Read an Excerpt
Crosses on the Ballot
Patterns of British Voter Alignment Since 1885
By Kenneth D. Wald
PRINCETON UNIVERSITY PRESSCopyright © 1983 Princeton University Press
All rights reserved.
The Context of the Study
... what sort of divisions can be found in British society which provide the basis for partisan opposition? — Leslie Lipson
This study of the pattern of voter alignments in Britain since the late nineteenth century is directed to a pair of topics which have long commanded the attention of political sociologists. First, it explores the relationship between social structure and voting patterns in a mass electorate. The goal is both to specify the various social formations which achieved political relevance after 1885 and to identify the mechanisms which translated social divisions into lines of partisan cleavage. Beyond a static portrayal of mass political behavior, the study has a second aim — to enhance understanding of the dynamic properties of the British party system. It attempts to explain why certain social divisions, once of considerable electoral significance, became less important over time until they were superseded altogether by new patterns of sociopolitical conflict. Despite its concentration upon a single political system, the study thus addresses problems which Lipset and Rokkan regard as "fundamental questions for comparative research."
Most of the material for the book has been drawn from an intensive study of electoral patterns between 1885 and 1910, dates which mark the first and last general elections held under the provisions of the Third Reform Act. Though not so neglected as it once was, this stretch of years remains something of a dark age in terms of electoral analysis. The British party configuration of this period has not been subject to the same careful scrutiny as the party system of the middle nineteenth century, which has been studied profitably through analysis of the rich store of data contained in pollbooks, the published records of individual votes at general elections. Nor has the party system between 1885 and 1910 received anything like the attention devoted to the era which followed it, a period for which the full resources of modern electoral analysis have been deployed. The relative neglect of the late Victorian-Edwardian party system is unfortunate because it was during this period that British mass politics first assumed a recognizably "modern" format.
During the period of the Third Reform Act, elections acquired their decisive modern function as the major mechanism linking the actions of the rulers with the wishes of the ruled. The fate of a Government, which had depended primarily on its ability to survive confidence votes in the House of Commons, came instead to depend upon the support it could command in the cities and counties at a general election. As the concept of the popular mandate gained legitimacy, elections were treated as referenda upon current issues, and the distribution of the vote was taken as a measure of public reaction to party policy. To a greater degree than ever before, the major issues of the day — Irish government, imperialism, free trade, the role of the Lords, the status of the Church — were debated with reference to electoral considerations. This development had its parallel outside Westminster as constituency life, formerly dominated by the peculiar features of the local community, responded more fully to national influences. These influences were channeled through political parties which increasingly functioned as agents of mass electoral mobilization. For leaders and followers alike, in other words, the votes counted.
The institutional structure for counting votes, the electoral system, also appeared in modern guise during the period. The scope and influence of electoral corruption and "influence" were lessened appreciably by the adoption of the secret ballot (1872) and the passage of corrupt practices legislation (1883). The old constituencies, based on a series of medieval communities which had long since ceased to correspond to social reality, were replaced by new election districts (1885) which more faithfully mirrored population distribution. The reforms embodied in the Franchise Act of 1884 led for the first time to a rough semblance of universal manhood suffrage. As recent research has emphasized, the combined effect of these changes still left a great many Britons outside the pale of the Constitution. Nevertheless, whatever the considerable defects and anomalies remaining in the electoral system, by 1885 "the word 'Reform' no longer naturally denoted 'Electoral Reform.'" The electoral process, formerly an institution for reformers to attack, had now become an avenue of attack.
The period under discussion also witnessed the modernization of another important element in British politics, the party system. Political parties replaced factions built around notables as the accepted nuclei for organizing a Government within Parliament. This development inevitably affected the individual Member of Parliament, downgrading him from an independent center of authority and criticism to a much diminished role as defender of the party leadership. Divisions within Parliament accordingly followed party lines to a much greater extent than before. In another respect, the partisan environment of Westminster took on a marked resemblance to the contemporary pattern. Then, as now, two parties divided the bulk of the popular vote, but they frequently had to rely for their parliamentary majority upon tacit alliances with assorted minor parties. It was during this period, moreover, that the three major parties of modern British political history — Liberal, Labour, and Conservative — first achieved representation in the same Parliament. As questions of distributive justice began to force their way onto the political agenda, the partisans staked out positions that sound familiar to the observer of British politics in the post-1945 setting.
