of social cleavages in Germany ( N Independent variables Left vs. Religious party family Green vs. Religious party family Liberal vs. Religious party family Constant .92 ∗ (.07) −24.83 ∗ (.15) −2.43 ∗ (.22) −1.73 ∗ (.14) = 1969) Year (reference 1975 −.09 (.08) −16.12 ∗ (<.01) −1.37 ∗ (.38) .99 ∗ (.15) 1976 −.03 (.08) −16.05 ∗ (<.01) −.22 (.28) 1.26 ∗ (.15) 1977 −.06 (.08) −16.12 ∗ (<.01) −.84 ∗ (.34) .82 ∗ (.16) 1978 .00 (.07) −16.09 ∗ (<.01) .75 ∗ (.21) 1.00 ∗ (.15) 1979 −.07 (.07) −16.16 ∗ (<.01) −1.19 ∗ (.31) .69 ∗ (.15) 1980 .04 (.06) 23.51 ∗ (.16) −1.43 ∗ (.27) 1.17 ∗ (.14) 1982 −.62 ∗ (.07) 24.18 ∗ (.16) −1.57 ∗ (.37) .90 ∗ (.15) 1984 −.19 ∗ (.08) 24.41 ∗ (.16) −1.80 ∗ (.42) .16 (.17) 1986 .00 (.08) 24.49 ∗ (.16) −1.38 ∗ (.38) .72 ∗ (.16) 1987 −.06 (.10) 24.19 ∗ (.16) −.52 (.39) .70 ∗ (.19) 1988 .13 (.08) 24.35 ∗ (.17) −.14 (.27) .58 ∗ (.17) 1990 .09 (.08) 24.42 ∗ (.16) .33 (.24) .96 ∗ (.15) 1991 −.32 ∗ (.08) 23.80 ∗ (.17) −.23 (.26) 1.03 ∗ (.15) 1992 −.19 ∗ (.08) 24.20 ∗ (.17) .21 (.23) .81 ∗ (.15) 1994 .14 (.08) 24.65 ∗ (.16) 1.05 ∗ (.22) .89 ∗ (.16) 1996 −.26 ∗ (.08) 24.22 ∗ (.16) −.03 (.24) .46 ∗ (.16) 1998 .15 (.08) 24.00 ∗ (.17) .60 ∗ (.23) −.01 (.17) = no religion) Religious group (reference Catholic −1.39 ∗ (.04) −1.56 ∗ (.07) −1.20 ∗ (.11) −1.31 ∗ (.05) Protestant −.52 ∗ (.04) −.87 ∗ (.06) −.84 ∗ (.10) −.35 ∗ (.06) = female) Gender (reference Male −.03 (.03) −.19 ∗ (.08) .43 ∗ (.08) −.04 (.04) = non-labor force) Class category (reference Un/semi-skilled manual workers .88 ∗ (.09) .31 ∗ (.08) .50 ∗ (.11) .18 ∗ (.07) Skilled manual workers .85 ∗ (.12) .03 (.11) .45 ∗ (.15) .07 (.08) Routine non-manual workers .06 (.04) .34 ∗ (.08) .26 (.15) .37 ∗ (.06) Service Class .04 (.03) −.05 (.24) −.20 (.12) .68 ∗ (.05) Self-employed and farmers −1.31 ∗ (.07) −.59 ∗ (.14) −.20 (.19) .12 (.08) Interactions Un/semi-skilled manual worker × year −.02 < .01) ∗ ( __ b __ b __b Skilled manual worker × year −.03 ∗ (.01) __ b __ b __ b Service class × year __ b .03 ∗ (.01) __ b __ b a An asterisk next to a coefficient denotes significance at the .05 level (2-tailed test). b Coefficient estimate constrained to 0. Appendix G Multinomial logistic regression coefficients = 21,008) a (SE in parentheses) for preferred model of social cleavages in Netherlands ( N Independent variables Other vs. Left party family Liberal vs. Left party family Liberal vs. Religious party family Constant 1.50 ∗ (.10) −2.87 ∗ (.35) 1.11 ∗ (.10) Year (reference =1970) 1971 −.17 (.13) .76 (.42) −.43 ∗ (.13) 1972 .08 (.09) 1.31 ∗ (.34) −.33 ∗ (.10) 1974 .55 ∗ (.11) 1.57 ∗ (.38) −.21 (.12) 1977 .30 ∗ (.08) .87 ∗ (.35) .08 (.09) 1979 .31 ∗ (.14) .08 (.66) .12 (.13) 1981 .14 (.13) .33 (.53) .13 (.13) 1982 .36 ∗ (.13) .32 (.56) .18 (.13) 1985 .28 ∗ (.11) 1.99 ∗ (.36) −.06 (.11) 1986 .06 (.13) 1.02 ∗ (.44) −.45 ∗ (.14) 1989 −.10 (.15) 1.04 ∗ (.45) −.59 ∗ (.15) 1990 −.64 ∗ (.13) 2.59 ∗ (.36) −.33 ∗ (.12) 1994 .19 (.16) 3.73 ∗ (.37) .58 ∗ (.15) 1996 .04 (.18) 3.51 ∗ (.39) .57 ∗ (.17) 1998 .45 ∗ (.16) 3.58 ∗ (.38) .34 ∗ (.16) = no religion) Religious group (reference Catholic −3.39 ∗ (.12) −1.50 ∗ (.29) −2.55 ∗ (.13) Protestant −2.51 ∗ (.12) −1.36 ∗ (.30) −2.06 ∗ (.13) = female) Gender (reference Male .26 ∗ (.08) −.29 ∗ (.08) −.05 (.04) = non-labor force) Class category (reference Un/semi-skilled manual workers .32 ∗ (.07) .32 ∗ (.16) −.54 ∗ (.15) Skilled manual workers .46 ∗ (.07) .66 ∗ (.15) −.65 ∗ (.15) Routine non-manual workers .05 (.06) .17 (.13) .39 ∗ (.18) Service class −.19 ∗ (.06) .31 ∗ (.10) .52 ∗ (.05) Self-employed and farmers −1.22 ∗ (.11) 1.83 ∗ (.45) .29 ∗ (.08) Interactions Un/semi-skilled manual worker × year __ b __ b .03 ∗ (<.01) Skilled manual workers × year __ b __ b .03 ∗ (<.01) Self-employed × year __ b −.12 ∗ (.03) __ b Male × year −.01 ∗ (<.01) __ b __ b Catholic × year .04 ∗ (<.01) −.03 ∗ (.01) .02 ∗ (<.01) Protestant × year <−.01 (<.01) −.04 ∗ (.01) −<.01 (<.01) a An asterisk next to a coefficient denotes significance at the .05 level (2-tailed test). b Coefficient estimate constrained to 0. Appendix H Multinomial logistic regression coefficients = 39,789) a (SE in parentheses) for preferred model of social cleavages in the US ( N Independent variables Liberal vs. Cons. party family Other vs. Cons. party family Constant 1.63 ∗ (.07) −4.81 ∗ (.71) = 1964) Year (reference 1966 −.28 ∗ (.11) −38.50 (>.01) 1968 −1.00 ∗ (.08) 3.97 ∗ (.71) 1970 −.70 ∗ (.20) 2.01 ∗ (.74) 1972 −1.32 ∗ (.08) 2.01 ∗ (.73) 1973 −1.49 ∗ (.08) 2.57 ∗ (.73) 1974 −1.49 ∗ (.07) 2.85 ∗ (.72) 1975 −1.34 ∗ (.08) 2.78 ∗ (.72) 1976 −1.29 ∗ (.08) 3.22 ∗ (.72) 1977 −.59 ∗ (.08) 1.92 ∗ (.72) 1978 −.73 ∗ (.08) 2.55 ∗ (.73) 1980 −.64 ∗ (.08) 2.53 ∗ (.73) 1982 −.64 ∗ (.08) 2.47 ∗ (.73) 1983 −.92 ∗ (.08) 2.00 ∗ (.74) 1984 −.94 ∗ (.08) 2.12 ∗ (.73) 1985 −1.37 ∗ (.08) 1.96 ∗ (.73) 1986 −1.34 ∗ (.09) 2.12 ∗ (.73) 1987 −.90 ∗ (.08) 2.29 ∗ (.73) 1988 −1.40 ∗ (.09) 2.45 ∗ (.73) 1989 −1.13 ∗ (.09) 2.31 ∗ (.72) 1990 −1.49 ∗ (.09) 2.10 ∗ (.73) 1991 −1.43 ∗ (.09) 2.10 ∗ (.73) 1993 −.61 ∗ (.09) 4.67 ∗ (.73) 1994 −.50 ∗ (.08) 4.65 ∗ (.71) 1996 −.50 ∗ (.08) 4.46 ∗ (.71) 1998 −.12 ∗ (.08) 4.45 ∗ (.71) = no religion) Religious group (reference Catholic −.23 ∗ (.06) −.71 ∗ (.07) Protestant −.71 ∗ (.04) −.79 ∗ (.06) = female) Gender (reference Male −.02 (.05) .22 ∗ (.05) = non-labor force) Class category (reference Un/semi-skilled manual workers .01 (.10) .49 ∗ (.08) Skilled manual workers −.35 ∗ (.10) .45 ∗ (.08) Routine non-manual workers −.22 ∗ (.08) .25 ∗ (.07) Service class −.05 (.07) −.08 (.06) Self-employed and farmers 2.94 ∗ (.47) .10 (.10) Interactions Un/semi-skilled manual workers × year −.01 ∗ (<.01) __ b Skilled manual workers × year −.01 ∗ (<.01) __ b Male × year −.01 ∗ (<.01) __ b Catholic × year −.01 ∗ (<.01) __ b a An asterisk next to a coefficient denotes significance at the .05 level (2-tailed test). b Coefficient estimate constrained to 0. ",

