76 Comparative Studies of Elections
to which minorities are safeguarded against oppression by majorities. The decline of cleavage voting in many countries can be regarded as an indicator of the resolution of conflicts, as the audit of decades of successful democratic government.
3.3. The Agenda/or Comparative Studies of Elections
Many, if not most studies of elections are not comparative in character. They may contribute to comparative research by providing empirical information for comparative analysts. Yet, in a number of ways the national character of many studies is also an obstacle. As Van der Eijk and Schmitt (1991: 260) remarked:
'..the national peculiarities and resulting incomparabilities of election studies in various Western countries have repeatedly curtailed the scope and plagued the execution of efforts to arrive at a comparative and more general understanding of the electoral process and its place in the operation of democratic political systems. The situation is somewhat of a paradox: while the field of electoral research is among the oldest, and certainly most developed areas of empirical social research, it has not generated the kind of large-scale cross-national survey projects which have been so successful in the development of other areas of comparative mass political behavior".
Important items on the agenda for future comparative studies of elections have been listed elsewhere by van der Eijk. Franklin et al. (1992: 427). Their major suggestions are the following:
1. research into the circumstances and processes which either promote or discourage the emergence, development or maintenance of systems of orientations that may serve to constrain both party and voter behavior after the decline of cleavage politics has run its course. Obviously, when cleavage-based interests lose their power to structure behavior of parties and voters, something else most take its place for elections to retain a meaningful communicative function in the interaction between elites and non-elites. Relatively stable orientations may develop on the basis of values, ideologies or personalities, but

it is equally possible that for some period of lime party systems and voters lose their 'moorings' and will drift in whatever direction they are propelled by unpredictable events. Research by, amongst others, Granberg and Holmberg (1988) suggests that it is of great importance to what extent political parties, other relevant elites and media offer the voters (and one another) a 'model' of political coherence. This, in turn, seems to depend on the structure of political parties where individual politicians' careers are made or broken. Parties which control to some extent the careers of politicians (as is the case in most European party systems) offer more opportunity in this respect than the much less-homogeneous American political parties which have little influence on the careers of individual political entrepreneurs.
2. research into the genesis of whatever relatively stable political orientations voters may hold. The cleavage perspective on politics suggests that relevant orientations were more ascribed than acquired, and that they had their origins in parental transmission, supported by homogeneous social settings and contexts. This suggestion does not hold good anymore once cleavages lose most of their erstwhile impact on voters. So, by what kind of processes do people acquire (he stable orientations which they often evidently have (e.g. Van der Eijk/Niemoller, 1992).
3. one of the consequences of the decline of cleavage-based voting is the increase in electoral competition between parties. What once were 'natural' reservoirs of support for parties have ceased to be so. This is even so in the presence of stable ideological or value orientations which guide voters in their party choice. Obviously, the degree and structure of electoral competition is an important determinant of what earlier was referred to as the course of politics. To the extent that the structure of such competition is relatively stable, it may also be considered as one of the indicators of the character of a political system, much in the way in which stable alignments have been used to that avail. Promising lines of research which lend themselves easily for cross-national research have been pioneered by Van der Eijk and Niemoller (1984), and have been applied for the countries of the European Community by Van der Eijk and Oppenhuis (1991).
4. a final urgent priority for comparative studies of elections is the organization of truly cross-national comparative studies of voters, parties and their interactions. For parties' behavior in electoral settings an important tine

78 Comparative Studies a/Elections

of research has been initiated by the Manifesto Research Group (Budge et al, 1987). For organizing comparative voter studies, the institution of direct electiolu to the European Parliament, conducted simultaneously in all member-8tate" of the Community, has been an invaluable and powerful stimulus. In 1989 this yielded the first large-scale and truly comparative study directed primarily at voter behavior. Results from comparative analyses on the basis of this study can be found in Schmitt and Mannheimer (1991) and Van der Eijk and Franklin (forthcoming).
To conclude: electoral studies are an important part of comparative roearch into the relations between 'politics - polity - policy*. However, traditional electoral research tended to remain country-based, stressing intra-systemic features rather than inter-system characteristics. Analyzing the comparative impact of elections on politics requires a research design that focussea on the common denominator of the electoral process: the functional equivalence of the institutions and the related behavior of the political and societal actors that shape the process.

