This book is based on a conference sponsored by the Joint Committees on Japanese Studies and Western Europe of the American Council of Learned Societies and the Social Science Research Council, and the Japan Society for the Promotion of Science. Support for the activities of the Joint Committees was provided by the Ford Foundation and the National Endowment for the Humanities.
The One-Party Dominant Regimes
Edited by
T.  J.  PEMPEL
CORNELL UNIVERSITY PRESS
ITHACA    AND    LONDON
Copyright © 1990 by Cornell University
All rights reserved. Except for brief quotations in a review, this book, or parts thereof, must not be reproduced in any form without permission in writing from the publisher. For information, address Cornell University Press, 124 Roberts Place, Ithaca, New York 14850.
First published 1990 by Cornell University Press.
International Standard Book Number 0-8014-2367-8 (cloth) International Standard Book Number 0-8014-9696-9 (paper) Library of Congress Catalog Card Number 89-22111 Printed in the United States of America Librarians: Library of Congress cataloging information appears on the last page of the book.
@ The paper used in this publication meets the minimum requirements of the American National Standard for Permanence of Paper for Printed Library Materials Z39.48—1984,
CENTRAL-BI6UOTEKET
Preface                                                                                                 vii
Contributors                                                                                         xi
Introduction. Uncommon Democracies: The One-Party Dominant Regimes                                                                                                  1
T. J. Pempel
1    Single-Party Dominance in Sweden: The Saga of Social Democracy                                                                                    33 G0STA Esping-Andersen
2    Conditions of Labor-Party Dominance: Sweden and Britain Compared                                                                                     58
Jonas Pontusson
3    The Political Economy of Labor-Party Dominance and Decline
in Israel                                                                                        83
Michael Shalev
4    Defense Controversies and One-Party Dominance: The Opposition in Japan and West Germany                                      128 Hideo Otake
5    Establishing Party Dominance: It Ain't Easy                                162 Giuseppe Di Palma
6    The Political Economy of Conservative Resurgence under Recession: Public Policies and Political Support in Japan, 1977-1983                                                                                 189 Takashi Inoguchi
32   /   T. J. Pempel
fold,  albeit with far fewer benefits than were  accorded to its  core
constituency.
This same pattern can be found with the Italian DC, as Sidney Tarrow shows. In contrast to the attempted "hard hegemony" of conservatives in France, the DC ruled with a "soft hegemony" that left it far more able to respond tactically to its opponents and to changing political opportunities than the DCs more rigid French counterparts.
My concluding essay attempts to tie the chapters and case studies together by examining three central questions: What are the necessary conditions for the establishment of one-party dominance? What conditions seem to cut across all the cases? How can one examine the commonalities that seem most important among the four major cases? In this context, I examine the means by which a single party gains and maintains power as well as the specific difficulties encountered in maintaining or changing the party's socioeconomic base while still holding electoral office. I also argue that long-term rule can be used to impose an agenda that makes continued rule more likely. In this sense, the conclusion focuses both on the causes and consequences of dominance. It argues essentially that the two feed each other and that long-term dominance by a single party involves clever tactics of electoral mobilization, ideological positioning, and governance. When blended skillfully, these are mutually reinforcing and suggest that one-party dominance is an art far more than it is an inevitability.
1. Single-Party Dominance in Sweden: The Saga of Social Democracy
G0STA Esping-Andersen
The Swedish Social Democratic party (SAP, or Sveriges Socialdemokra-tiska Arbetarparti) must be examined in any explanation of one-party dominance in democratic polities. It is, worldwide, the party that has governed longest and most consistently over the past five decades. This achievement is the more astounding because it is explicitly a (democratic) socialist party.
After a number of aborted efforts, the SAP came to power in 1932 and remained the governing party for the next forty-four years. In 1976, it was forced into opposition for two successive election periods but was then returned to office in 1982,1985, and 1988. It is thus the longest governing democratically elected party in the Western world. If we consider the six-year period of opposition as a de faao continuation of Social Democratic politics under a center-right label (as I do), the dominance of the SAP within Swedish politics is even more overwhelming. Indeed, Sweden arguably has experienced one-party hegemony.