year = "2006",

month = mar,

doi = "10.1016/j.ssresearch.2004.06.005",

language = "English (US)",

volume = "35",

pages = "88--128",

journal = "Social Science Research",

issn = "0049-089X",

publisher = "Academic Press Inc.",

number = "1",

}

TY - JOUR

T1 - Cleavage-based voting behavior in cross-national perspective

T2 - Evidence from six postwar democracies

AU - Brooks, Clem

AU - Nieuwbeerta, Paul

AU - Manza, Jeff

N1 - Funding Information: Comparative and historical analysis visibly broadens scholarly debates, but past research on the impact of social cleavages has been constrained by either significant methodological limitations or a tendency to limit hypothesis testing to single countries. Virtually no previous studies have attempted to test hypotheses about social cleavage impacts in comparative context using a fully differentiated party family scheme and a multi-category typology of cleavages. The relatively broad scope of our analyses, covering six advanced capitalist democracies for a 30 year period, provides a new perspective from which to address these questions. While future investigations will be needed to refine the understandings developed here, we have shown the possibilities for systematic comparative-historical investigation, and we welcome further work that spans additional countries and time periods as a means of advancing research on cleavage voting in capitalist democracies. It is critical to emphasize, in this context, the relevance of the time period covered by our analyses to debates and unresolved controversies over trends in social cleavage-based voting within capitalist democracies. In particular, the theoretical underpinnings of most predictions concerning a decline or displacement of cleavage voting identify historical factors and novel processes that are hypothesized as emerging primarily during the historical era since the 1960s. These would include the following: new left-libertarian (and neo-conservative or new right) movements; novel (or re-emergent) ideological conflicts concerning issues of rights, the environment, and the character of civil society; recent patterns of postindustrial change; partisan volatility affecting left and other previously-dominant parties during the 1980s; and perhaps more recently, the maturation of globalization and transnational processes (e.g., Franklin et al., 1992; Clark et al., 1993; Dalton and Wattenberg, 1993; Pakulski and Waters, 1996 ). Accordingly, while it could thus be hypothesized that social cleavage voting within specific countries was possibly higher prior to the decade of the 1960s, if true, that line of argumentation might also call into question much of the theoretical basis for anticipating contemporary patterns of cleavage decline. Further research, then, situating the post-1960s era within a longer time frame, while employing suitable measures and methods for the study of social cleavages, may thus help to refine scholarly understanding of the historical origins and magnitude of trends in cleavage-based voting. The substantive findings of this study help to advance recent debates concerning the continuing significance vs. declining relevance of social cleavages to voter alignments and political conflict within capitalist democracies (see Evans (2000) for a recent review). Widely proclaimed assertions regarding the existence of working-class dealignment and its significant consequences for declining cleavage voting within Western democracies are partially supported by these analyses. More specifically, in four of the six countries we have investigated, unskilled workers, traditionally the bulwark of Left party support, have evolved over time to more closely resemble the average voter, making their partisan political alignments far less distinctive, and potentially providing far less incentive for left politicians to develop policies designed to appeal (disproportionately) to unskilled workers. By the same token, however, we find no evidence for a universal decline in the partisan relevance of social cleavages, including class in the post-1960s era. While some countries in our analyses conform to the declining cleavage scenario (Britain, Germany, and to a somewhat lesser extent, the Netherlands), others show some tendency toward a growing magnitude of social cleavages as a whole (Australia and the US), and one country (Austria) has experienced a pattern of stability coupled with a very recent increase. This larger pattern is visible in part because we have simultaneously examined trends among three major social cleavages. This enables various refinements to the overall picture of social cleavages in democracies. For example, we find that cleavage decline in Britain and Germany is a product of decreases in the magnitude of a single cleavage (class); gender and religion-based differences in voter alignments have experienced little change within these two countries. As suggested in the introduction to this paper, the existing scholarly literatures on social cleavages and political behavior is characterized by a sharply polarized debate between scholars who consistently assert decline vs. those who