PART II: Institutions, Actors and Rationality

On the empirical side, one can find for instance studies on election rules and procedures. This area is characterized by keywords such as electoral systems, suffrage and enfranchisement, electoral (and sometimes constitutional) law, the odministration of electoral law, the history of electoral procedures, registration, electoral reform, electoral fraud, electoral formulae, electoral bias, electoral recall, redistribution (the British term) or reapportionment (the American equivalent), thresholds, indirect elections, electoral colleges, political finance, etc.
A quite different area of study is that of election results and their aggregate distributions. This is an area of macro-analysis, where relevant terms include electoral geography, electoral cleavages, normal vote analyses, electoral change, constituencies, (again) redistribution and reapportionment, electoral sociology, popularity functions, electoral cycles, etc.
A third large part of the literature is concerned with voter behavior. Characteristic terms in this area of micro-analysis are, amongst others, electoral participation, party choice, party identification, issue preferences, candidate evaluations, political ideologies, (non)rational choice, as well as all kinds of attitudinal referents such as competence, apathy, efficacy, alienation, (post)materialism, etc.
Yet another area focusses on the agents who compete for votes, and the way in which they do so: candidates, parties and their campaigns. The relevant vocabulary of this area includes: recruitment, selection and nomination, election campaigns, political financing, political communication and mass media, propaganda, primaries, canvassing.
In addition to empirically oriented literaturei> there is a vast body of work which is primarily normative or conceptual in nature. Implicitly or explicitly, much of this work centers around the relationship between elections and democracy. Some of this work is very general in nature, some of it centers around the (mathematical) (im)possibility of voting procedures being proof against manipulation, and some is specifically directed towards the pros and cons of different electoral systems. Still different bodies of literature, which may be empirical as well as normative in character, deal with phenomena such as referenda, plebiscites, initiatives and recall.
In addition to the topics mentioned above, o multitude of additional

subjects can be found in the literature of electoral studies, as well as additional and alternative ways of categorizing them. In view of this diversity, one might wonder whether it makes sense at all to speak about the field of electoral studies. After all, what do have studies in common which deal with, for example, the determinants of turnout in American elections (Wolfinger/Rosenstone 1980). funding of political parties in Britain. (Ewing. 1987), constitutional design and elections (Powetl, 1989), the impact of issues on party choice in the European elections of 1989 (Kuechler, 1991), and similarities between voters and their representatives in the Netherlands (Thomassen, 1976)? At times one could also wonder about the relevance to political science of a large number of the sometimes esoteric topics researched. Why should political scientists concern themselves in great detail with phenomena which at first sight seem to belong to the realm of (social) psychology, or law, or sociology, or history? Already in 1967 it was, correctly, observed that
"work in the field is now a recognized sub-discipline within political science. Limited consideration, however, is given to the relevance of voting and elections for the political system as a whole" (Rose/Mossawir. 1967: 173).
The relevance of this observation has hardly diminished since then.
The contribution and relevance of electoral research to political science is difficult to assess from a categorization of the real-world topics addressed in such studies. As is the case in other disciplines as well, research into specific phenomena is relevant only because it contributes (or is assumed to contribute) to solving larger 'puzzles*, the relevance of which is more obvious. The more developed a field of scientific inquiry becomes, the larger the number of seemingly esoteric details and sub-specializations. One can view these only in their proper place by stepping back into more general perspectives. In this case this means first of all to appraise the proper place of studies of elections in the discipline of political science.
From the perspective of political science the rationale for electoral studies is to determine what impact elections have on politics. Slightly more elaborated, this involves the task to elucidate -empirically as well as nonnatively-the impact of the institution of regular elections on the character of the political