The conventional theory of competitive democracy is difficult to reconcile with the unique degree of electoral stability that marks Swedish social democracy. From one election to another, the party has hardly deviated more than two or three points from its average of about 46 percent of the electorate. On two occasions, its share surpassed the magic 50 percent mark; first in the unusual 1940 wartime election and then in 1968 (with 50.1 percent). Its worst performances since 1932 occurred in 1956 (with 44.6 percent) and
A large part of the description and analysis presented in this chapter is based on Gssta Esping-Andersen, Politics against Markets: The Social Democratic Road to Power (Princeton: Princeton University Press, 19S5). An earlier version was presented to the Social Science Research Council working group on One-Party Pemocracies at Cornell University in April 1984. I thank T. J. Pempel for his generous comments and patience.
33
34   /   G0STA Esping-Andersen
in 1976 (with 42.8 percent). Following both these losses, the party was able to recuperate its habitual electoral position.
Dominance, to say nothing of hegemony, is obviously not a question of electoral returns alone; it must assume a capacity to shape policy making and, more generally, to structure society in accordance with a party's priorities. It is this quality that most sharply demarcates the SAP from its socialist brethren in Scandinavia and elsewhere and that underpins the argument that the SAP is one of the few examples of party hegemony worldwide. Sweden does not just happen to harbor the world's strongest socialist movement; decades of labor movement influence have rendered Sweden itself social democratic.
The 1930s were a watershed for social democracy in Scandinavia as a whole. The Swedish Social Democratic party evolved, came to power, and consolidated its position pretty much in tandem with its Danish and Norwegian counterparts. The model was largely identical: both the parties and their affiliated trade unions grew very rapidly after World War I, adopted a reformist political line, and managed to obtain governing majorities on the basis of an alliance with the Agrarian party. In all three nations, the postwar era has been essentially identified with social democracy. Until World War II, the Danish party was broadly regarded as the leading force of the Scandinavian labor movements. Subsequent developments, however, led to sharply divergent trajectories. While Swedish Social Democracy established its hegemonic status, the Norwegian and especially the Danish parties began to decay from late 1960s onward. Indeed, the two latter became incapable of maintaining their accustomed role as the natural parties of governance, their capacity to control governments slipped, and they were confronted with both a growing leftist opposition and the eruption of populist rightist anti-welfare-state movements.
These contrasting profiles are evident in the parties' electoral shares. In contrast to the Swedes' extraordinarily stable 44 to 46 percent, the Danes slipped badly in the 1960s and even more in the 1970s; the Norwegians paralleled the Danish fate in the 1970s (see Table 1.1).
Table 1.1. Average electoral shares of the Scandinavian Social Democratic parries, 1950s, 1960s, and 1970s {percentages)
	Denmark	Norway	Sweden
1950s	41	47	46
1960s	38	45	48
1970s	34	38	44
Source: Gosta Esping-Andersen, Politics against Markets: The Social Democratic Road to Power (Princeton, N.J.: Princeton University Press, 1985), pp. 327-29.
Single-Party Dominance in Sweden   f   35
None of the three parties has escaped the fate of being ousted from office, but the experience of each was quite different: the Swedes returned from their six-year opposition status with their accustomed strength; the Norwegians were ousted in the mid-1960s and were never again able to assert themselves as the natural governing party; and the Danes shuttled in and out of government throughout the postwar era with the need to forge increasingly difficult and precarious coalitions. In Denmark, the nonsocialist parties came to dominate the cabinets of the 1970s and 1980s, and the Social Democrats slipped to an average 32 percent of the vote in 1981-84. One of the most significant contrasts has to do with challenges to the parties from either left or right. Again, the Swedish case is unique. As early as the 1930s, Swedish Social Democracy effectively emasculated the competition from the left (in the form of the Communist party), and the right has been chronically unable to forge a united front against Social Democracy (despite presiding over a numerical majority on numerous occasions). Even during the period 1976—81, the "bourgeois coalition" splintered repeatedly and was forced to continue in the end as a minority cabinet. In contrast, both Danish and Norwegian Social Democratic parties have suffered from the right's superior ability to coalesce and present itself as an effective governing force. In Denmark, especially, the Social Democratic party became increasingly squeezed between a strong (noncommunist) left (averaging 10 to 12 percent of the vote in the 1970s and 1980s) and an electoral move toward the right.