categorically reject such a scenario. Taken as a whole, however, our results suggest an important limitation of these polarized interpretations. Indeed, the clear association between country context and the extent and type of social cleavage change establishes a pattern of cross-national complexity with respect to which past debates over cleavage decline appear ill-suited to understand. Following an emerging strain of commentary on political-sociological research on voting behavior ( Evans, 1999; Evans and Whitefield, 1999; Weakliem and Heath, 1999 ), our results suggest the utility of more detailed and systematic comparisons between countries as a means of advancing debates. In addition to developing such comparisons, further research may be able to shed productive light by considering the potential relevance of such factors as nation-specific electoral systems ( Powell, 2000 ), welfare state type (e.g., Esping-Andersen, 1990; Korpi and Palme, 1998 ; O’Connor et al., 1999 ), other long-standing sources of party-based political cleavages (e.g., Przeworski and Sprague, 1986; Kitschelt, 1994 ), and public opinion (e.g., Page and Shapiro, 1992; Powell, 2000 ). Cross-national differences in trends affecting social cleavages also cast significant doubt on the other interpretive pole of recent debates, namely, that the institutionalization of social group-based conflict within capitalist societies renders them impervious to any process of change affecting voter alignments. While the class cleavage has not declined universally, four of the six countries investigated in this study show clear evidence of working-class dealignment, and the very large decline of class voting is inconsistent with expectations of persistence within all Western democracies. Generic expectations of stability also provide little basis for understanding countries such as Australia and the US in which the magnitude of specific social cleavages has increased. Our results also suggest a degree of similarity in the average size of overall social cleavages within the six democracies we have investigated. Indeed, we find that country-specific patterns of cleavage change have lead to growing comparability in the magnitude of the overall social cleavage: By the mid-1990s, the average difference between two social groups in the probability of favoring a given party family was approximately .10. Although such estimates represent a non-trivial difference in political behavior (and some pairwise contrasts produce much larger differences), we would caution against over-estimating their magnitude. The quite large deviance remaining after estimating our preferred models suggests the further importance of other sources of the vote. Given these results, further research and debate, including with respect to related debates concerning political change ( Brooks et al., 2003 ), may profit by broadening its scope to consider other factors behind the partisan alignments of voters. Appendix A List of original datasets Study Type of study No. of cases Archive Archive No. AUS65 Other 1925 SSDA M168 AUS67 Other 2054 ICPSR 07282 AUS73 Other 4939 SSDA M169 AUS79 Other 2016 SSDA M170 AUS84E NES 3012 SSDA M171 AUS85I ISSP 1528 ZA 1490 AUS86I ISSP 1250 ZA 1620 AUS87E NES 1825 SSDA 445 AUS87I ISSP 1663 ZA 1680 AUS90E NES 2037 SSDA 570 AUS92I ISSP 2203 ZA 2310 AUS93E NES 3023 SSDA 763 AUS96E NES 1797 SSDA 943 AUS98E NES 1897 SSDA 1001 AUT74P Pol.Act. 1585 ICPSR 7777 AUT85I ISSP 987 ZA 1490 AUT88I ISSP 972 ZA 1700 AUT89I ISSP 1997 ZA 1840 AUT91I ISSP 984 ZA 2150 AUT94I ISSP 977 ZA 2620 AUT95I ISSP 1007 ZA 2880 ENG64E NES 1769 ICPSR 7250 ENG66E NES 1874 ICPSR 7250 ENG70E NES 1355 ICPSR 7004 ENG740 NES 2365 UK-DA 0666 ENG79E NES 1893 UK-DA 1533 ENG83E NES 3955 UK-DA 2005 ENG85I ISSP 1530 ZA 1490 ENG86I ISSP 1416 ZA 1620 ENG87E NES 3826 UK-DA 2568 ENG87I ISSP 1212 ZA 1680 ENG88I ISSP 2614 ZA 1700 ENG89I ISSP 2594 ZA 1840 ENG90I ISSP 1197 ZA 1950 ENG91I ISSP 1257 ZA 2150 ENG92I ISSP 1066 ZA 2310 ENG92E NES 3534 DA 2981 ENG93I ISSP 1261 ZA 2450 ENG94I ISSP 984 ZA 2620 ENG95I ISSP 1058 ZA 2880 ENG97I ISSP 1087 ZA 3090 GER69E NES 939 ZA 0525 GER69F NES 1158 ZA 7098 GER75P Pol.Act. 2307 ICPSR 07777 GER76Z ZUBUS 2036 ZA 1233 GER77Z ZUBUS 2002 ZA 1233 GER78C ZUBUS 2030 ZA 1233 GER78X ZUBUS 2012 ZA 1233 GER79X ZUBUS 2007 ZA 1233 GER79Z ZUBUS 2012 ZA 1233 GER80A ZUBUS 2955 ZA 1795 GER80C ZUBUS 1939 ZA 1233 GER80P Pol.Act. 2095 ZA 1188 GER80Z ZUBUS 1997 ZA 1233 GER82A GSS 2991 ZA 1795 GER84A GSS 3004 ZA 1795 GER86A GSS 3095 ZA 1795 GER88A GSS 3052 ZA 1795 GER90A GSS 3051 ZA 1800 GER91A GSS 3058 ZA 1990 GER92A GSS 3548 ZA 2140 GER94A GSS 3450 ZA 2400 GER96A GSS 2800 ZA 2800 GER98A GSS 3234 ZA 3000 NET70 NES 1838 STEIN P1036 NET71 NES 906 ICPSR 7768 NET72E NES 3175 STEIN P0353 NET74P Pol.Act 1201 ICPSR 7777 NET76 Other 755 STEIN P0653 NET77E NES 1856 STEIN P0354 NET77L Other 4159 STEIN P0328 NET79P Pol.Act 806 ZA 1188 NET81E NES 2305 STEIN P0350 NET82E NES 1541 STEIN P0633 NET85S Other 3003 STEIN P1012 NET86E NES 1630 STEIN P0866 NET87 Other 795 STEIN P1194 NET89E NES 1745 STEIN P1000 NET89M Other 956 STEIN P1098 NET90S Other 2384 STEIN P1100 NET94E NES 1812 STEIN P1208 NET96 Other 