system and the course of political processes. Two elements of this description de-ve clarification; character and course.
The term character is used in this context to denote how democratic o political system is, or how just it is, or equitable, open, efficient, stable, peaceful, etc. It refers to characterizations of the state of the political system, which are generally of a normative or moral kind. The meaning of such terms is never self-evident or uncontented, but can be appraised by analyzing the conceptual and theoretical foundations of the context within which they are used (refer to, e.g. Connolly, 1983). Character refers also to the way in which symbols, values, parties, population groups etc. are associated with one another, or, in the jargon of the discipline, to existing alignments in a political system which to a large extent constrain the processes of daily politics.
The course of political processes refers to the dynamics of a political oystem, the ongoing processes of power-acquisition, agenda formation, decision making, policy implementation, conflict resolution, etc. Character and course are necessarily interrelated. The course of political processes may alter the character of a system, which in turn facilitates or obstructs those processes develop in specific directions.
The rationale for electoral studies presented above presumes that the institution of regular elections does indeed affect character or course of politics. But how justified is this assumption? Is it more than merely a reiteration of one of the central ideological tenets of representative democracies? This chapter is based on the proposition that it is more than just that, at least to the extent that one is willing to make the assumption that character and course of politics are affected by the behavior of the individuals and groups. It is manifest that many people including those who command social and political institutions are convinced of the political relevance of elections and that they act accordingly. Were they not, the large investments would be inexplicable which they make in time, effort, money and other resources to conduct elections, to attempt to influence their outcome and to participate in them.
3.1.1 Compwwive Studies of Elections
HavioJt established in what the contribution can be of studies of elections to political science, the next question is what sets comparative studies of elections

opart. For the study of politics, such comparative) studies derive their value not from taking electoral behavior and election results as the phenomena whose variation is to be explained by systemic differences, but rather how variations in doctoral phenomena (behavior, procedures, outcomes, etc.) affect politics. Furthermore, in the line of the rationale for electoral studies discussed above, comparative studies of elections also analyze the consequences of systemic or contextual differences on the impact which elections have on politics. The 'dependent* variables are thus, just as in the case of not explicitly comparative studies, the character of the political system and the course of political processes.
Systemic or contextual differences may be institutional in character, or political, cultural, social or economic. They should be identifiable as variables, and should not be referred to in terms of proper names of systems, periods, etc. (Przworski/Teune, 1970).
In principle, comparisons can be made across different dimensions. One possibility is, for example, comparison between different administrative or organizational levels of a single system, as when one compares within a single country local, regional and national elections. Alternatively, comparisons may be longitudinal, as when one compares within a single system elections at different times, i.e. elections conducted under different historical, economic, and possibly also different institutional, cultural or political circumstances. A third possibility is comparison between systems. Most frequently this involves national elections in different countries. Although not entirely correct, the last-mentioned case is most frequently referred to as 'comparative'. This chapter concentrates in particular on such cross-national comparative studies.'
Cross-level and longitudinal comparisons have great value in comparative research. Cross-level comparisons within a single country often show the significance of systemic differences which are overlooked in cross-national comparisons. For example, differences in the autonomy of the political arena for which the elections are conducted have great impact, both on the behavior of actors in the electoral process (voters, parties, media, etc.),and on the systemic consequences of elections .e.g. Reif/Schmitt, 1980; Van der Eijk et al., 1992). Cross-level and longitudinal comparisons have the disadvantage, however, that many interesting systemic characteristics are by and large invariant.

3.1.2 Political Consequences of Elections
The overwhelming majority of publications on elections hardly addresses the question of the impact of elections on politics and often neglects the consequences of systemic differences on this impact. Most of the literature starts from the unspecified assumption that elections matter politically. This assumption is then believed to justify a shift of focus to all kinds of phenomena which may affect elections and which, by virtue of the assumption just made, are therefore also possibly relevant belt in an even less specified way to the study of politics. Instead of studying election results, for example, one focusses on the behavior of voters, or of parties. Sometimes the choice of topics reflects a further regression into antecedent phenomena, for instance when not voter behavior, but factors possibly influencing it become the phenomenon to be explained. Such research may very welt make invaluable contributions to the study of politics, but only if its findings can be integrated in a more direct focus on politics. This requires a more integrative perspective which specifies the ways in which elections (or more antecedent events) have political consequences. A number of authors has attempted to make such attempts by classifying the consequences which elections may have for politics (e.g. Rose/Mossawir, 1967;
Converse. 1975; King, 1981;Kirkpatrick, 1981; Beck. 1986; Van der Eijk, 1992). Although they do not arrive at identical results, their suggestions are largely overlapping and clearly compatible with one another. The most important classes of political effects they distinguish will be discussed below.1
When this text speaks of 'the impact of elections on polities', this is not meant to indicate a simple one-way causal process. In most cases, reciprocal causal influences appear to be much more plausible, resulting in a chicken-and-egg relationship between cause and effect. King (1981, p, 294) states this very aptly as follows: "...to discuss the effects of democratic elections on the political systems in which they occur is rather like discussing the effects of skyscrapers on New York or of Judaism on Israel. Each is typically defined in terms of the other....'.In spite of this, the assertion of this chapter is that it makes sense to talk about impact (or, as some authors do, of 'function') even when this is part of reciprocal relationships and feedback loops.