For all these reasons, then, the period of British politics bounded by 1885 and 1910 merits sustained study. It presents an opportunity to examine problems of major interest to students of voting behavior and to do so in a broader historical context than is customary. The study is further warranted by the inability of scholars to achieve anything approaching consensus on the salient features of political life during the period. Indeed, as scholarly interest in the period has deepened, disagreement has intensified to a point where virtually every claim or generalization generates a counter-argument. This description is particularly apt for the two main topics addressed in the study, the contours of electoral cleavage and the source of long-term change in party fortunes.
Contemporary observers seem to have treated political parties not as collections of voters sharing a considered attachment to basic political values nor as floating masses of individuals responding like Pavlovian dogs to the bribes offered by party leaders. Much like modern political scientists, they recognized that party coalitions were often based on shared social characteristics and fortified by the dead hand of tradition. As one MP wrote,
To many persons party symbols and party associations have taken the place of all party meaning [so] that to vote "blue" or to vote "yellow" has become the traditional practice in many families. ... The answer of not a few to all solicitations is "my father always voted 'blue' (or 'yellow') and so did my grandfather, and so shall I."
What groups were significant in binding voters to their parties? According to the conventional wisdom expressed in many modern accounts of the period, religious or denominational differences provided the basic line of political demarcation. The electoral arena is portrayed in these accounts as a context in which the protagonists fought out battles which had their origin in Reformation-era conflicts. The kinds of issues which provoked partisan conflict, it is argued, have a quaint tone when compared to the characteristically modern rhetoric of class conflict.
Other scholars reject this religious interpretation, however, and describe party competition under the Third Reform Act as a relatively straightforward clash of economic interests. The proponents of this alternative perspective regard the pre-1918 party system as a somewhat more primitive version of the class-stratified configuration that has characterized post-1945 Britain. Perhaps, to consider yet another logical possibility, these contrasting views can be reconciled. If class and religion overlapped to a considerable degree, it would make sense to treat them not as competitive sources of cleavage but as mutually reinforcing influences on voter behavior. We must also entertain the possibility that a variable which has not thus far figured prominently in British historical election analysis — such as region — exercised a considerable impact on voter alignments. The merits of these competing claims can be assessed in part through a systematic analysis of voting patterns — a major task of this study.
If the nature of electoral stability under the Third Reform Act has proved so difficult to gauge, the problem of change has generated even more disagreement. The rise of the Labour Party at the expense of the Liberals, a trend first suggested during the period, has been described by one reviewer as a subject likely to replace the rise of the gentry as the foremost problem in modern British political development. This transformation of the party system has been attributed to all manner of forces: changes in the basis of electoral cleavage, bad tactics by Liberal strategists, the increasing structural differentiation of industry, reapportionment of constituencies, the growth of secularism, the extension of the franchise in 1918, wartime disagreements among the Liberal leadership, etc. These factors are not all equally amenable to verification with the methods favored here; but because many of them are based upon unproven claims about voting behavior, the systematic analysis of electoral data from the pre-1918 period may contribute significantly to clarifying the debate.
These empirical disputes among students of British political development tie into some broader theoretical issues that concern specialists in political sociology, voting behavior and political parties. By paying due regard to these issues, an intensive study of Britain may contribute to theoretical development in the field of political behavior. For example, the social analysis of voting has spawned many useful insights about the process by which social divisions impinge upon voting and considerable speculation about hierarchies of cleavage. The British data offer an opportunity to refine and test some of these insights for a period for which survey data are unavailable. The general phenomenon of change in party systems can also be advanced by examining trends in British voting. The electoral history of many nations seems to be marked by an alternating series of stable party configurations followed by an abrupt period of electoral discontinuity in which stable allegiances are disrupted and new party coalitions emerge. The periods of electoral stability, which have been described variously as "party systems" and "sociopolitical periods," tend to exhibit a characteristic political agenda, a high degree of persistence in mass voter alignments and a relatively fixed distribution of the vote. In explaining why such stable systems seem to give way to rapid decomposition and reformulation, researchers have speculated about the role of factors such as new political issues, demographic changes, alterations in the legal-institutional structure of elections, and generational displacement in dissolving a seemingly stable pattern of political conflict. The process of electoral change in Britain can thus be regarded as another potential incidence of a recurring phenomenon and be treated as a "case" in the study of the transformation of party systems.