790 STEIN P1370 NET97I ISSP 2267 ZA 3090 NET98E NES 2101 STEIN P1415 USA64E NES 1834 ICPSR 07235 USA66E NES 1291 ICPSR 07259 USA68E NES 1673 ICPSR 07281 USA70E NES 1694 ICPSR 07298 USA72G GSS 1613 ICPSR 09505 USA73G GSS 1504 ICPSR 09505 USA74G GSS 1484 ICPSR 09505 USA74P Pol.Act 1719 ICPSR 07777 USA75G GSS 1490 ICPSR 09505 USA76G GSS 1499 ICPSR 09505 USA77G GSS 1530 ICPSR 09505 USA78G GSS 1532 ICPSR 09505 USA80G GSS 1468 ICPSR 09505 USA82G GSS 1860 ICPSR 09505 USA83G GSS 1599 ICPSR 09505 USA84G GSS 1473 ICPSR 09505 USA85G GSS 1534 ICPSR 09505 USA86G GSS 1470 ICPSR 09505 USA87G GSS 1819 ICPSR 09505 USA88G GSS 1481 ICPSR 09505 USA89G GSS 1537 ICPSR 09505 USA90G GSS 1372 ICPSR 09505 USA91G GSS 1517 ICPSR 2685 USA93G GSS 1606 ICPSR 2685 USA94G GSS 2992 ICPSR 2685 USA96G GSS 2904 ICPSR 2685 USA98G GSS 2832 ICPSR 2685 Data references ICPSR:7004 Butler, D., & D.E. Stokes, Political Change in Britain, 1969–1970 ICPSR:7214 Campbell, A., P. Converse et al., American National Election Study, 1956 ICPSR:7215 Campbell, A., P. Converse et al., American National Election Study, 1958 ICPSR:7216 Campbell, A., P. Converse et al., American National Election Study, 1960 ICPSR:7235 SRC (Survey Research Center), American National Election Study, 1964 ICPSR:7250 Butler, D., & D.E. Stokes, Political Change in Britain, 1963–1970 ICPSR:7259 SRC (Survey Research Center), American National Election Study, 1966 ICPSR:7281 SRC (Survey Research Center), American National Election Study, 1968 ICPSR:7282 Aitkin, D., M. Kahan & D.E. Stokes, Australian National Political Attitudes, 1967 ICPSR:7298 CPS Center for Political Studies, American National Election Study, 1970 ICPSR:7768 Verba, S., N.H. Nie & J.-O. Kim, Political Participation and Equality in Seven Nations, 1966–1971 ICPSR:7777 Barnes, S.H., & M. Kaase et al., Political Action: An Eight Nation Study, 1973–1976 ICPSR:9505 Davis, J.A. & T.W. Smith, General Social Survey Cumulative File, 1972–1991 ICPSR:xxxx Davis, J.A. & T.W. Smith, General Social Survey Cumulative File, 1972–1998 NSD:NOR77e Names of principal investigators: Valen, H., B. Aardal, Institute of social research, Oslo Norwegian Election Study, 1977 [computer file]. Data collector: Statistics Norway, Oslo SSDA:M168 Broom, L., F.L. Jones & J. Zubrzycki, Social Stratification in Australia, 1965 SSDA:M169 Broom, L., P. Duncan-Jones, F.L. Jones, P. McDonnel & T. Williams, Social Mobility in Australia Project, 1973 SSDA:M170 Aitkin, D., MacQuarie University Australian Political Attitudes Survey, 1979 SSDA:M171 Kelley, J.L., R.G. Cushing & B. Headey, Australian National Social Science Survey, 1984 SSDA:0445 McAllister, I. & A. Mughan, Australian Election Survey, 1987 SSDA:0570 McAllister, I., R. Jones, E. Papodalis & D. Gow, Australian Election Survey, 1990 SSDA: 0763 Roger Jones, Ian McAllister, David Denemark, David Gow, Australian Election Study, 1993 SSDA: 0943 Roger Jones, Ian McAllister, David Gow, Australian Election Study, 1996 SSDA: 1001 Clive Bean, David Gow, Ian McAllister, Australian Election Study, 1998 STEIN:P0136 Heunks, F. M., M.K. Jennings, W.E. Miller, P.C. Stouthard & J. Thomassen, Dutch Election Study, 1970–1973 STEIN:P0328 Centraal Bureau voor Statistiek (CBS), Life Situation Survey, Netherlands 1977 STEIN:P0350 Werkgroep Nationaal Kiezersonderzoek, Dutch Parliamentary Election Study, 1981 STEIN:P0353 Werkgroep Nationaal Kiezersonderzoek, Dutch Parliamentary Election Study, 1972 STEIN:P0354 Werkgroep Nationaal Kiezersonderzoek, Dutch Parliamentary Election Study, 1977 STEIN:P0633 Werkgroep Nationaal Kiezersonderzoek, Dutch Parliamentary Election Study, 1982 STEIN:P0653 Hermkens, P.L.J. & P.J. van Wijngaarden, Criteria for Justification of Income Differences, Netherlands 1976 STEIN:P0866 Van der Eijk, C., G.A. Irwin & B. Niemoeller, Dutch Parliamentary Elections Study, 1986 STEIN:P1012 Felling, A.J.A., J. Peters & O. Schreuder, Social Relevance of Religion in the Netherlands, 1985 STEIN:P1098 Arts, K., E. Hollander, K. Renckstorf & P. Verschuren, Media-equipment, Media-exposure and Media-use in the Netherlands, 1989 STEIN:P1100 Felling, A., J. Peters, & O. Schreuder, Social Relevance of Religion in the Netherlands, 1990 STEIN:P1194 Hermkens, P.L.J. & P.J. van Wijngaarden, Criteria for Justification of Income Differences, Netherlands 1987 STEIN:P1415 Aarts, K., H. van der Kolk, M. Kamp, J. Thomassen, Dutch Parliamentary Election Study, 1998 STEIN:P1208 Stichting Kiezersonderzoek Nederland, Dutch Parliamentary Election Study, 1994 UK-DA:0666 Crewe, I., B. Saerlvik & J. Alt, British Election Study: October 1974, Cross-section UK-DA:1533 Crewe, I., B. Saerlvik & D. Robertson, British Election Study: May 1979, Cross-section UK-DA:2005 Heath, A.F., R.M. Jowell, J.K. Curtice & E.J. Field, British Election Study: June 1983 UK-DA:2568 Heath, A.F., et al., British Election Study: 1987. ZA:0426 Klingemann, H.-D., & F.U. Pappi, German Pre- and Post-Election Study, 1969 ZA:0525 Kaase, M., U. Schleth, W. Adrian, M. Berger & R. Wildenmann, German Election Study: August–September 1969 ZA:1188 Allerbeck, K.R., M. Kaase, H.