". official recognition and legitimation of values which are inseparably embodied in the electoral process and its mstitutionalization. In western democratic systems, such values include those of popular participation in politics, political equality, majoritarianism, free expression of political preferences, representation, accountability of power-holders. In spite of their apparent negation by existing electoral practices in authoritarian or totalitarian systems, these values retain even there some of their liberal-democratic and in those contexts subversive meanings. Differences in electoral institutional! zation and actual practices generate variation in the extent to which these, and possibly other, values are legitimized.
b. allocation of power. Manifestly, the most direct political consequence of elections is that individuals or groups (parties) are invested with the power of elected positions and offices. Which positions varies between systems. Most common is that legislatures are elected. Less common is the election of executives and still less common that of judiciaries. How important elected offices are is dependent on their relation to other positions of power, such as economic, military, religious, etc. Irrespective of whether or not such relationships are codified in constitutions and positive law, their actual contents can only be assessed by an analysis of the systems under consideration. In those countries where the executive branch is not elected, there is usually a more or less direct link between the composition of government and that of the elected legislative. Depending upon the party system, such a link may be direct (as is, for instance, usually the case in Great Britain) or mediated by a process of coalition formation which is not controlled, but at most circumscribed by electoral verdicts.
Depending upon the manner in which it is institutionalized, the electoral process may also affect the structure of the competition for votes (i.e. the party system) and the composition of (elected) leadership in a political system.
Finally, elections serve as mechanisms for the transformation of other sources of power into political power. Political systems differ in what they encourage, allow or inhibit in this respect. Quite often, limitations are set on the use of financial or economic power in electoral settings. Votes are usually supposed not to be (directly) for sale. Yet, where poverty is rampant and economic inequality u large such practices are exceedingly frequent, which has

nude cynical observers remark that elections are one of the few effective mechanisms for redistributing wealth. In a less direct way, economic resources may cicctorally be transformed into political power by financing campaigns etc. In a similar way, religious power, government power, military power, etc. can under certain circumstances be transformed into political power by electoral means,
c. legitimation of the political community, the regime and officeholders. Particularly in the age of general elections, where substantially the entire adult population has the right to vote, elections define political communities. Often these definitions are uncontested, elections then reinforce them by default. In some cases, however, elections may undermine the legitimacy of the political community as formally defined, for instance when segments of a population boycott elections, or when election results induce secessions. Obviously, it is not the election per-se that undermines legitimacy, but rather the manifestation of such disagreement which elections permit.
Whereaa the legitimacy of the political community is often uncontested, that of officials is more frequently subject to dispute. One of the important functions of elections is to bestow legitimacy on officeholders. Often, the legitimizing effects of elections for officeholders are thought to be so inevitable that they can even be obtained by fraudulent practices. Indications are, however, that this is an oversimplification at the very least, and that the strength of whatever legitimizing effects elections may have depends on a number of characteristics of the electoral process. One can think in this regard of, amongst others, the freedom and fairness of the election, the breadth of choices offered, how directly the occupation of certain offices is tied to an election result, disproportionatity between election results and distribution of seats, etc.
Legitimacy of regimes may be reinforced or undermined by election results. Reinforcement may be the default, at least to the extent that elections withheld popular support for anti-regime forces while being free and fair and open for such forces to compete. When these conditions are not fulfilled, the regime legitimacy accruing from elections may be very low indeed, as the experience of Eastern Europe since 1989 has demonstrated.
As Easton (1965) already suggested, legitimacy for the political community, regime and officeholders may be distinct, yet interrelated. This is particularly the case for community and regime, where the borderline between

