Rational Switch

Vernon Bogdanor

  • Democracy at the Polls: A Comparative Study of Competitive National Elections edited by David Butler, Howard Penniman and Austin Ranney
    American Enterprise Institute for Public Policy Research, 367 pp, £5.75, March 1982, ISBN 0 8447 3403 9

It was R. B. McCallum who invented the word ‘psephology’ to describe the study of elections. Yet in 1955 he wrote of the act of voting as the last haven of free choice in an increasingly bureaucratised society, an ultimate redoubt to be defended at all costs against the assault of the social sciences. ‘The secrecy of the ballot, the pencilled cross in the secluded polling booth’, was, he said, ‘the great eleusinian mystery of the democratic state. It must be respected as an article of faith.’ Not surprisingly, political scientists have conspicuously ignored this warning. Indeed, as the introduction to Democracy at the Polls points out, the analysis of elections has become one of the main growth areas in the discipline: the behaviourist finds the quantitative methods needed to study elections peculiarly congenial; whilst the electoral process itself, whose rules and regulations vary little between one democracy and another, seems especially suited to comparative study. Analysis of elections also has a normative purpose. ‘The reality of liberty,’ claimed Ernest Barker, ‘consists in the details and the substance of actual institutions.’ Psephology can, therefore, be seen as the discipline that will give flesh and blood to the generalisations made by such classical writers on democracy as Mill and Bryce. For elections throw a searchlight onto the working of the democratic process in different countries, illuminating the assumptions upon which democratic government rests. General elections, to paraphrase Namier, are ‘locks on the stream of ... democracy, controlling the flow of the river and its traffic’.

The American Enterprise Institute, in possession of both intellectual and financial resources beyond those of its European counterparts, decided in the mid-1970s to initiate a series called ‘At the Polls’, whose purpose it would be to study the electoral process in 22 of the 28 countries classified as ‘free’ or ‘partly free’ by Freedom House, and with populations of over three million. Of these 28 countries, no less than 16 are in Western Europe, while six are members of the Commonwealth. Democracy at the Polls is an attempt to establish generalisations from the volumes studying individual countries. What does the analysis of elections teach us about the nature of political choice, or about the impact of different electoral systems upon party politics? How are representation and accountability actually secured through the institutional procedures of modern democracy? And, above all, can modern political science solve the riddle first posed by Oliver Cromwell, who was ‘as much for government by consent as any man, but if you ask me how it is to be done, I confess I do not know.’

Alas, Democracy at the Polls, although containing contributions from the leading students of electoral politics in Europe and the United States, reveals all too clearly the poverty of political science as an explanatory discipline. For the skills of the political scientist seem to be displayed more in the construction of classificatory typologies than in explaining the processes of democratic government. What has gone wrong? In part, the contributors to this book have simply been unlucky. The late 1970s, when the ‘At the Polls’ studies were published, was a period of turbulence and change, when assumptions about party politics were shown to be without foundation.

In their seminal work on party structures and voter alignments, S. M. Lipset and Stein Rokkan showed that the basic electoral cleavages in Western Europe were formed at the time when universal suffrage was conceded, and had remained remarkably stable from the 1920s till the mid-1960s. Electoral geographers such as André Siegfried showed the profound historical continuity of voting patterns in France: for example, a Département hostile to the Revolution in 1789 would be likely to support a candidate of the Right a hundred and fifty years later; while a Département sympathetic to the Revolution would be found supporting the Popular Front in 1936. Even the ravages of war were unable to alter deeply-rooted habits of electoral behaviour. When the first elections in the German Federal Republic were held in 1949, it was discovered that patterns of support for Christian, conservative and socialist parties mirrored almost exactly the voting patterns of the last free elections before Hitler. In Italy, despite political instability and an electoral system encouraging the formation of new parties, 92 per cent of the seats in the Chamber of Deputies in 1979 were occupied by Deputies belonging to parties present in the 1946-8 Constituent Assembly; and these in turn all had their roots in the political traditions of the pre-Fascist era. Parties without such deep roots had a record of almost unredeemed failure.

Since the mid-Sixties, however, the party systems of most of the countries of Western Europe have been in flux. New parties, many of them of the ‘protest’ variety, have mushroomed: the SDP in Britain, D 66 and DS 70 in the Netherlands, the Centre Democrats and Mogens Glistrup’s Progress Party in Denmark, Christian parties in Scandinavia, the Greens in France and Germany; while ethnic politics have come to dominate Northern Ireland, Scotland and Belgium. Moreover, the nature of the electoral support for the more traditional political parties seems to have changed. Instead of being able to rely upon solid blocs of support, parties in countries such as Britain and the United States seem to be faced with an increasingly volatile and sceptical electorate. For this reason, it is by no means clear whether the growth of new parties in Western Europe heralds a process of realignment – such as occurred in Britain in the Twenties – or a process of dealignment whereby parties cease to be able to rely upon particular socio-economic groups for regular support. Does the SDP in Britain signify, as its leaders on the left such as Shirley Williams would like it to, a realignment amongst the forces of progress; or does it, on the contrary, signify the growth of an ‘anti party’, a party of protest supported by the electorate as a symbol of distrust of party politics and of the political system in general? It is not possible yet to answer this question, nor the similar questions which can be asked in relation to the new political movements on the Continent, and it is doubtful whether the methods of political science displayed in this book are capable of resolving them.