One problem with the existing literature, as we have just seen, is that scholars have given radically different answers to a number of basic questions. Such disagreement alone warrants further study of the period. Equally important, existing studies have failed to capitalize fully upon modern techniques of electoral analysis. To remedy that omission requires a new kind of study with methods suitable to the analysis of mass politics.
Without doing too much violence to their unique qualities, previous voter studies of the period can be placed into four categories: single election studies, works of electoral geography, intensive local profiles, and, a more recent development, historical survey research. Four of the eight general elections fought during the period have been the subject of monographs similar to the Nuffield College series on post-World War II British elections. These single election studies characteristically examine the record of the Government that called the election, its fortune in by-elections, the kinds of issues raised during the campaign, the tactics of interest groups and candidates, and the general pattern of the returns. In the typical works of electoral geography which were popular early in the twentieth century, the authors mapped the results of elections and commented upon the spatial distribution of party support and the relationship between party vote and parliamentary representation. Recent work in the genre has utilized more sophisticated techniques to explore the relationship between constituency voting patterns and aspects of social structure. It is difficult to characterize works in the third division of historical voting research, intensive local profiles of particular areas or constituencies. The questions which motivate such studies are often quite specific to the locale under investigation, which might range from a single constituency or city to a large region. With some notable exceptions, the authors of these studies use the traditional techniques of narrative history. In recent years, scholars have attempted to harness the techniques of public opinion research to historical inquiry. By probing the memories of the elderly during in-depth interviews, scholars have helped to elucidate features of the period which were evident only to persons who witnessed events for themselves.
Each of these approaches has made a distinctive contribution to furthering our understanding of political life between 1885 and 1910; the present study, which is built upon its predecessors, will incorporate such techniques where appropriate. But this survey attempts to overcome the limitations of scope and method which characterize the existing body of research. First, the investigation is comprehensive in its coverage of the period. Unlike the single election studies, and most local surveys and geographical analyses, it deals with voting behavior over the entire twenty-five years of the period and, on occasion, extends the analysis to cover developments well beyond 1910. This scope is essential if the study is to pay proper regard both to the durability of party alignments and the forces working for change. The scope is further broadened by the attempt to examine developments throughout all of Great Britain. Rather than allow local peculiarities to dominate the inquiry, an attempt is made to discover patterns with more general applications. Second, the study takes advantage of modern advances in the ecological analysis of voting behavior. These new methods certainly do not rule out impression and insight nor do they provide a set of facts which mysteriously "speak for themselves." But modern quantitative analysis provides more explicit standards of inference than conventional historiography and permits systematic comparisons of the power of various explanations. A study with these features should advance our knowledge of the period and provide conclusions which are themselves available for additional testing and further development.
The centerpiece of this work is a quantitative analysis of the relations between various social forces and the patterns of party support in general elections from 1885 to 1910. The units of analysis consist of sets of parliamentary constituencies that were combined until their boundaries corresponded closely to the units for which census data were enumerated. From the census material and miscellaneous sources, we have extracted tangible measures of the social forces which have been cited as important factors in structuring the aggregate vote. By using a variety of statistical techniques, primarily regression analysis, we can measure with some precision the predictive power of the social variables and the extent and manner of their relationship to the political variables.
These particulars do not apply to all phases of the analysis. In some instances, the political variables were derived from the results of local government elections, and some of the social variables under investigation had to be collected from data sources which went far beyond the bounds of 1885 and 1910. A more important limit was my recognition that "electoral data are only the outward and visible sign of an invisible political situation which has to be intuitively apprehended in the light of many variables, not all of which are numerical ones." Therefore, to make sense of the statistical patterns which emerged from the quantitative analysis, we will take advantage of the techniques associated with conventional historiography. Thus, the findings of intensive studies in selected locales are cited if they illuminate a problem. Similarly, illustrations and examples are brought forward in support of arguments that cannot always be verified in a manner that satisfies the strict standards of empirical social research. In an area so open to investigation, an area in which the proper questions have not yet been identified nor the necessary concepts specified with any degree of precision, the combination of traditional historical method with the techniques of modern social science seems the most promising avenue for advancing knowledge.
Excerpted from Crosses on the Ballot by Kenneth D. Wald. Copyright © 1983 Princeton University Press. Excerpted by permission of PRINCETON UNIVERSITY PRESS.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.