them may be hard to draw, as the cases of Canada and Belgium illustrate.
d, control over. influence on and legitimacy of policy. Control over policy is rarely exerted by way of the electoral process. Referenda are of course an exception,! as are the very rare cases where (he entire electoral contest is dominated by a single clearly defined and discriminating issue. Influence on policy formation may be a more common phenomenon. It may involve that voters base their choice on the difference between parties' or candidates ideologies or values in the expectation that these will guide policy choices on future issues. A different possibility is that parties anticipate (some) voters* preferences in the expectation of acquiring (or retaining) their support. Legitimacy of (government) policy is usually the consequence of the legitimacy of those who enact it, at least to the extent that the policies concerned fall within the normal scope of power of officeholders. Legitimacy of policy may be undermined to the extent that it violates explicit pre-election promises.
In most cases, prudence should be exercised in attributing influence over policy to voter choice. Careful analysis on the level of voter perceptions, motivations, and actual behavior is required before the an electoral majority for a party or coalition is interpreted as popular support for its platform (Khngemann et al., 1993).
e, expressing and integrating conflicting interests in a society. As Lipset and Rokkan (1967: 4-5) pointed out "the establishment of regular channels for the expression of conflicting interests ...helped to stabilize the structure of a great number of nation states". Obviously, the institution of regular, free and competitive elections is one of the most important of such channels. It allows the formation of groups or parties which represent the expression of diverse sides of social conflicts, and it provides a framework for the acquisition of shares of power within a larger system, and therefore a slake in that larger system, at least
As long as this larger system may contribute to the furthering of the various conflicting interests, elections promote their gaining an interest in the larger system itself, thus performing an important integrative social and political function.
f. communication between elites and non-elites. The perspective of elections offering a mechanism for ordinary citizens to collectively exert control, or at least influence, over the behavior of governing elites may be overly naive. Politi-

cal leaders and elites can, and do influence the electoral process by shaping the agenda of public discussion, by their participation in public debate, by influencing reality through the process of governing, etc. When taking this into account, the electoral process shows itself as a channel of communication between elites and ordinary citizens. To the extent that it operates effectively, it transmits messages from a population to its political leaders- These involve support, disaffection, preferences, etc. It also transmits information from elites to the masses which they govern: information, explication, persuasive messages. and so on, all in the context of elections and election campaigns. This channel of communication can be said to be effective to the extent that the flows of information carried by it are intelligible for both sides. This aspect will be discussed separately below.
Elections are. of course, not the only channel of communication between masses and elites. People may use petitions, demonstrations, lobbies, all kinds of activities to convey their opinions and preferences to their governors. Likewise, elites may in all different forms attempt to inform, persuade, cajole, bribe or mislead populations. Yet, the unique character of the electoral process is that it is the only of such channels of communication which embodies the democratic principle of political equality, and that its effects on leadership composition are unequivocal.
3.2. Electoral Change ana the Decline of Cleavage Voting
This section summarizes an important and voluminous tradition of (comparative) research on elections which is focussed upon the explanation of party choice. As will be apparent, such research is of particular importance because it helps to illuminate central aspects of the character of a political avBtem: the nature of the long-standing conflicts in a society, their relation to political parties and social groups, their importance as anchors or orientation and choice of individual voters. With this focus, it contributes also to understanding other political consequences of elections, such as the bases upon which electoral power is acquired, the grounds of legitimation of political communities, regimes, and officeholders, the kinds of policy areas in which