If we examine the social bases of voting, the variable performances of the three social democracies becomes even more pronounced. As is often argued,1 a basic problem facing socialist parties is to hold on to their traditional working-class base while simultaneously mobilizing among other electorates. Available data paint a clear picture of Swedish Social Democracy's unique capacity to override this dilemma. Its sustained electoral stability is clearly owing to its ability to maintain the loyalty of workers while simultaneously making major inroads into the new salaried strata; the party's weakness in the late 1970s was related to a marginal drift of both groups to the right. Crucial to the Swedish party's long-term strength is its continual ability to mobilize newly entering young voters.
The Danish case is an extreme contrast. Over the 1970s, the Danish party lost large sections of blue-collar voters to both the left and the right, and it has been only marginally successful in mobilizing the new salaried groups (who, interestingly, are a prominent part of the New Left's clientele). Perhaps the most serious symptom of the party's decomposition in Denmark is its growing inability to attract younger voters. It is rapidly becoming a party of old-age pensioners.
1. Adam Przeworski, Capitalism and Social Democracy (Cambridge: Cambridge University Press, 1985); Esping-Andersen, Politics against Markets.
36   /   GßSTA Esping-Andersen •
These contrasts illustrate the basic point that, despite common origins, the Swedish Social Democratic party commands an exceptional status in a Scandinavian as well as a global context. The question is why? What are the causes of Swedish Social Democracy's extraordinary performance?
An answer to this question would be impossible if our inquiry were limited to the Swedish case study. There would be too many contending variables to account for. A comparison with the other Nordic countries offers a way to reduce the number of possible explanations substantially: a host of historical and sociological factors can quickly be dismissed because they would obtain more or less equally for Norway and Denmark as well.
Nonexplanations
It would be tempting to search for an explanation of Swedish Social Democracy's success in the country's history of industrialization and democratization, as have numerous scholars.2 Sweden's gradualist and nonviolent path toward a modern capitalist democracy is not unusual, however. In addition to Denmark and Norway, this route was also shared by the Swiss, the Dutch, and a host of other nations. A unique feature of Scandinavia was the early decay of absolutist rule and reactionary aristocratic forces, as well as the early collaboration between farmers and the working class in the struggle for full democracy. But, again, these conditions were shared by both Denmark and Norway. Indeed, Denmark was the pioneer, Sweden the follower. The early experience with a popular red-green alliance of farmers and workers was decisive for social democracy's breakthrough in the 1930s in Scandinavia at large but does not differentiate Sweden.
In all the Nordic countries, the process of industrialization came late and rather gradually. The cataclysmic, disruptive, and polarizing consequences experienced in many European nations were largely absent in Scandinavia. It is true that Swedish industrialization has unique features that single it out from its Nordic neighbors: it was, at once, highly concentrated and geographically decentralized. This is an important issue that will be examined later in this chapter.
The literature frequently cites social and cultural homogeneity in contributing to the success of social democracy. There is no doubt that Sweden appears extraordinarily homogeneous in language, religion, ethnicity, and cultural tradition, in contrast to Norway's linguistic and center-periphery cleavage structure, but Denmark is also homogeneous.
Characteristics of the political system are a third factor often invoked in
2. Seymour Martin Lipser, Political Man (New York: Doubleday Anchor, 1960); Stein Rokkan, Citizens, Elections, Parties (Oslo: UnJversitets Forlaget, 1972).
Single-Party Dominance in Sweden   I   37
the explanation of cross-national differences in political parties' performance. Thus it can be argued that a system of strict proportional representation, as exists in Sweden, is conducive to a multiparty competitive system in which the individual parties are encouraged to develop distinct sociopolitical profiles and strong internal organization. This is certainly the case in Sweden, but it is equally so in many other nations. All three Nordic countries are remarkable in the degree to which the social classes and interests are directly represented in the party system, with a distinct Agrarian party and labor parry mobilizing and representing essentially the entire underlying clientele. But there is nothing distinctly Swedish in this respect. Indeed, another property of multiparty systems—the difficulty for any single party to mobilize an absolute majority alone—seems to have been superseded in the Swedish case and provides another instance of the extraordinary performance of Swedish Social Democracy.