-D. Klingemann, Ph.C. Stouthard, F.J. Heunks, J.J.A. Thomassen, J.W. van Deth, S.H. Barnes, B.G. Farah, R. Inglehart, and M.K. Jennings, Political Action II, 1979–1980 ZA:1233 ZUMA Zentrum fuer Umfragen, Methoden, und Analysen, ZUMA-Standard Demographie (Zeitreihe), Germany 1976–1980 ZA:1490 International Social Survey Programme (ISSP), International Social Survey Programme: Role of Government, 1985–1986 ZA:1620 International Social Survey Programme (ISSP), International Social Survey Programme: Social Networks and Support Systems, 1986 ZA:1680 International Social Survey Programme (ISSP), International Social Survey Programme: Social Inequality, 1987 ZA:1700 International Social Survey Programme (ISSP), International Social Survey Programme: Family and Changing Sex Roles, 1988 ZA:1795 Allerbeck, K.R., M.R. Lepsius, K.U. Mayer, W. Müller, K.-D. Opp, F.U. Pappi, E.K. Scheuch, and R. Ziegler, Allgemeine Bevolkerungsumfrage der Sozialwissenschaften Allbus Kumulierter Datensatz, 1980–1988 ZA:1800 ZUMA Zentrum fuer Umfragen, Methoden, und Analysen, German Social Survey (Allbus), 1990 ZA:1840 International Social Survey Programme (ISSP), International Social Survey Programme: Work Orientations, 1989 ZA:1950 International Social Survey Programme (ISSP), International Social Survey Programme: Role of Government II, 1990 ZA:2150 International Social Survey Programme (ISSP), International Social Survey Programme: Religion, 1991 ZA:2450 International Social Survey Programme (ISSP), International Social Survey Programme: Environment, 1993 ZA:2620 International Social Survey Programme (ISSP), International Social Survey Programme: Family and Changing Gender Roles II, 1994. ZA:2880 International Social Survey Programme (ISSP), International Social Survey Programme: National Identity, 1995 ZA:3090 International Social Survey Programme (ISSP), International Social Survey Programme: Work Orientations II, 1997 ZA:3000 ZUMA Zentrum fuer Umfragen, Methoden, und Analysen, German Social Survey (Allbus), 1990–1998 Appendix B Party families and political parties by country Party family Party name Australia a Left Communist Party, Australian Labor Party, Queensland Labor Party, Lang Labor Party Liberal Australia Party, Liberal Movement, National Alliance, Service Party of Australia, Australian Democrats, Country-Liberal Party Other The Greens, a Tasmanian Greens, a Western Australia Greens, a Queensland Greens, a Victorian Greens a National Party Call to Australia Party, Democratic Labor Party, Australian Liberal Party, Other Parties Austria b Left Communist Party Socialist Green United Greens of Austria, Alternative List, Green Alternative Liberal Liberal Forum Religious Austrian Peoples Party Other b Association of Members of the Social Security System, No-Citizens’ Initiative against the Sale of Austria, Czechs, National Socialist, German Workers’ Party, Hitler movement, Fatherland Front, Democratic Progressive Party, Freedom Party, Other Parties Germany Left German Communist Party, German Peace Union, Action for Democratic Progress Party of Democratic Socialism, Social Democrats, All-German People’s Party Green Greens, Ecological Democratic Party, Alliance 90-Greens Liberal Free Democrats Religious Christian Democratic Union, Christian Social Union Other Centre Party, All-German Party, German Social Union, The Greys, Bavarian Party, South Schleswig Voters’ League, Refugee Party, Federal Union Economic Reconstruction League, German Reich Party, National Democratic Party, Republicans, German Party, Other Parties The Netherlands c Left Communist Party, Netherlands Labour Party, Democratic Socialists ’70, Pacifist Socialist Party, Socialist Party, Green Left, c Greens c Liberal People’s Party for Freedom and Democracy, Democrats 66 Religious Anti-Revolutionary Party, Catholic People’s Party, Christian Historical Union, Political Reformed Party, Catholic National Party, Reformed Political Union, Radical Political Party, Roman Catholic Party, Christian Democratic Appeal, Reformed Political Federation, Evangelical People’s Party Other Centre Party, Centre Democrats, People’s Party of the Right, Middle Class Party, United Old Persons’ League, 55 + Union, Other Parties United Kingdom c Left Communist Party, Independent Labour Party, Labour Party, Social Democratic and Labour Party, Green Party c Liberal Liberal Party, Liberal Democrat Party Conservative Conservative Party, National Liberal Party Other National Front, Sinn Féin, Ulster Unionist Party, Democratic Unionist Party, Alliance Party of Northern Ireland, Plaid Cymru (Party of Wales), Scottish National Party, Other Party United States Liberal Democratic Party Conservative Republican Party Other Communist Party d Socialist Labor Party d Socialist Party d Progressive Party d Progressive Party (H. Wallace), d American Party, Independent (J. Anderson), Independent (Ross Perot), Libertarian Party, Other Party a The very small number of Green party family supporters in Australia are combined with Other. b The very small number of Other party family supporters in Austria are treated as “non-voters” and deleted from the analyses. c In the Netherlands and Great Britain the very small number of Green party family supporters are combined with Left party. d In the US, third