Most of the work on policy dynamics focuses on the agenda-setting stage of the policy cycle and argues that policy issues wax and wane in public attention, generating either a cyclical or evolutionary pattern of governmental activity in particular policy sectors. Anthony Downs's notion of a periodic “issue-attention cycle” and Frank Baumgartner and Bryan Jones's notion of a stepped or “punctuated equilibrium” pattern of policy change are prominent in the literature, but have received little empirical and virtually no cross-national verification. Utilizing the analysis of time-series data gathered on nuclear energy and acid rain issues appearing on government and public agendas in Canada over the period 1977–1992, this article elaborates the elements of the two models and subjects both to empirical testing. The article finds little support for either model in the Canadian case and argues the assumptions behind the models must be altered to account for this anomalous case.

Rational choice theory has drawn attention to the phenomenon of structure-induced equilibrium in situations of potential cycling. When there is no majority, first preference or Condorcet winner, the outcome is determined by agenda control and institutional rules of decision making. Within that context, the status quo has a special advantage because of the parliamentary amendment procedure, in which the status quo, as the default option to the bill in formal form, is not voted upon until the last stage. The unsuccessful attempts of the Canadian government of Prime Minister Brian Mulroney to respond legislatively to the Supreme Court's Morgentaler decision illustrate these general principles of rational choice. The government was unable to get legislation passed because, with cyclical configurations of opinion in both the House of Commons and the Senate, institutional rules, especially the order of voting required by the parliamentary amendment procedure, favoured the status quo.

In this article we seek to find a way out of the antinomic debate pitting the advocates of the substantialist thesis against those who defend constructivist theories regarding the nation as a “place so named” (lieu dit). Using examples from the case of Quebec—with a view to illustrating a line of argument much more than establishing an empirical interpretation—we attempt to reconcile perspectives from a semantic of action with those from a pragmatic of action within the framework of a theory-in-progress of narraction.

In this article, it is argued that Canada's relatively low rate of political participation is related to its electoral system being nonproportional, but that a complementary factor is to be found in its political institutions being discontinuous. Discontinuous institutions are manifested in relatively weak links between political organizations active municipally, regionally (provincially) and nationally. While the relationship between proportional representation (PR) and high turnout has been well established in the literature, there is still a puzzle surrounding the theoretical explanation for it. The author argues that the key to the solution to this puzzle lies less in the additional potential benefits to the voter in a PR system than in the reduction of costs, specifically information costs under such a system. PR is seen to frame incentives and disincentives for political actors in such a manner as to result in a reduction of the cost of political information. This is especially the case when PR is embedded in integrated (non-discontinuous) political institutional arrangements. The most salient manifestation of this effect is seen in comparative turnout levels in municipal elections.

This article examines the 1992 Progressive Conservative leadership election in Alberta, Canada, and assesses the degree to which this universal ballot produced a leadership electorate distinctive from that enfranchised at the previous convention. The universal ballot electorate possessed a limited background in the party, was not heavily involved in the campaign but included a much higher proportion of women and substantially more elderly voters. Participation remained essentially a middle-class endeavour. Although many of the concerns raised by critics of the universal ballot are confirmed by the Alberta experiment, there are positive features to the Alberta model and the move to universal balloting is not without its merits.

This article offers an improved understanding of the often-neglected religious cleavage in Canada. Using data from the 1988 Canadian Election Study, the authors examine the questions of voting intentions, opinion on abortion, support for unions and support for military spending. This article first shows that Roman Catholics have different opinions on these questions than Protestants, but that these differences disappear as Catholics' exposure to the news media increases. The authors also demonstrate that the more religious the Catholics, the more resistant they are to media messages. The largest differences between Catholics and Protestants are thus found between regular churchgoers with low media exposure, while the smallest differences are found between heavily exposed non-practisers.