Sweden has a somewhat higher threshold for party representation (5 percent of the vote) than Denmark and Norway (2 percent). A higher threshold helps stall the formation of new parries and the splintering of parties. Yet it would be difficult to regard this as a cause of single-party dominance. Other nations with high thresholds (such as West Germany) have seen new parties emerge and upset the position of the existing parties. And in.Sweden, the 1988 elections brought about the entry of a new Green, ecological party with roughly 6 percent of the vote. But in contrast to Germany, the Swedish Greens seem to rob votes from and destabilize the right rather than the Social Democrats. They certainly did not impair SAP's capacity to govern yet another terra. Characteristics of the political system alone cannot explain SAP's relative success.
In sum, in explaining single-party dominance in Sweden we can largely ignore many of the standard textbook arguments. To account for Swedish Social Democracy's dominance, we will first have to distinguish the issue of how the SAP came to power from the question of how it stayed in power. Clearly, long-term dominance is difficult to predict from the process of power attainment; maintenance of power requires consolidation and concurrent reproduction of the conditions that fuel power.
Second, we must identify the causes of SAP's dominance within the comparative framework of Scandinavian social democracy because this will allow us to single out specifically Swedish circumstances. As I shall argue, there are three factors that make Swedish conditions peculiar: the social structure within which Social Democracy emerged and matured; the party's ability to forge critical cross-class political alliances; and the SAP's unusual capacity to institutionalize and invigorate its own power bases through its reformist policies. The latter factor is of particular importance. Yet the final explanation cannot be based on the added impact of these three factors but only on the way they came to interact and reinforce each other in Sweden.
38   /   G0STA Esping-Andersen
The Origins of Swedish Social Democratic Ascendance
In classical socialist thought, the long-term power of socialist parties was assumed to follow from the process of proletarianization: as workers became the majoritarian class, the socialist parties would be able to command majority power via the ballot box. The theoretical revisionism of Eduard Bernstein3 turned out, however, to be closer to the truth. By and large, strictly working-class majorities failed to materialize.
As the socialists began to realize this logic of class structural development, they faced the choice between a continued strict adherence to the working-class party model, which would conceivably lock the party into a perennial minority status, or an alliance strategy of collaboration across the class structure to compensate for the numerical shortcoming. Choosing the latter option would almost certainly compel the party to dilute its orthodox programmatic tenets.
Still, of course, the tensions resulting from the absence of clear working-class numerical majorities can be lessened by the degree to which the available working-class clientele can be effectively unified and mobilized under the social democratic banner. Similarly, the ease with which a party can forge necessary cross-class coalitions depends on the social character of potentially allied classes. Hence the nature and composition of the underlying social structure defines the conditions under which social democracy can mobilize, gain, and sustain majorities.
Social structural conditions have been favorable for Swedish Social Democracy. Paradoxically, its rise to power in the 1930s and 1940s was based not so much on the working class as on the rural classes.
The process of industrialization in Sweden had a number of internationally peculiar characteristics. Centered around iron ore mining, hydroelectric power, and timber, it was from the start dominated by large industrial concerns but was scattered across a geographically vast nation. The industrial establishments were located in small, remote, and isolated rural localities, not, as elsewhere, in burgeoning cities. This geographical factor had profound effects on political and organizational development. First, it helped smooth the process of "proletarianization"; local, rural-based industries helped absorb the peasantry into industrial life without the social trauma of massive population movements.
Well into the twentieth century, the Swedish economy was dominated by a mass of poor and economically unviable farms. Emigration to the United States was considerable. The decentralized mode of industrialization, however, helped absorb substantial rural labor surpluses; the result was the rise of a peasant-worker class with blurred class divisions and a strong traditionalist, local-based solidarity.
3. Eduard Bernstein, Evolutionary Socialism (1899; rpt. New York; Schocken Books, 1971).
Single-Party Dominance in Sweden   I   39
Geographical dispersion meant that the emerging labor movement's strategy for mobilization had to be based on centralization. Concerted industrial action, bargaining, and strikes could be effective only under conditions of strong centralized coordination. This is one reason why the Swedish trade union movement rapidly moved in the direction of industrial unionism and national-level representation. A second reason is the absence of a strong skill- and craft-exclusive tradition; the Swedish working class has always been comparatively undifferentiated.