and independent candidates are combined in the Other category. Appendix C Multinomial logistic regression coefficients = 17,944) a (SE in parentheses) for preferred model of social cleavages in Australia ( N Independent variables Other vs. Left party family Liberal vs. Left party family Constant −.54 ∗ (.20) 2.45 ∗ (.15) = 1979) Year (reference 1984 1.48 ∗ (.24) .43 ∗ (.19) 1985 .66 ∗ (.26) −.14 (.20) 1986 2.55 ∗ (.26) .17 (.22) 1987 1.52 ∗ (.01) −.52 ∗ (.16) 1990 −.33 (.22) −.1.49 ∗ (.16) 1992 2.33 ∗ (.22) −.33 (.18) 1993 2.80 ∗ (.22) .01 (.18) 1995 2.21 ∗ (.21) −.95 ∗ (.17) 1996 2.36 ∗ (.20) −.89 ∗ (.18) 1998 2.33 ∗ (.22) −.72 ∗ (.18) = no religion) Religious group (reference Catholic .52 ∗ (.09) .4 8 ∗ (.09) Protestant .76 ∗ (.08) .21 ∗ (.08) = female) Gender (reference Male .30 ∗ (.07) .40 ∗ (.07) = non-labor force) Class category (reference Un/semi-skilled manual workers −.44 ∗ (.10) −.06 (.14) Skilled manual workers −.33 ∗ (.15) .01 (.14) Routine non-manual workers −.12 ∗ (.10) −.16 ∗ (.10) Service class −.4 8 ∗ (.08) −.64 ∗ (.08) Self-employed and farmers 2.00 ∗ (.46) −.44 ∗ (.17) Interactions Self-employed × year −.06 ∗ (.02) ___ b a An asterisk next to a coefficient denotes significance at the .05 level (2-tailed test). b Coefficient estimate constrained to 0. Appendix D Multinomial logistic regression coefficients = 5754) a (SE in parentheses) for preferred model of social cleavages in Austria ( N Independent variables Left vs. Religious party family Green vs. Religious party family Liberal vs. Religious party family Constant −2.43 ∗ (.18) −23.14 ∗ (.41) −.63 ∗ (.25) = 1974) Year (reference 1985 −.34 ∗ (.10) 21.23 ∗ (.41) −.26 (.24) 1988 −.21 (.12) 22.64 ∗ (.40) .02 (.26) 1989 −.06 (.09) 23.23 ∗ (.36) 1.03 ∗ (.18) 1991 1.29 ∗ (.17) 23.03 ∗ (.38) .86 ∗ (.20) 1994 .00 (.1 1) 23.16 ∗ (.38) .48 ∗ (.22) 1995 −.06 (.12) 23.97 ∗ (.37) 1.92 ∗ (.18) = no religion) Religious group (reference Catholic −2.11 ∗ (.16) −2.04 ∗ (.22) −.1.88 ∗ (.20) Protestant −.1.43 ∗ (.22) −.1.12 ∗ (.32) −.50 ∗ (.27) = female) Gender (reference Male .02 (.07) .05 (.14) .32 ∗ (.1 1) = non-labor force) Class category (reference Un/semi-skilled manual workers .44 ∗ (.1 1) −.25 (.32) .10 (.20) Skilled manual workers .61 ∗ (.12) .03 (.30) .61 ∗ (.18) Routine non-manual workers .08 (.10) .83 ∗ (.19) .27 (.17) Service class −.32 ∗ (.10) .97 ∗ (.17) −.02 (.16) Self-employed and farmers −.1.82 ∗ (.13) −.1.13 ∗ (.33) −.30 (.17) a An asterisk next to a coefficient denotes significance at the .05 level (2-tailed test). Appendix E Multinomial logistic regression coefficients = 27,758) a (SE in parentheses) for preferred model of social cleavages in Britain ( N Independent variables Cons. vs. Left party family Other vs. Left party family Liberal vs. Left party family Constant −.70 ∗ (.07) −4.29 ∗ (.35) −2.17 ∗ (.11) = 1966) Year (reference 1970 .05 (.08) .67 (.43) −.10 (.15) 1974 .22 ∗ (.08) 1.92 ∗ (.36) 1.12 ∗ (.10) 1979 .67 ∗ (.08) 1.26 ∗ (.39) 1.01 ∗ (.13) 1983 .67 ∗ (.07) 1.35 ∗ (.37) 1.57 ∗ (.1 1) 1985 .04 (.09) 1.02 ∗ (.40) .63 ∗ (.13) 1986 .19 ∗ (.09) 1.66 ∗ (.38) .66 ∗ (.13) 1987 .61 ∗ (.07) 1.30 ∗ (.36) 1.53 ∗ (.1 1) 1988 .33 ∗ (.08) 2.12 ∗ (.35) .20 (.13) 1989 .52 ∗ (.08) 3.18 ∗ (.34) .60 ∗ (.12) 1990 .13 ∗ (.10) 2.77 ∗ (.36) .04 (.16) 1991 .41 ∗ (.09) 2.12 ∗ (.37) 1.00 ∗ (.14) 1992 .48 ∗ (.10) 2.42 ∗ (.37) 1.11 ∗ (.14) 1993 .29 ∗ (.09) 2.15 ∗ (.37) .86 ∗ (.14) 1994 −.08 (.10) 1.94 ∗ (.38) .82 ∗ (.14) 1995 −.19 (.10) 1.74 ∗ (.38) .84 ∗ (.14) 1997 −.07 (.10) 1.97 ∗ (.38) .57 ∗ (.15) = no religion) Religious group (reference Catholic −.22 ∗ (.05) −.78 ∗ (.13) −.35 ∗ (.07) Protestant .57 ∗ (.03) −.16 ∗ (.07) .35 ∗ (.04) = female) Gender (reference Male −.04 (.03) −.22 ∗ (.07) −.14 ∗ (.04) = non-labor force) Class category (reference Un/semi-skilled manual workers −.85 ∗ (.12) −.01 (.12) −.15 ∗ (.07) Skilled manual workers −.62 ∗ (.05) −.11 (.11) −.34 ∗ (.06) Routine non-manual workers .18 (.11) .36 ∗ (.11) .49 ∗ (.06) Service class .70 ∗ (.04) .43 ∗ (.10) .81 ∗ (.05) Self-employed and farmers 1.52 ∗ (.18) .61 ∗ (.18) .53 ∗ (.11) Interactions Un/semi-skilled manual worker × year .02 ∗ (.01) __ b __ b Routine non-manual workers × year .01 ∗ (<.01) __ b __ b Self-employed × year −.02 ∗ (.01) __ b __ b a An asterisk next to a coefficient denotes significance at the .05 level (2-tailed test). b Coefficient estimate constrained to 0. Appendix F Multinomial logistic regression coefficients = 39,874) a (SE in parentheses) for preferred model of social cleavages in Germany ( N Independent variables Left vs. Religious party family Green vs. Religious party family Liberal vs. Religious party family Constant .92 ∗ (.07) −24.83 ∗ (.15) −2.43 ∗ (.22) −1.73 ∗ (.14) = 1969) Year (reference 1975 −.09 (.08) −16.12 ∗ (<.01) −1.37 ∗ (.38) .99 ∗ (.15) 1976 −.03 (.08) −16.05 ∗ (<.01) −.22 (.28) 1.26 ∗ (.15) 1977 −.06 (.08) −16.12 ∗ (<.01) −.84 ∗ (.34) .82 ∗ (.16) 1978 .00 (.07) −16.09 ∗ (<.01) .75 ∗ (.21) 1.00 ∗ (.15) 1979 −.07 (.07) −16.16 ∗ (<.01) −1.19 ∗ (.31) .69 ∗ (.15) 1980 .04 (.06) 23.51 ∗ (.16) −1.43 ∗ (.27) 1.17 ∗ (.14) 1982 −.62 ∗ (.07) 24.18 ∗ (.16) −1.57 ∗ (.37) .90 ∗ (.15) 1984 −.19 ∗ (.08) 24.41 ∗ (.16) −1.80 ∗ (.42) .16 (.17) 