This paper sets out to consider two related questions: the significance of elections in single-party systems in underdeveloped countries, and the kind of study which political scientists have made of such elections.1

Studies of medical politics usually emphasize one of the following types of inquiries: (a) analyzing the internal politics of medical organizations, as with Oliver Garceau's classic study of the American Medical Association; (b) describing and explaining the roles individual physicians play in the political life of the community as voters, officials, or citizen participants in civic life; or, (c) assessing the impact of medical groups and organizations on public policy, particularly health policy. Harry Eckstein's widely known study of the British Medical Association is primarily a study of the third type, a discussion of the channels of influence, the tactics, and the effectiveness of the BMA in shaping public policy to its ends.

French politics have been undergoing a process of restructuring since the coming of the Fifth Republic in 1958. This process, like any process of political change, has been a complex one involving many factors and variables. The most obvious of these factors has been the one most commented upon by journalists and contemporary historians – the influence of General de Gaulle and his policies (the new constitution, foreign policy, etc.). What is only beginning to emerge from the academic debate is the relationship between Gaullist politics and the other variables.

One of the issues that has dominated American political science in the last twenty years has been the debate between those who assert that America is ruled by an elite, and those who believe that the pluralist model is a more accurate description. Robert Dahl, who is the most influential of the pluralists, has attacked ‘elitists’ on two fronts: negatively, he has argued that they misperceive the nature of power and are sloppy in their criteria for determining a ruling elite; positively, in Who Governs?,he has attempted to apply his own conceptualization of power to an American community and thus give empirical backing to the pluralist model.

Political science or political studies or government is a rather odd subject, something of a mixture, as indicated by its diverse origins and the various influences which have moulded it. In our universities it was affected by and grew out of the study of the classics, philosophy, law and jurisprudence, economics, history, geography, literature, and so on. The list of pre-existing and established academic subjects which acted as a sort of hydra-headed godfather is considerable. An important impulse in its development was provided, too, by ideology and practical purpose. I have in mind here the ethos represented, for instance, by the Webbs and the foundation of the LSE, by the teaching of people such as Cole, Laski, Hobhouse, Tawney and, more remotely, Green and Jowett. The idea was that the subject was one of considerable pragmatic value and offered practical advantages to be achieved, perhaps, by some sort of reforming doctrinal emphasis. But the general effect of this variety of influences and stress is that there is no obvious or uniform academic focus: the subject seems (in terms of the intellectual and other forces at work) essentially eclectic. Nevertheless something like a traditional form did emerge. Generally there was a distinction between ‘institutions’ and ‘theory’, though naturally, given the dissimilar sources and influences mentioned, what has been comprised by these titles has varied somewhat. ‘Institutions’, for example, could cover a concentration on the formal machinery of government simply or it might expand to embrace parties, groups, the formation of opinion and so on.

Power has been so extensively analysed and so variously defined that a raw student (or even a seasoned practitioner) who looks at a dozen or so definitions of power might well wonder about the nature of the enterprise. The situation seems to be that there is some agreement on what power is — the word might otherwise be expected to have fallen into disuse — and some very sharp disagreement. But one is at something of a loss to know, firstly where the agreement is, secondly what is at issue between rival definitions, and thirdly which disagreements are basic and which mere refinements. A new analyst might well be puzzled about where to begin; and I rather think that all those who analyse power must be quite uncertain about what it would be to succeed in their task.

National stereotypes have an understandable attraction, but are often a snare and a delusion. There has been no lack of commentators, in recent years, to remind the French that their structures of local administration were laid down at the Revolution, and have been little changed since. Fragmented into nearly 38,000 urban and rural communes, which vary enormously in population and wealth, the country has more local authorities than the other five states of the EEC and Britain put together. Most people, on both sides of the Channel, accept the idea of reform, but we find something of a contrast when we look at what has happened in practice. President Pompidou said at Lyon on 31 October 1970:

Local government politics, particularly in urban areas, is now dominated by the major national parties though the process and causes of the take-over are not well documented. These notes, based on research in Doncaster, relate the process of the take-over of the anti-Labour position by the Conservative party and draw some possible conclusions.

In his brief note on turnout and marginality in a recent issue of this journal Newton [K. Newton, ‘Turnout and Marginality in Local Elections’, II (1972), 251–5], raises some interesting questions. In our view, however, the boldness of some of his conclusions is not warranted by the evidence he presents or by his supporting argument. In particular, we think three points deserve comment.

I shall deal with the points made by Denver and Hands in the order in which they raise them.

(1) Since my article tries to show that the relationship between marginality and both turnout and changes in turnout is, at best, a faint one, the assumption that ‘parties are more efficient in, or put greater effort into, marginal wards’ is in no way necessary to the argument. On the contrary, the assumption seems to be characteristic of those who claim a relationship between marginality and high turnout.

Denver and Hands point to accumulated evidence that concentrated efforts can have a dramatic effect on turnout. It would be foolish to deny that this can be the case or has, indeed, been the case, but the figures seem to suggest that the effect is usually small to insignificant. Most of their British evidence is based upon general not local election results. While there is clear evidence to show that turnout increases when a constituency becomes marginal (Brighton, Kemptown in 1966 is the classic case), the evidence about local elections is muddled and inconclusive, and has often been used to support conclusions which do not follow.