To what extent can the dramatic differences in social policy efforts in rich and poor countries be accounted for by genuinely political explanations? The hypothesis that is advanced in this article rests upon the combination of two schools of thought in comparative social policy analysis: socioeconomic models which focus attention on levels of economic wealth, need and demand for social security, and models of a comparative-historiographic and political-institution-alist nature. Empirical applications of socio-economic models with lagged dependent variables reveal the existence of two deviant families of nations: overspenders in social policy (such as overspenders of socialdemocratic complexion and of Christian democratic composition) and underspenders (such as the superpowers in East and West as well as Japan and East Germany). The residuals that can be derived from these models are amenable to an explanation which resides in comparative-historiographic political analysis of social policy.

The Reform Theory and the Political Economy Theory postulate contradictory effects of government size on citizens' satisfaction with urban services. The former asserts that citizens' satisfaction increases with increasing size of urban governments because large units are more efficient and allows citizens to participate effectively in public policy-making. The latter postulates that citizens are more satisfied in the smaller jurisdictions because small units are more homogeneous, efficient and democratic. A series of tests performed in this study overwhelmingly supports the Political Economy Theory: citizens in small jurisdictions hold more favourable attitudes towards participation and democracy, and the smaller units are more homogeneous and more efficient in the provision of services. This in turn leads to more favourable evaluations of public services.

The article examines approaches to the study of the party-policy link and argues that two generally held assumptions - the unidimensionality of party political conflict and the party as unitary actor - should be treated as matters for empirical dedermination rather than a priori assumption. Having briefly reviewed interpretations of dimensionality in the Irish party system, the article puts forward an empirical approach to both issues based on multidimensional scaling of preference data from a sample of parliamentarians. The results of the application of such an approach in the Irish case indicate at least a two dimensional policy space and considerable intra-party factionalism. This evidence is then used to interpret the party-policy link in three areas: capital taxation, abortion and divorce, and Northern Ireland policy.

An overview of the literature on the party-policy link suggests a weak, non-existent or even perverse relationship between parties in government and governmental policies in Belgium. It is argued that this weak relationship could be due to the methodology applied in this type of research, i.e. the treatment of the party variable on a much too high level of aggregation. The relative strength of the coalition parties, the parties’ general ideological profile on a specific moment, the relative strength of the factions within the parliamentary party and in the Cabinet are not taken into account. On the other hand, several organisational factors with regard to the relation between parties, parliamentary groups, ministers and the administration tend to facilitate the impact of parties on policies in Belgium. Interview data with ministers suggest that the policy variable also needs further disaggregation. In particular, the qualitative aspects of policies should be more closely investigated.

The literature on party government, coalition formation, and links between party and policy tends to assume that parties in government command legislative majorities that can be used to enact desired policies. This assumption, however, does not apply in general. In Scandinavia, and especially in Denmark, the minority type of government is predominant. Minority governments cannot govern by means of their own votes. Two questions for research are therefore obvious. Why do Scandinavian countries deviate from the normal pattern of government formation? And what are the consequences of minority government for policy making? While the first question has been treated in recent research, this paper carries the analysis a step further by exploring the policy consequences of minority government in Denmark during the 1980s. It is shown that the government may in fact not always govern, that the government may actually be the opposition, and, consequently, that the party-policy link can indeed be extremely complex.