representation, control and influence is important, and the terms in which elites and non-elites can communicate with one another.
On the basis of a comparative analysis of West European countries, Upset and Rokkan concluded in 1967 that
"...the contemporary party systems reflected, with but few exceptions, the cleavage structures of the 1920s. (...)the party alternative, and in remarkably many cases the party organisations, are older than the majorities of national electorates" (p. 169).
They characterized this state of affairs as 'frozen party systems', a term referring not only to party systems in a narrow sense, but also to the relation between political parties and electorates. In other words, the party systems of the 1960s not only reflected the cleavage structure of the 1920s, but also that of the 1960s, and this cleavage structure itself, with its consequences for voters' behavior and election results, was equally frozen. This implicit conclusion was empirically tested by Rose *nd Urwin (1970), who were indeed able to demonstrate that during a period of rapidly changing circumstances the electoral support for most relevant parties had been remarkable stable. Rapid economic growth, the emergence of welfare arrangements and decolonization in the period 1945-1969 had evidently had little impact on election results in Western countries during
mis period.
Just about at the time that Lipset and Rokkan and Rose and Urwin
published their results, the situation started to change. Many countries experienced dramatic increases in electoral change. New parties sprang up, old ones were often in disarray, the erstwhile predictability appeared to have vanished. The changes appeared to be different in different countries. They did not set in at ins san" time, they affected different parties, they took different forms. In the Netherlands the most striking was the stark decline of Christian-democratic parties in the 1967-1972 period. In Denmark, 1973 showed a sudden fall in the electoral support of the four traditional parties from 84 to 58 percent of the vole, arid a doubling of the number of parties in the national parliament from five to ten. In Britain the Labour Party split in 1982, resulting in such unprecedented electoral support for a third party (an alliance of Liberals and the split-off Social Democrats) that the traditional two-party dominance in

British politics almost lost its self-evident character. Yet other countries showed their own timing and manifestations of electoral changes which contributed to the demise of the perspective of frozen party systems. What had happened? How was this to be explained?
The dominant mode of analyzing and explaining electoral behavior and election outcomes was by means of the kind of social cleavages Lipset and Rokkan had referred to. These included the cleavage between classes, between urban and rural interests, between different kinds of religious interests, and between these and secular ones as exemplified in the modem state, between cultural, ethnic and linguistic segments of a population. The most relevant of these cleavages had, according to Upset and Rokkan, resulted in the formation of parties in the late 19lh and early 20th century which represented the interests of populations groups which were demarcated by cleavage distinctions. The extension of Ihe suffrage in this period offered the possibility for political parties to establish their own cleavage-based clientele, to build their organizations and to become part of the political system. When universal suffrage was attained, new parties could only establish an existence at the expense of the already existing ones, which usually proved too difficult. At the same time, the existing parties tried to convince their respective clienteles of the continuing relevance of cleavage distinctions, and of the need to safeguard the interests of the various segments of the population by cleavage-based parties. Such processes account for the stability referred to by Lipset and Rokkan and by Rose and Urwin.
The success of the cleavage perspective for the period until the late 1960s accounts for various attempts to rescue it as an analytic approach for the period thereafter. First of all, the cleavage perspective did not predict only stability. When demographic, social and economic developments alter the size of cleavage-based groups in a population, it was to be expected that the parties deriving their support from those groups would be affected accordingly. This line of explanation ran, however, in unsurmountable problems. One problem involved timing. Why would all kinds of contextual developments, which were already under way for decades, all of sudden impact on the size of clienteles? How to reconcile abrupt electoral changes with very gradual and slowly evolving social and economic changes? Even worse, electoral changes were often not unidirectional in their effects, whereas the contextual developments which alter the size of cleavage-based groups are. A second kind of attempt to rescue cleavage

based approaches was to attempt to adapt the definition of the cleavage structure which was supposed to underlie the electoral support of parties. This has most frequently been attempted with respect to the class cleavage. The simple distinction between labour class and middle class, which might have been adequate for the early 20th century, was not so any more later. A 'new middle-class' was often proclaimed to have emerged, which would constitute the electoral basis for new parties, or, in their absence, for electoral volatility. Here too, however, empirical evidence was hard to reconcile with this notion. Apart from the problem of how to sensibly define new class distinctions, newly defined groups did not behave as expected, or were hard to distinguish in their behavior from others. A third way to rescue the idea of cleavage politics is to hypothesize new cleavages, which earlier were of no or less relevance. The newly politicized issue of women's rights, the different interests associated with public versus private employment, or wit^ public versus private consumption, or with materialist versus post-materialist concerns have all at one time or another been hypothesized as new cleavage dimensions which could account for electoral change and volatility. None of these, however, has had much empirical success in explaining electoral behavior and election outcomes (see for this: Dalton el al., 1984;Franklin et al., 1992).
The conclusion of diminished relevance of the cleavage perspective was drawn in a number of non-comparative studies in the 1980s. Van der Eijk and Niemoller (1985: 367) referred to it as "the political emancipation of individual citizens who can now choose, rather than be predestined by social position". Franklin (1985: 175-9) termed it "the posl-colleclivist era", and Rose and McAllister (1985) alluded to it in the title of their book Voters begin to choose. All refer to a new situation in which voters' position in terms of socio-demographic cleavages had lost its former stronghold on their individual behavior. Even where such rather general interpretations were given to electoral changes, the specific national contexts of studies generated quite different interpretations of the origins of these changes. In the Netherlands, for example, much emphasis was given to processes of secularization which undermined the potency of religious cleavage divisions, in Britain it was the declining importance of class. Only recently did a large-scale comparative investigation of patterns of electoral change in Western countries demonstrate that similar patterns of diminishing relevance of cleavage voting exist or are under way in many political

