Explaining the foreign policy behaviour of states has proved a particularly difficult task for theorists of international relations. For centuries it relied on an analogy between states and individuals in the state of nature, so that an endemic tendency to international anarchy resulted from states having ‘interests’; systemic, determinist theories could therefore explain foreign policy by appealing to such notions as national interest and power maximization. The elusive, contestable character of these notions later led many analysts to focus on the empirical decision-making process for explanations of foreign policy behaviour. Yet these attempts have run into a fundamental problem: the proper weight to be attached to the perceptions and reasons of the actors. Some of the literature takes the actors very seriously and relies either on a psychology of perception or on a decision-theoretic model of individual choice. Some of it, on the other hand, by-passes the actors altogether and concentrates on such structural features as bureaucratic position. In this article we shall argue the case for a concept of role, requiring a less mechanical view of action than the standard approaches allow, both separately and in combination.

The current American debate about the relationship between business and government represents the most significant reopening of that issue since the New Deal. The debate is in part about government's role in the economy, but the issue of business's role in politics is being joined as well, joined in fact on several fronts. There are, of course, the polemics of corporations and their critics, in which business is cast alternately as victim and villain. The issue also divides more serious students of American politics and has fostered a wealth of theorizing about the role of the state. Finally, the issue of business influence pervades discussions about campaign finance.

Local government borrowing is often referred to as a highly technical, not to say arcane, activity of local government. It should not be dismissed so easily. In reality, such borrowing illuminates broader relations between government and market by highlighting interdependencies which exist between central government, local government and financial markets.

When streams of research are isolated from one another by methodological style and ideological character, the fact that they share similar explanatory principles is easily overlooked. Thus, since the 1950s many quantitative and ‘pluralistic’ American studies have argued that political leaders are more likely than the public to support procedural rules of the game. And since at least the 1930s, many qualitative and ‘left-wing’ European commentaries have argued that, in matters of socio-economic policy, members of parliaments become more moderate than their parties' activists. These important claims are embedded in two partial theories which have previously been treated as unrelated, the theories of institutional support and of deradicalization. And yet, different as these theories may be in many respects, they are driven by similar socialization principles which accompany movement from one role in the political system to another. Such socialization principles are a conservative force inculcating both institutional support in procedural rules of the game and deradicalization in orientations towards public policy.

The standing of the public choice approach to politics is ambiguous. Most people now accept that there is something in it, even if the ‘something’ is restricted to Arrow's Social Choice and Individual Values, Downs's An Economic Theory of Democracy, Olson's The Logic of Collective Action, and (perhaps) Riker's The Theory of Political Coalitions. The Logic, for instance, is a cult abbreviation propagated by Olson himself but accepted by others. Three new adjectives (Downsian, Arrovian and Olsonian) are now established, although I have not yet found ‘Rikerian’. By analogy (Keynesian, The General Theory), the language seems to denote acceptance. But the seminal works (they are always seminal, as dawn in the Odyssey is always rosy-fingered) came out between 1951 and 1965; so where are their successors? Why have they had relatively little impact on student texts and informed public opinion? In short, what has been going on?

A very great deal, in fact. The first generation proposed bold hypotheses; their successors have put enormous efforts into refining them and proposing new ones, and (unfortunately) rather limited efforts into testing them. There have been four main lines of development. It would be impossible to review all four critically in a single article; therefore I make only fairly brief comments on the first three and devote most of this article to the fourth, which I think is one of the most promising and of the greatest interest to political scientists.

‘Tactical voting’ refers to voting contrary to one's nominal preferences. The usual form of tactical voting described in the literature consists of ‘third’ party supporters in plurality electoral settings voting for one of the two major parties in their constituency. This Note aims to demonstrate both theoretically and empirically the existence of the converse phenomenon, i.e. followers of one of the two big parties voting for a small one. We shall call this phenomenon ‘inverse tactical voting’ (ITV).