Thus, to a degree not found elsewhere, the Swedish union movement was favored by homogeneity and a capacity for centralized, national unification. Organizational growth was rapid in the 1920s and 1930s.4 By 1940, trade union organization exceeded 1 million, or 50 percent of wage earners; by the 1960s, about 90 percent of blue-collar and close to 70 percent of white-collar workers were unionized, the former in the central LO (Landsorganisationen) federation, the latter in the independent TCO (Tjanstemannens centralorganisation) federation.5 Compared to many other nations (Denmark excluded), Sweden did not experience sagging unionization in the 1970s and 1980s.
The homogeneity, organizational density, and centralism of the labor movement had profound effects on SAP's fortunes. As in Denmark and elsewhere, the independent strength of the union movement helped move the socialist party away from ideological orthodoxy toward practical reform politics. It was thus comparatively easy for the Swedish socialists to shelve ideological orthodoxy, move out of the traditional working-class ghetto, and embrace the broader and electorally more rewarding popular reform strategy.
The Swedish Social Democrats have always been more theoretically inclined than other Scandinavians. The party was, in fact, fairly radical during the 1920s. Its election campaign in 1927 was devoted to economic socialization {far-reaching inheritance confiscation). On the other hand, its ideological elasticity can be seen from the swiftness with which it abandoned the orthodoxy of socialization when faced with indifference or outright opposition from both outside and within the working class. The trade unionists played an important role in moving the party from Marxism to Keyne-sianism in the crucial period of the 1930s.
The organizational power of the Swedish trade unions is unparalleled elsewhere and, as we shall see later, it has become a key factor in the postwar strength of Social Democracy. But it was not a significant cause of the party's rise to power. The LO suffered a major defeat in the 1909 general conflict, and membership growth subsequently grew slowly. In the 1920s and 1930s, the Swedish union movement was weaker than the Danish, albeit more unified. More important, Swedish industrial relations in
4.  Nils Elvander, Skandmavisk Arbetarrörelse (Helsingborg: liber, 1980).
5.  Esping-Andersen, Politics against Markets, pp. 57—70.
40   /   G0STA Esping-Andersen
that period were unusually tumultuous, boasting some of the world's highest strike rates.6
Indeed, before the war it was not the unions' strength that catapulted SAP to power, but the opposite: full trade union recognition and the stabilization of Swedish industrial relations had to await the affirmation of Social Democratic governance in the late 1930s.7
The key to Swedish Social Democracy's ascent to power thus lies not in the labor movement but in the class structural mix that prevailed before the war. On one side, the party's capacity to mobilize working-class votes was vastly aided by its homogeneity, complemented by the pervasive sense of social solidarity that not only characterized die Swedish working class but society at large. On the other side, the party's rise to power could.only be premised on some form of accord with the rural classes, who were the numerical key to a parliamentary majority required for consolidation of power.
Sweden was, until World War II, the most rural of the Nordic countries. But its rural structure was peculiar. Genuinely viable family farming was limited to the south; the vast majority of the people were fragile peasants and/or wage workers. The former group was parallel to those in Denmark, although numerically relatively smaller. The latter was a strata peculiarly Swedish in that its economic interests and needs were often the opposite of those of the larger farmers but convergent with those of industrial workers. In short, the rural economy was duaUstic. Since World War I, the wealthier farmers had been effectively organized and politically represented by the Agrarian party; large sections of the rural "quasi-proletariat" were potential recruits to Social Democracy. This was a situation that both favored and threatened the SAP when in the 1930s, the Depression pauperized agricul-. ture.
For Social Democracy, then, the creation of a cross-class coalition was a necessity; the rural structure made it objectively possible. Yet neither necessity nor objective potential could have resulted in a viable political coalition were it not for two critical conditions: the nineteenth-century experience of cooperation in the struggle for full democracy and the ability to develop a common program that would override perceived conflicts between workers and farmers.
