Accounting for regional divides in British electoral behaviour during the 1980s has proved controversial. Critics have argued that, once individual characteristics are taken into account, regional context is unimportant. Supporters of the regional divide have pointed to the connection between economic conditions and voting, but have been unable to show direct links between local economies and voters' decisions. The paper employs new data to examine the controversy. Voters at the 1992 General Election are shown to judge their regional economies, and to act on those judgements, even when their personal positions are controlled for.

The relationship between voters' support for the government and thirteen items tapping their evaluation of the economic situation is analyzed. One data set is collected under Conservative rule and another under Social Democratic rule. The answers are first aggregated into three factors. Two are egotropic: the economic situation of the family, and its unemployment situation. One is sociotropic: worry about the economic development in the country. The impact of the first is strongly significant, and greater than found in comparable studies. The impact of the second is mostly significant, while the impact of the third is the least reliable. We have thus found a clear case of pocketbook voting.

Have the meanings of ‘left’ and ‘right’ changed during the last twenty years? In this article the ten-point left-right self-placement scale is correlated with three central value orientations (religious/secular, economic left-right and materialist/post-materialist values) to examine whether associations between these value orientations and the self-placement scale have changed from the early 1970s to 1990. Four theories about the changing meaning of the left-right language are presented. These theories about the irrelevance, persistence, transformation and pluralisation of the meaning of left and right are tested by using Eurobarometer data from eight West European countries and the second wave of the European Value Study from 1990. The data provide strong support for pluralisation theory. Left-right semantics have an impressive absorptive power, describing an over-arching spatial dimension capable of incorporating many types of conflict. Left-right semantics are significantly correlated with religious/secular values, remain highly correlated with the dominant industrial value orientations (economic left-right values), and are increasingly associated with materialist/post-materialist value orientations. The new meanings of left and right are added to the old meanings.

The aim of this study is to assess the relative strength of the reciprocal causal relationships between the political agenda (the party agenda), the mass media agenda and the public agenda. Although the research literature is rather confusing and inconclusive, three causal patterns have often been suggested. The economic theory of representative democracy (‘public choice’ theory) assumes bottom-up agenda-setting. The political agenda is assumed to respond to the public agenda. However, theories on political communication suggest top-down agenda-setting. The political agenda would set the media agenda, which in turn would set the public agenda. The central tenet of mediacracy theory is the proposition that the media agenda sets both the public agenda and the political agenda. This article uses data on economic issues in the Netherlands in the period 1980–1986 and linear structural equations models to test these three causal patterns. The results warrant both bottom-up and top-down agenda-setting, but the mediacracy model is rejected. The conclusion should be that the economic theory of democracy, which prevails in the political science journals, has to incorporate findings from political communication research.

This paper reviews the methodological issues raised by different commentators' attempts to categorise countries' social security systems into ‘regime types’ or ‘models’. It is argued that there are two major problems with these categorisations: (1) the difficulty of finding empirical counterparts to the theoretical concepts used; (2) the complexity of actual social security systems, arising in particular from the diversity of benefits within each system. Despite these problems, we argue that it is possible to maintain a theoretically-informed approach to the comparative analysis of social security, by adopting a more flexible methodology. The approach presented in this paper identifies some of the main principles on which social security systems are based, specifies their empirical counterparts, and allows each national system to be analysed in terms of its particular conjunction or combination of principles.