While all industrial nations have been plagued by the setback to economic growth since the mid-1970s, a wide range of variation marks the economic performance, policy outputs and policy outcomes of Western nations. The basic question addressed in this paper is the extent to which economic, socioeconomic and political variables account for the differences in economic performance, unemployment, public debt and the growth of social security expenditure. In general, the forces which shape public policies in periods of economic crisis tend to be different in character from the major determinants of policy-making in periods of prosperity. The analysis does not support the view that bourgeois and socialist governments produce clear-cut policy differences. It suggests that the major determinants of policies are the power relations in extra-parliamentary arenas, the level of national economic strength prior to the crisis, the extent to which ‘solidaristic’ values characterize the political culture, and the extent to which a correspondence exists between the power relationships in the political ‘superstructure’ and in the party system and industrial arenas. An expanding capitalist order is, in theory, not incompatible with low unemployment and a developed welfare state. However, the governments’ room to manoeuvre is strongly restricted by political and technological developments. One of the major political restrictions is the ‘paradoxical’ outcome of elections.

Existing theories about the causes of coups d'état are examined by multiple regression analysis with a dummy dependent variable, using cross-section data for, in turn, all independent countries and restricted areas. The exercise aims to demonstrate the value of the method in examining general hypotheses in their entirety as opposed to the usual piecemeal correlation analyses. The exercise also suggests the relative strengths and weaknesses of the theories examined. In particular it produces supportive results for Zolberg's theory but finds that it would be better applied generally than limited to the countries of tropical Africa.

Recent evidence suggests that significant changes in consciousness are occurring in Western democracies. This paper argues that in Britain, at least, the full extent and import of these changes can only be assessed by a panel-based mass survey which builds upon the ‘Political Action’ research of the 1970s. Such a survey should examine propositions derived from personal risk theory, social learning theory and from a more detailed examination of the individual's ‘objective’ environment.

The theory of the Silent Revolution relates value change to the process of population replacement. Materialist and post-materialist values are seen as reflections of the conditions that prevailed during one's pre-adult years. This assumption is tested here by analysing a Dutch panel, questioned in 1974 and in 1970. The change actually found at the individual level is not in line with the predictions of the theory and leads to a dilemma. The first option implies a rejection of the socialization hypothesis - one of the cornerstones of the theory. The second option places considerable doubts on the reliability of the instrument. Moreover, the validity of the instrument seems to be questionable too. The remarkably similar marginal distribution of the materialist and post-materialist value types in both years, accompanied by massive shifts between the types, points into the direction of political attitudes and non-attitudes instead of basic orientations or basic outlooks.

Jan Van Deth’s penetrating critique of the theory of value change presented in The Silent Revolution raises some key questions about the nature of value change and related measurement problems (see pp. 63-79 of this volume). It also presents significant new evidence about the individual-level stability of materialist/postmaterialist values, based on a recent Dutch panel survey.

An important insight emphasized by proponents of ‘politico-economic’ models is that economic policy is made by politicians in the context of political institutions. As a result, economic policy reflects the interaction between economic structures and decisions taken in this political context. Attempts to model this interaction formally should be encouraged, as we believe that purely ‘economic’ or ‘political’ accounts of economic policy give at best a partial view. We do not accept, however, that the only possible way of modelling economic policy in its political context is that offered by Frey and Schneider (1982). Our approach is also ‘politico-economic’, since it is based upon an explicit view of institutional behaviour.

The Frey and Schneider model is weak as a description of institutional policymaking. Moreover, their empirical work does not stand up to careful examination.

We feel pleased and honoured that Alt and Chrystal (1982) respond at such length to our contribution in this journal and elsewhere (Frey and Schneider, 1978; 1981; 1982). Without mentioning it, the two authors address themselves to a limited part of our paper only, the one referring to the United Kingdom, leaving out of account the models and empirical estimates for the United States and West Germany.

Alt and Chrystal claim to present a politico-economic model ‘rooted in actual institutional behaviour’. It is difficult to detect, however, any analysis of such (governmental) behaviour. To cite the intended actions of a committee at a particular period and for a particular country is no substitute for a serious study of what motivates and constrains politicians and administrators. Alt and Chrystal maintain that these decision-makers want to keep expenditures a stable share of anticipated national income, but no reason is given why they should desire to do so. ‘Explaining’ government expenditures by national income of the same period is rather uninformative, and does not allow us to make any true forecast of what the government is going to do, except ex post facto, when the level of national income is known.