Sweden before World War II was rent with three axes of fundamental class conflict: the tumultuous and unsettled conflia in the labor market between the unions and employers' organizations; conflict between industry and farmers on issues such as tariffs and protectionism; and chronic antagonism between farmers' organizations and the labor movement. Facing de-
6.  Walter Korpi and Michael Shalev, "Strikes, Power and Politics in the Western Nations, 1900-1976," in Maurice Zeitlin, ed., Political Power and Social Theory 1 (1980), pp. 301-34.
7.  Walter Korpi, The Working Class in Welfare Capitalism (London: Routledge & Kegan Paul, 1978).
Single-Party Dominance in Sweden   I   41
:Iining food prices and severe economic problems, Swedish farmers saw mionism as a major cause of their deteriorating profitability; in contrast, workers naturally opposed higher food prices. During the Depression, Swedish farmers began to drift toward Nazi sympathies.
The Swedish Social Democrats, having followed the political events in Europe, saw the pacification of the farmers as a precondition for stable democracy as well as for coming to power. Yet, unlike many European countries, the presence of a deep conflict between farmers and industrial interests precluded the formation of a reactionary, rightist alliance. It was this combination of factors which in Sweden helped launch the red-green coalition of workers and farmers in 1933.
The prelude to the Red-Green alliance was not especially auspicious. The SAP took office in 1932 with no clear parliamentary majority. Deep conflicts over food prices seemed to preclude the possibility of support from the Agrarian party, and the SAP thus faced the prospect of governing without the power to implement its program for restoring employment and introducing long-delayed welfare state reforms. The turning point came in the wake of a parliamentary crisis in 1933, when the nonsocialist parties vetoed SAP's (rather modest) Keynesian employment policies. The SAP negotiated a package deal with the Farmers party, providing for agricultural price supports in return for agrarian backing of SAP's deficit-financed economic revitalization policies. Once established, the coalition with the farmers permitted the SAP to build its position as the dominant political force for the long-range future. But it is difficult to interpret the conditions that allowed its establishment as such as a major cause of SAP's five decades of dominance. More or less similar conditions prevailed in Denmark and, a little later, in Norway. It can be strongly argued that the Red-Green coalition was a necessary stepping-stone to power. But in and of itself, it would certainly hold no guarantees, as subsequent developments in Denmark illustrate.
The Soda! Bases of Postwar Dominance
Single-party dominance over a span of half a century requires an extraordinary capacity to adjust to changing social structure. The postwar era was, by any yardstick, one of massive social change. Was there any factor particularly favorable to Social Democracy in light of Sweden's process of modernization?
At first glance, one would probably answer in the negative. As everywhere else, the rural population diminished rapidly, the relative size of the industrial working class was stagnant and even declining, and the new white-collar middle strata became increasingly dominant. Concurrently, educational levels increased, birth rates declined, and the population was aging. On none of these counts does Sweden stand out especially.
42   /   G0STA Esping-Andersen
At second glance, however, it is possible to locate a critical source of SAP's strength in the way these universal processes unfolded, in the process of social class formation, organization, and unification. The Swedish social structure is uniquely capable of nurturing strong social organization; Swedish Social Democracy has benefited from this capacity and has nurtured its conception.
To understand this situation, we have to return to the mix of conditions inherited from the nineteenth century. Sweden was, until the beginning of the twentieth century, an isolated, industrially backward, and poor nation. It was relatively insulated from foreign influences, culturally homogeneous, and with a simple conflict structure. The matrix consisted of a small, albeit powerful, bourgeoisie, an emasculated landed aristocracy, a sizable but not dominant class of independent farmers, and a mass of small quast-pro-letarianized peasants. Industrialization coincided with the struggle for democracy and liberalization, a struggle that pitted the rural classes, the urban liberals, and the embryonic labor movement against industry and aristocracy, the latter strongly associated with the conservative party. The class confrontation was more apparent than in other nations because of the concentration of ownership within a small and highly visible group of powerful families.
The capacity of the peasantries and working classes to struggle effectively was aided by their high level of literacy and education, and thus by their ability to articulate and represent political demands, and by their capacity to unify and organize in collective organizations such as agricultural cooperatives and trade unions. The combination of geographical dispersion and cultural homogeneity was then and has been subsequently a chief inducement to the centralization and national coordination of interest representation. Within the respective social classes and strata, mobilization was made easier; free-rider problems were less important than elsewhere. More crucial for Social Democracy and the trade unions were the homogeneity and unity of their base. The unions never had to struggle seriously with internal divisions between craft segments, skilled, and unskilled workers; the party was never seriously wrought with ideological dissension, polarization, or fragmentation.