1986 .00 (.08) 24.49 ∗ (.16) −1.38 ∗ (.38) .72 ∗ (.16) 1987 −.06 (.10) 24.19 ∗ (.16) −.52 (.39) .70 ∗ (.19) 1988 .13 (.08) 24.35 ∗ (.17) −.14 (.27) .58 ∗ (.17) 1990 .09 (.08) 24.42 ∗ (.16) .33 (.24) .96 ∗ (.15) 1991 −.32 ∗ (.08) 23.80 ∗ (.17) −.23 (.26) 1.03 ∗ (.15) 1992 −.19 ∗ (.08) 24.20 ∗ (.17) .21 (.23) .81 ∗ (.15) 1994 .14 (.08) 24.65 ∗ (.16) 1.05 ∗ (.22) .89 ∗ (.16) 1996 −.26 ∗ (.08) 24.22 ∗ (.16) −.03 (.24) .46 ∗ (.16) 1998 .15 (.08) 24.00 ∗ (.17) .60 ∗ (.23) −.01 (.17) = no religion) Religious group (reference Catholic −1.39 ∗ (.04) −1.56 ∗ (.07) −1.20 ∗ (.11) −1.31 ∗ (.05) Protestant −.52 ∗ (.04) −.87 ∗ (.06) −.84 ∗ (.10) −.35 ∗ (.06) = female) Gender (reference Male −.03 (.03) −.19 ∗ (.08) .43 ∗ (.08) −.04 (.04) = non-labor force) Class category (reference Un/semi-skilled manual workers .88 ∗ (.09) .31 ∗ (.08) .50 ∗ (.11) .18 ∗ (.07) Skilled manual workers .85 ∗ (.12) .03 (.11) .45 ∗ (.15) .07 (.08) Routine non-manual workers .06 (.04) .34 ∗ (.08) .26 (.15) .37 ∗ (.06) Service Class .04 (.03) −.05 (.24) −.20 (.12) .68 ∗ (.05) Self-employed and farmers −1.31 ∗ (.07) −.59 ∗ (.14) −.20 (.19) .12 (.08) Interactions Un/semi-skilled manual worker × year −.02 < .01) ∗ ( __ b __ b __b Skilled manual worker × year −.03 ∗ (.01) __ b __ b __ b Service class × year __ b .03 ∗ (.01) __ b __ b a An asterisk next to a coefficient denotes significance at the .05 level (2-tailed test). b Coefficient estimate constrained to 0. Appendix G Multinomial logistic regression coefficients = 21,008) a (SE in parentheses) for preferred model of social cleavages in Netherlands ( N Independent variables Other vs. Left party family Liberal vs. Left party family Liberal vs. Religious party family Constant 1.50 ∗ (.10) −2.87 ∗ (.35) 1.11 ∗ (.10) Year (reference =1970) 1971 −.17 (.13) .76 (.42) −.43 ∗ (.13) 1972 .08 (.09) 1.31 ∗ (.34) −.33 ∗ (.10) 1974 .55 ∗ (.11) 1.57 ∗ (.38) −.21 (.12) 1977 .30 ∗ (.08) .87 ∗ (.35) .08 (.09) 1979 .31 ∗ (.14) .08 (.66) .12 (.13) 1981 .14 (.13) .33 (.53) .13 (.13) 1982 .36 ∗ (.13) .32 (.56) .18 (.13) 1985 .28 ∗ (.11) 1.99 ∗ (.36) −.06 (.11) 1986 .06 (.13) 1.02 ∗ (.44) −.45 ∗ (.14) 1989 −.10 (.15) 1.04 ∗ (.45) −.59 ∗ (.15) 1990 −.64 ∗ (.13) 2.59 ∗ (.36) −.33 ∗ (.12) 1994 .19 (.16) 3.73 ∗ (.37) .58 ∗ (.15) 1996 .04 (.18) 3.51 ∗ (.39) .57 ∗ (.17) 1998 .45 ∗ (.16) 3.58 ∗ (.38) .34 ∗ (.16) = no religion) Religious group (reference Catholic −3.39 ∗ (.12) −1.50 ∗ (.29) −2.55 ∗ (.13) Protestant −2.51 ∗ (.12) −1.36 ∗ (.30) −2.06 ∗ (.13) = female) Gender (reference Male .26 ∗ (.08) −.29 ∗ (.08) −.05 (.04) = non-labor force) Class category (reference Un/semi-skilled manual workers .32 ∗ (.07) .32 ∗ (.16) −.54 ∗ (.15) Skilled manual workers .46 ∗ (.07) .66 ∗ (.15) −.65 ∗ (.15) Routine non-manual workers .05 (.06) .17 (.13) .39 ∗ (.18) Service class −.19 ∗ (.06) .31 ∗ (.10) .52 ∗ (.05) Self-employed and farmers −1.22 ∗ (.11) 1.83 ∗ (.45) .29 ∗ (.08) Interactions Un/semi-skilled manual worker × year __ b __ b .03 ∗ (<.01) Skilled manual workers × year __ b __ b .03 ∗ (<.01) Self-employed × year __ b −.12 ∗ (.03) __ b Male × year −.01 ∗ (<.01) __ b __ b Catholic × year .04 ∗ (<.01) −.03 ∗ (.01) .02 ∗ (<.01) Protestant × year <−.01 (<.01) −.04 ∗ (.01) −<.01 (<.01) a An asterisk next to a coefficient denotes significance at the .05 level (2-tailed test). b Coefficient estimate constrained to 0. Appendix H Multinomial logistic regression coefficients = 39,789) a (SE in parentheses) for preferred model of social cleavages in the US ( N Independent variables Liberal vs. Cons. party family Other vs. Cons. party family Constant 1.63 ∗ (.07) −4.81 ∗ (.71) = 1964) Year (reference 1966 −.28 ∗ (.11) −38.50 (>.01) 1968 −1.00 ∗ (.08) 3.97 ∗ (.71) 1970 −.70 ∗ (.20) 2.01 ∗ (.74) 1972 −1.32 ∗ (.08) 2.01 ∗ (.73) 1973 −1.49 ∗ (.08) 2.57 ∗ (.73) 1974 −1.49 ∗ (.07) 2.85 ∗ (.72) 1975 −1.34 ∗ (.08) 2.78 ∗ (.72) 1976 −1.29 ∗ (.08) 3.22 ∗ (.72) 1977 −.59 ∗ (.08) 1.92 ∗ (.72) 1978 −.73 ∗ (.08) 2.55 ∗ (.73) 1980 −.64 ∗ (.08) 2.53 ∗ (.73) 1982 −.64 ∗ (.08) 2.47 ∗ (.73) 1983 −.92 ∗ (.08) 2.00 ∗ (.74) 1984 −.94 ∗ (.08) 2.12 ∗ (.73) 1985 −1.37 ∗ (.08) 1.96 ∗ (.73) 1986 −1.34 ∗ (.09) 2.12 ∗ (.73) 1987 −.90 ∗ (.08) 2.29 ∗ (.73) 1988 −1.40 ∗ (.09) 2.45 ∗ (.73) 1989 −1.13 ∗ (.09) 2.31 ∗ (.72) 1990 −1.49 ∗ (.09) 2.10 ∗ (.73) 1991 −1.43 ∗ (.09) 2.10 ∗ (.73) 1993 −.61 ∗ (.09) 4.67 ∗ (.73) 1994 −.50 ∗ (.08) 4.65 ∗ (.71) 1996 −.50 ∗ (.08) 4.46 ∗ (.71) 1998 −.12 ∗ (.08) 4.45 ∗ (.71) = no religion) Religious group (reference Catholic −.23 ∗ (.06) −.71 ∗ (.07) Protestant −.71 ∗ (.04) −.79 ∗ (.06) = female) Gender (reference Male −.02 (.05) .22 ∗ (.05) = non-labor force) Class category (reference Un/semi-skilled manual workers .01 (.10) .49 ∗ (.08) Skilled manual workers −.35 ∗ (.10) .45 ∗ (.08) Routine non-manual workers −.22 ∗ (.08) .25 ∗ (.07) Service class −.05 (.07) −.08 (.06) Self-employed and farmers 2.94 ∗ (.47) .10 (.10) Interactions Un/semi-skilled manual workers × year −.01 ∗ (<.01) __ b Skilled manual workers × year −.01 ∗ (<.01) __ b Male × year −.01 ∗ (<.01) __ b Catholic × year −.01 ∗ (<.01) __ b a An asterisk next to a coefficient denotes significance at the .05 level (2-tailed test). b Coefficient estimate constrained to 0.