oytoms, even in countries where they were not yet identified before, and in "y"tems where (aggregate) electoral change so far bad been quite small. Franklin. Mackie and Valen, who directed this study (1992). concluded that a more general process of change exists than what has been detected earlier in separate country studies.
The declining importance of social cleavages for electoral choice was demonstrated for a large number of countries, some of which had shown not very much turbulence in election outcomes, and each of which has to be characterized by a separate set of traditionally relevant cleavage distinctions. A simultaneous longitudinal (1960s-I970s-1980s) and cross-national comparison documented a strongly diminished capability of cleavages to explain voting behavior in Norway, Denmark, Sweden, Belgium, the Netherlands, France, Britain, Australia, New Zealand and Ireland. In the United States and Canada, not much of a decline could be found, but there impact of cleavages on voter behavior was already very low in the 1960s, allowing for only small downward changes. The only two exceptions to the general pattern are Italy and Germany, where the impact of cleavage positions on voting had remained the same since the 1960s, or had even increased somewhat. At any single point, a comparison between these various counties would not have been very illuminating. In the 1960s they varied considerably in the strength of cleavage voting, which was very strong in some countries, intermediate in most of the rest, and weak only in a few. The time at which this strength declined, the rate of decline, and the remaining strength in the 1970s and 1980s do all vary. Only comparison over a longer time-span than a few years or even a single decade allowed the common pattern of decline to show itself clearly.
Franklin et al. (1992) suggest that the observed loss of capacity of social cleavages to determine voter behavior in many countries is part of a long-term process which antedates and follows the window of observation provided by longitudinal comparisons over a period of two to three decades. Exactly when this process starts differs from country to country, causing quite different 'views' at any specific point in time. This suggestion implies on the one hand that all of the systems under study (western democracies with competitive elections) experience a stage in which voter choice is strongly determined by social cleavages, whatever those may be in each specific country. On the other hand it implies that this electoral importance of cleavages will eventually disappear

largely. TniB perspective ix not only based on extrapolation of documented changes so as to encompass periods prior to and after the actual observations, but follows also from a careful breakdown of results according to generations. In virtually all countries studied, party choice of younger generations is less strongly determined by cleavages than that of older ones. Consequently, the extrapolation of over-all changes in cleavage impact is in agreement with the consequences of a simple and inescapable process of generational replacement prior to and after the actual period studied. Yet the question remains what the nature, origins and implications are of this suggested general process.
3.2.1 Originsand Implications a/Declining Cleavage Voting
The decline of cleavage voting which had apparently occurred in many western countries, or which is still under way, cannot! adequately be understood in terms of country-specific factors, such as secularization, decline of class voting, etc. Such interpretations are not necessarily incorrect, but cannot account for the commonality of the process in different systems. The common denominator is o decline in the relevance of traditional social groups (on whatever basis these may be constituted) as anchors for voters and political systems. In a discussion of the origins of this process Van der Eijk/Franklin et al. (1992: 420) state that
"In the most general terms, the decline of socially structured voting can be explained in the same way as the decay of many other empirical structures, in terms of the fact that none of the countries we have studied constitute closed systems. Therefore, their structures may welt have contained the seeds of their own decay. Leakages and contamination might be seen as being 'bound* to bring about an eventual decline in cleavage politics."
This general principle is illustrated in four different ways. each of which contributes to the decay of an erstwhile cohesive structure of alignments between social conflicts, the social groups and their interests which are defined by these conflicts, and political parties.
First of all, social structure itself has changed in quantitative terms. Some cleavage defined groups grew in size, others diminished. Such slowly occurring changes have their origins in demographic, social and economic