The traditional, preindustrial mechanisms of social integration were crucial for Social Democracy's ascent. Equally important, the way they were transplanted into modern industrial society favored the labor movement considerably. SAP's ability to count on an extraordinary level of loyalty, compliance, and social solidarity today has its roots in nineteenth-century Sweden, a society that was insulated from influences from abroad, organized around small, localistic, and cohesive communities in which solidarity and conformity not only came naturally but were strictly enforced. In large measure, the Swedish labor movement could employ this culture of preindustrial solidarity in its mobilization efforts during much of the twentieth
Single-Party Dominance in Sweden   I   43
century because urbanization came exceedingly late and the natural inclination toward social conformity was carried on from generation to generation. Traditionalist consensus mechanisms were easily transposed into a social democratic framework, and when combined with the structure of class-organizational affiliation, resulted in an extraordinary level of collective identity formation. Thus both the level of working-class identification and party identification are substantially higher in Sweden than in either Denmark or Norway.8
The blend of these sociocultural characteristics marks Sweden as a special case in Scandinavia and in Europe at large. They are imperative for understanding the rise of Sweden's extraordinary organizational network, without which it is difficult to imagine persistent Social Democratic hegemony. _ The LO is closely linked to the SAP (the TCO is independent) and has since the war been a key power resource for the party in attaining finances, party membership, and voter mobilization. Sweden is one of the few countries in which trade union members collectively affiliate with the labor party. Partly as a consequence, membership in the SAP was, and remains, unusually high. At the eve of World War II, SAP claimed 14 percent of all eligible voters; in 1976, the ratio stood at 25 percent. These figures contrast sharply with those in Denmark, where the party membership ratio dropped from 37 to 13 percent in the same period.9
Parallel to the organizational density of trade unionism and the party, Swedish Social Democracy has managed to erect an organizational empire that penetrates most of Swedish society, ranging from the cooperative movement, tenant organizations, and pensioner associations to education and leisure societies. As critics of Swedish Social Democracy argue, the labor movement has more or less organized the entire country.10 That is, of course, an exaggeration, but the point is that such a degree of penetration serves as a powerful resource for political mobilization, cultivation of Social Democratic ideals, and reproduction of social solidarity. Although quantitative comparisons are difficult to make, there seems no doubt that activity and participation rates in the Social Democratic empire are much stronger than in either Denmark or Norway.
The link between organizational mobilization and party strength is evident and can be tapped in various ways. First, the societal penetration of the labor movement (and other interest organizations) helps maximize electoral participation. In fact, Sweden consistently boasts one of the world's highest rates of electoral participation. Voter turnout is positively related to SAP's
8.  Diane Sainsbury, "Scandinavian Party Politics Re-examined: Social Democracy in Dechne." West European Politics 7 (1984): 84-88.
9.  Gurmar Sjoblom, "The Role of Political Parties in Denmark and Sweden," in Richard Küti, ed., Tbc Future ofParty Government: European and American Experiences (Berlin: De Grjyter, 1987), pp. 169-70.
10. Erik Anners, Den Socialdemokratiska Magtapparaten (Boraas: Askild and Karnekull 1976).
44   /   G0STA Esping-Andersen
electoral performance, the bivariate correlation being ,498 for the period 1918—81. Electoral turnout explains 20.6 percent of the variance of the party vote (T = 2.44).