PY - 2006/3

Y1 - 2006/3

N2 - We investigate trends and cross-national variation in the impact of class, religious, and gender cleavages on voting behavior in six advanced capitalist democracies in the postwar period. Earlier research on cleavage voting has been criticized for utilizing outdated "two-class" models of class structure, simplistic left/right party distinctions, flawed statistical approaches, and incompatible and/ or limited of cross-national empirical evidence. We take such criticisms seriously and seek to overcome them. Using multinomial logistic regression models, we analyze data from a new dataset, the International Social Cleavages and Politics (ISCP) file, which contains comparable, over-time cross-national data from 112 nationally representative election surveys of voters in six Western democracies in the period 1964-1998. The six countries examined in the paper (Australia, Austria, Germany, Great Britain, the Netherlands, and the United States) are unique because of the existence of time series data available for all three cleavages. Our analyses examine the changing magnitude of the class, religion, and gender cleavages for up to five distinct party families for each country. Unskilled workers have become less distinctive in their partisan alignments over time, but other classes have experienced offsetting changes, yielding little evidence of a universal decline in the class cleavage. Further analyses suggest an important degree of stability in the aggregated effects of all social cleavages, while also revealing significant cross-national differences and trends in the magnitude of specific cleavages. These results refine debates concerning the possible decline of social cleavages; implications for research are discussed in conclusion.

AB - We investigate trends and cross-national variation in the impact of class, religious, and gender cleavages on voting behavior in six advanced capitalist democracies in the postwar period. Earlier research on cleavage voting has been criticized for utilizing outdated "two-class" models of class structure, simplistic left/right party distinctions, flawed statistical approaches, and incompatible and/ or limited of cross-national empirical evidence. We take such criticisms seriously and seek to overcome them. Using multinomial logistic regression models, we analyze data from a new dataset, the International Social Cleavages and Politics (ISCP) file, which contains comparable, over-time cross-national data from 112 nationally representative election surveys of voters in six Western democracies in the period 1964-1998. The six countries examined in the paper (Australia, Austria, Germany, Great Britain, the Netherlands, and the United States) are unique because of the existence of time series data available for all three cleavages. Our analyses examine the changing magnitude of the class, religion, and gender cleavages for up to five distinct party families for each country. Unskilled workers have become less distinctive in their partisan alignments over time, but other classes have experienced offsetting changes, yielding little evidence of a universal decline in the class cleavage. Further analyses suggest an important degree of stability in the aggregated effects of all social cleavages, while also revealing significant cross-national differences and trends in the magnitude of specific cleavages. These results refine debates concerning the possible decline of social cleavages; implications for research are discussed in conclusion.

UR - http://www.scopus.com/inward/record.url?scp=28844479736&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=28844479736&partnerID=8YFLogxK

U2 - 10.1016/j.ssresearch.2004.06.005

DO - 10.1016/j.ssresearch.2004.06.005

M3 - Article

AN - SCOPUS:28844479736

SN - 0049-089X

VL - 35

SP - 88

EP - 128

JO - Social Science Research

JF - Social Science Research

IS - 1

ER -

Cleavage-based voting behavior in cross-national perspective: Evidence from six postwar democracies

Abstract

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this