The autonomy of a state is thought to depend on the state's structures. This widely accepted view of the policy-making process has given rise to an uncontested assertion: states that possess centralized administrative structures are better able than decentralized states to formulate and implement policies independently of societal pressures.

Is a decentralized state more permeable than a centralized one? We seek to answer this question by treating it as a hypothesis. The empirical material derives from a case study of a reform project proposed by the French state and objected to by the group in question, the notaires.

Mrs Thatcher's decisive and determined stand during the Falklands crisis in 1982 has been widely credited with restoring the electoral fortunes of the Conservative party in the run-up to the 1983 general election. This article argues that the Falklands war produced a boost to Conservative popularity of at most three percentage points for a period of only three months. Government popularity was already accelerating as a result of macroeconomic factors before the outbreak of the Falklands crisis, in particular ‘personal economic expectations’ proved to be of critical theoretical and empirical significance, and can be modelled satisfactorily on the basis purely of objective macroeconomic indices. Thus macroeconomic factors were at the root of the revival of Mrs Thatcher's political fortunes, and most of the boost to government popularity which occurred in the spring of 1982 derived from intelligent (or cynical) macroeconomic management. The Falklands crisis merely coincided with a jump in government popularity which would have occurred anyway in the wake of Geoffrey Howe's 1982 Budget.

At the heart of the feminist theory of consciousness-raising is a very precise hypothesis about the conditions in which most women – and members of any other socio-political out-group – will overcome their socialization into the culture of the dominant in-group to acquire political consciousness. The hypothesis is that separate interaction directly among themselves in autonomous, all-female groups will lead women to develop a new consciousness of women as a political category with interests distinct from those of men. This article uses new data about local women politicians in Scotland to test the hypothesis that there will be a strong, symmetrical and independent relationship between a woman politician's political orientation towards women and her experience of separate interaction. This relationship holds good for experience of any kind of separate interaction, even if it is confined to groups of an entirely non-political character, thereby confirming the causal inference that politicization is the consequence of separate experience and not ils precondition.

There are few systematically comparative cross-national studies of urban policy or service provision, partly because there is little in the way of empirical testable theory that might guide research and partly because the variety of institutional methods of providing and financing urban services in different nation states makes it difficult to carry out such comparisons. Urban systems theory, however, applies to all levels of urban society from local collections of towns and villages right up to the world system of cities. It also enables the development of empirically testable hypotheses linking the hierarchical system of cities with the level of expenditure on public services in metropolitan areas. An examination of data for seventy-six metropolitan regions in France, eighty-three in Italy, seventy-four in West Germany, and sixty-three in England and Wales supports the hypothesis. The article concludes that urban systems theory offers a theoretically well-developed and empirically powerful means of carrying out systematic cross-national comparative urban research.

The importance of political parties for inter-governmental differences in public policies has long been the subject of dispute. Given their comparative intentions and their concern with end-products rather than decision processes, statistical analyses of policy outputs should provide a valuable tool in evaluating this question. This they have seemingly failed to do, largely because they have devoted insufficient attention to the conceptual and theoretical underpinnings of their investigations. However, with Sharpe and Newton's Does Politics Matter? a theoretical approach has been introduced into the local outputs literature which should herald a new era in comparative statistical investigations and the welcome emergence of conceptual and theoretical advances. In addition, their efforts raise the promise of forging a closer alliance between the outputs literature and more detailed, qualitative investigations of policy-making processes in individual local authorities.

There is a generally accepted view among academic observers of British politics that no strong relationship exists between the ideological position of a party's candidate and the vote-winning capacity of his or her party. This is said to be a reflection partly of the electorate's lack of knowledge of particular candidates and partly of the strength of party allegiance. Political developments in the 1980s, however, suggest that there is an a priori case for re-examining this accepted view, at least as far as the Labour party is concerned.

In a recent publication Reece proposes to substitute for the ‘swing’ method what he labels a proportional-loss model. Although his argument against the logic of the swing method is perfectly valid, this short Note will demonstrate that what he offers as a substitute is just as unsound intellectually.