The study of elections, as it developed during the Forties and Fifties, seemed to cast doubt upon many of the hopes of liberal democrats, and to confirm that disenchantment with the workings of liberal democracy which informed the writings of social scientists such as Michels and Max Weber. Democracy worked as much because of the deficiencies of the citizen as because of his virtues. The voter, far from conscientiously weighing up issues, made his voting decision on the basis of class or family allegiance; he identified with a political party on what the liberal theorist was bound to regard as irrational grounds. Worst of all, the floating voter – the voter who actually decided the outcome of elections – was less, not better, informed than his more stable counterparts, in that he was more likely to be swayed by ephemeral prejudices or the irrationalities of the election campaign. That voting behaviour was rooted in socio-economic factors became part of the folk wisdom of political science, as did a certain scepticism as to the virtues of the democratic electorate. Some political scientists, indeed, erected the socio-economic basis of voting behaviour into a precondition of successful democracy. Democratic government relied, not upon the electorate’s appraisal of policy issues, but upon the individual’s ties with his class, religion, subculture or famille spirituelle.

It is doubtful whether the voting behaviour of democratic electorates will lend itself to this kind of explanation for much longer. Studies of the basis of support both for the Liberal/SDP Alliance in Britain and for Reagan’s Republicans in the US have shown how difficult it is becoming to correlate electoral support with socio-economic indicators. Indeed, the nature of the Liberal vote in Britain has long frustrated political scientists, precisely because it seems resistant to explanation in socio-economic terms. For this reason, it has been regarded as fundamentally unsystematic and unpredictable in nature, and has offered splendid opportunities for political scientists to project their own classificatory schema upon reality.

There is increasing evidence, however, that support for the Liberal Party, and perhaps for the new movements in the Western democracies, is the product of a new concern with political issues. For, although the Liberal vote is difficult to explain in socio-economic terms, it can be explained in terms of negative issue voting. The voter uses his vote not necessarily to express a diffuse protest against the political system as a whole, but rather to express specific policy disagreements with the party which he had hitherto been accustomed to support. For voting behaviour is in general retrospective in nature, a verdict on the past rather than a hope for the future. To the extent that voting behaviour expresses the attitudes of the electorate to specific policy issues, the pessimistic view of the individual voter will have to be revised; and, indeed, recent studies of electorates in Western democracies have revealed an increasing degree of knowledge and sophistication compared with the ignorance and apathy that prevailed in the Fifties. Voters, on the whole, make their decisions on rational grounds. They are fully aware of the issues at stake, and aware also of the positions which the parties take on the issues. Nor are the floating voters more ignorant than those who remain with the same party for election after election. On the contrary, the floating voter is now shown to be more knowledgeable and informed about politics than his more conservative counterpart. It is the rational switcher who conforms to the liberal democratic model of voting behaviour, not the party voter, whose appraisal of the political situation is likely to be fairly lethargic.

Attitudes to politics reflect generational changes. The generation studied in the Fifties was the first which had grown to maturity during the era of universal suffrage. It is hardly surprising that its attitude towards politics was marked by a cautious conservatism and its voting behaviour characterised by deference to the traditional landmarks of family, class and religion. The current political generation, however, takes universal suffrage for granted, is sceptical of the pretensions of political parties, and disdains the crudity of the rhetoric used to cement party ties. Few practising politicians in the West have been able to come to terms with these changes: equally, the majority of political scientists have been unable to appraise their likely consequences. In a world in which socio-economic factors cease to be accepted as sure guides to political behaviour, the political scientist will have to develop more respect for the attitudes of the electorate, and a more sensitive grasp of the political context within which voters make their decisions. He will have to become more wary of crosscountry classifications and comparisons between societies with different historical backgrounds which conceal more than they explain. Those political scientists who have displayed the most insight into the changing nature of democratic government – scholars such as S. E. Finer, Stein Rokkan and Luigi Farneti – have been steeped in the history of the democracies which they have studied. For democratic elections are an expression of the internal political relationships of a society. They can be understood only as indicators of a society’s view of itself and of its expectations of the future. To make sense of these changes requires less the techniques of the behaviouralist than the sensitivity of the historian, and perhaps the creativity of the novelist. Democratic electorates continue to elude those who seek to encapsulate their activities in the formulae of the social sciences. At the heart of democracy there does indeed lie an act of faith.