and institutions of the "new" Europe,
others less so. But only a few are truly knowledgeable about the specifics
of a newly negotiated treaty such as Maastricht, Amsterdam, or Nice.
As we have seen in some of the cases discussed here, referendums can
be highly partisan contests, even without the appearance of party or can-
didate names on the ballot. If the positions of parties on an issue are well
known or if the referendum debate follows clearly understood ideological
lines, votingbehaviour may tend to conform to familiar and relatively
predictable patterns. In
Scandinavia," Electoral Studies
Jahn, Detlef, and Ann-Sofie Storsved.
1995. "Legitimacy through Referen-
dum: The Nearly Successful Domino
Strategy of the EU Referendums
in Austria, Finland, Sweden and
Norway," West European Politics 18:
Jenssen, Anders Todal, Pertti Pesonen
and Mikael Gilljam, eds. 1998. To Join
or Not to Join: Three Nordic Referen-
dums on Membership in the European
Union. Oslo: Scandinavian University
Midtbo, Tor and Kjell Hines. 1998. "The
Referendum-Election Nexus: An
Aggregate Analysis of Norwegian
FRANCOIS GARON is a Ph.D. candidate at the Ecole nationale d'administration
publique (ENAP). His research deals with the consultative process in Canada and
France, especially in the field of biotechnologies.
ELISABETH GIDENGIL is Professor in the Department of Political Science at
McGill University. She specializes in the study of votingbehaviour, public opinion,
gender, and the media. She has been a member of the 1993, 1997, and 2000 Cana-
dian Election Study teams. She is a co-author of Anatomy of a Liberal Victory: Mak-
ing Sense of the 2000 Canadian
. ‘Linkages Between Citizens and Politicians in
Democratic Polities.’ Comparative Political Studies 33 (6): 845–79.
Kobayashi, Masayoshi. 2001. Minna no yuubinbunkashi: Kindai nihon wo sodateta
jouhoudentatsu shisutemu. Tokyo: Nijuuni.
Kohei, Shinsaku, Ichiro Miyake, and Joji Watanuki. 1991. ‘Issues and VotingBehavior.’ In The Japanese Voter, edited by Scott C. Flanagan et al. New
Haven, CT: Yale University Press.
Koizumi, Jun’ichirou. 1999. ‘Forward.’ In Yuusei mineikaron: Nihon saisei no dai
kaikaku, edited by Koizumi Jun’ichirou and Matsuzawa Shigefumi. Tokyo:
THE SOCIAL BASES OF POLITICAL CLEAVAGE IN 1962*
Robert R • Al ford
The classic studies of votingbehaviour have usually examined the
way in which workers differ from businessmen, Protestants from
Catholics, rural residents from urban ones, young persons from
old persons, males from females. Numerous generalizations have
been erected, which have seemed to hold for most American elec-
tions, and a number of British ones. Workers are more likely to
vote for a left-wing party than are businessmen; so are Catholics,
urban residents, youth, and men. ( I ) No
differences between the types of polls.
The earliest polls on political opinion were of the action-oriented type
but they did not use scientific methods. As we noted earlier, the straw
polls conducted by various media aimed at predicting votingbehaviour.
The key characteristic of this polling method was the ability of any
voter to be included in the poll if they cared to be; voters just had to
respond to the questions. Of course, straw polls make no attempt to
ensure that everyone had the chance to be included in the poll, and, as
we have seen, they were dismissed for this
Aldrich, John H., and Forrest D. Nelson.
1984. Linear Probability, Logit, and
Probit Models. Beverly Hills, Calif.:
Archer, Keith. 1987. "A Simultaneous
Equation Model of Canadian VotingBehaviour," Canadian Journal of Politi-
cal Science 20: 553-72.
Archer, Keith, and Faron Ellis. 1994.
"Opinion Structure of Party Activ-
ists: The Reform Party of Canada,"
Canadian Journal of Political Science 27:
Banting, Keith, and Richard Simeon.
1983. And No One Cheered: Federal-
ism, Democracy and the Constitution
Act. Toronto: Methuen.
• Effect of education on own health and spouse’s health.
• Effect of education on consumer-choice efficiency, labour-market
search efficiency, adaptability to new jobs, marital choice, savings,
and attainment of desired family size.
• Effect of education on charitable giving and volunteer activity.
• Effect of schooling on social cohesion: votingbehaviour, reduced
alienation, and smaller social inequalities.
• Effect of education on reducing reliance on welfare and other social
• Effect of schooling on reduced criminal activity.
Many of the studies also find
the left of other citizens, though controls for
union membership tend to depress the effect of government employ-
ment by a discernible amount. It is interesting to note that Blake finds
the strongest effects of government employment on ideology among
those in professional and managerial/technical occupations, while in-
dividuals in other occupational categories hold relatively indistinct po-
litical views based on whether or not they work for the public sector.
Regarding votingbehaviour, Blais, Blake, and Dion (1991) find that gov-
ernment employees are
groups through the study of partisanship, votingbehaviour,
and political preferences. Even though the average citizen holds several
identities based on age, class, gender, and religion, evidence from cross-
national studies suggests that common experiences of social, cultural,
and political disadvantage disproportionately affect some immigrant
and ethnic minority groups more than others. For example, “non-
white” or “non-Western” immigrants in Australia appear to be more
exposed to negative stereotyping and to institutional forms of social