developments in the past century. They may gradually cause a change in the character of entire societies as erstwhile dominant groups are replaced by others. This in turn undermines the strength of linkages of voters to a party system blued on increasingly more outdated cleavages.
Second, social structure has changed in qualitative terms, rendering party systems based on cleavages of many decades ago increasingly less relevant for increasing numbers of voters. New occupational groups, new modes of acquiring income, new patterns of residence, changed systems of transport and communication all create new interests and conflicts which are not easily expressed, let alone represented by parties defined in terms of cleavages at the beginning of the century.
Third, the coherence that previously existed between central and reinforcing characteristics of particular social groups has reduced. The once existing reinforcement given to for example occupational class by education, income, housing type, health care, etc. has disappeared, to no small measure as the consequence of state interventions in the course of the century, and in particular of the creation of welfare arrangements. This lack of coherence puts increasing numbers of people under frequent cross-pressures. In such a situation, people may seemingly retain for a long time linkages with established cleavage parties, but these will increasingly rest upon inertia and be subject to sudden change.
Fourth, generational replacement is an evident factor preventing the closedness which a cleavage structure would require in order not to be subject to decay. It can be argued that parental transmission of attitudes and values can only be entirely successful if they are not reinforced by personal experiences felt by succeeding generations. Unless this is the case, successive generations will have increasingly less reason to re-transmit their inherited identity. Consequently, the strength of identifications with cleavage-defined groups will gradually be reduced, unless successive generations will personally experience the conflicts which gave rise to these cleavages in the first place.
The first three of these causes of decay of cleavage structures relate to 9. diminishing capacity of the old structure to accommodate newly emerging conflicts, interests and problems. The fourth specifies a condition which must be fulfilled for the first three to be consequential: if important old conflicts have not been resolved new ones are unlikely to achieve prominence. To the extent

that Western countries managed to find political solutions for the conflicts. social problems and unfulfilled aspirations which were the foundation of the cleavage structures in the beginning of the century, these cleavages ran out of steam and simultaneously allowed new conflicts and interests to gain prominence, which further undermined the bases of the traditional cleavages. Evidently, this process has run its course in many Western countries, leaving in the course of the 1960s the traditional cleavage-based party systems seemingly intact, but without much resistance against whatever challenge. Little wonder that wherever such challenges occurred, serious electoral change resulted.
It must be emphasized that the decay of cleavage structures is not portrayed as the inescapable result of the mere passage of time. Earlier, the condition has been formulated that the important conflicts which gave rise to the cleavage structure in the first place must be resolved for this outcome to occur. Northern Ireland is a clear example of what happens when this condition is not met. In the course of the 20th century, economic and social developments in Northern Ireland have not been very different from most of the rest of the United Kingdom. Yet, group identifications, cleavage structures and the traditional conflicts underlying them are as much in place, if not stronger, than at earlier moments in the century. The inability to adequately resolve the conflict between Protestants and Catholics in Northern Ireland resulted in a continuous barrage of personal and intense experiences for all generations which reinforced the saliency of the conflict and of personal identifications with either of the two sides. Regarded from this perspective, the decline of cleavage voting which was observed in elsewhere was the consequence of the successful resolution by political systems of deep-sealed conflicts of social interests. These countries succeeded in politicizing their major and most divisive conflicts in the form of cleavage-based political parties. This not only allowed the expression of group-based conflicts of interest, it also helped to resolve them by integrating them into a single national political system: "the establishment of regular channels for the expression of conflicting interests ...helped to stabilize the structure of a great number of nation states' (Lipset and Rokkan, 1967:4-5). Obviously, the institution of regular, free and competitive elections is one of the most important of such channels. The crucial difference between Northern Ireland and many other countries is the degree to which all conflicting groups in a society can express their interests legitimately and effectively, and the extent.

• [ На главную ]
• [ Библиотека ]
• [ Reader ]
• [ Программы ]
• [ Абитуриенту ]
• [ Фотоальбом ]
• [ О ФП ]
• [ Гостевая книга ]
• [ Ссылки ]
• [ Обратная связь ]