A second and more important link has to do with the organizations' role in directly mobilizing party votes. The overwhelming majority of LO trade union members support the SAP, although in the 1970s a decline set in (75 percent supported SAP in 1970, 68 percent in 1979). Among the white-collar TCO unionists, the percentage of SAP support is lower but also more stable (about 38 percent). Indications are that the decline in unionist support was reversed after 1979.1T
A third phenomenon of profound importance for SAP's sustained dominance is its ability to mobilize optimally among both its traditional working-class clientele and "allied" classes. Of special significance has been its success in shifting from rural white-collar mobilization when, in the 1950s and 1960s, changing social structure compelled such a shift. Thus the SAP has been able to count on possibly the world's highest score of class voting among manual workers while simultaneously increasing the white-collar share. Until the party slipped in the mid-1970s, it could normally rely on about 70-75 percent of blue-collar votes and about 48-52 percent of white-collar (lower and middle-level) votes. Again, this contrasts with Denmark and Norway, where the decline in blue-collar support has been sharp since the 1960s and white-collar support has been stagnant and, in Norway especially, declining. The stability of class political allegiances is, of course, a mirror image of SAP's stable political record.
Nevertheless, the social structural and organizational conditions that facilitate Social Democratic mobilization have only limited explanatory power unless it can be demonstrated that the party, through its political alliance strategies and policies, is able not only to mirror and formulate mass demands but also to cohere potential or real interest differences. This is especially so because the SAP has never been able to rely solely on traditional working-class support but has perennially faced the necessity of presenting a political package that can mobilize across the class structure. The relative size of the manual working class (including agricultural workers) has been a rather stable 50 percent of the labor force, declining markedly in the last two decades. If we additionally posit that the SAP can realistically hope to capture up to 80 percent of its votes, the conclusion is clearly that the party is chronically captive to the logic of coalition building.
Formation of Political Alliances and Social Democratic Power
In a multiparty system such as the Swedish, parties tend to represent distinct clienteles. In Scandinavia, where the degree of interest organization
11. Sören Holmberg, Svenska Valjara i Forandring (Stockholm: Liber, 1984); Sainsbury, "Scandinavian Parry Politics Re-examined."
Single-Party Dominance in Sweden   I   45
among the various social groups and classes is so strong, rhe likelihood that the parties directly mirror an underlying and well-identified collectivized social stratum is additionally enhanced.
Thus, from the cradle of democracy until today, the party system has been pretty much a replica of the social structure. The conservative parties represent business and industry; the farmers' parties represent farmers; the liberal parties, urban professionals and the middle class; and the social democratic parties came, naturally, to represent the working classes. Since no particular stratum is numerically dominant, the result is that either minority or coalition cabinets prevail.
The Swedish Social Democratic party has, since the early 1920s, been the largest party. But except for two extraordinary occasions (1940 and 1968), its electoral share has never surpassed the magic 50 percent mark. From a European perspective, there is nothing unusual in this, nor is SAP's high share of the total vote (and parliamentary seats) unique; a more or less similar share has been maintained by the Austrian socialists.
What is uniquely Swedish is the terrain for the formation of political alliances. In part because of class structural circumstances, and in part historical fate, the Nordic countries in general, and Sweden in particular, rendered a broad conservative-dominated social coalition difficult. In the nineteenth century, the urban liberals (often in liaison with the farmers) were pitted against the conservatives; in the early twentieth century, as the socialist movement rose, conflicts among the nonsocialist parties remained so strong that they were unable to unify against the socialist threat even when they tried. This failure to unify juxtaposes Scandinavia to most of the continental European nations and helps explain the ease with which Social Democracy could come to power, as well as the immunity of the Nordic countries to reactionary and fascist responses during the 1930s.
Over the postwar era, the nonsocialist bloc was unable to replicate the European model of broadly based, conservative mass parties such as the CDU in Germany, DC in Italy, or the Gaullist alliance in France. Instead, the prewar matrix of a fragmented center and right remained pretty much intact.12 Thus the Scandinavian Social Democratic parties were uniquely positioned to play the Machiavellian and, being the single largest parties, remained the only really plausible contenders to power. A decisive difference between Sweden, on one hand, and Denmark and Norway, on the other, is that in the two latter nations, the bourgeois parties managed to forge a more united front from the mid-1960s onward.
There is no doubt that the SAP in Sweden has benefited enormously from the chronic inability of the nonsocialist bloc to govern in unison. SAP's six years in opposition confirm this tendency, as the "bourgeois" coalition Hrst broke on the issue of nuclear power and, in its second term, on taxation. In subsequent elections, the coalition's incapacity to unify has been recon-
12. Francis Castles, The Social Democratic Image of Society (London: Routledge & Kegan Paul, 1978).
46   /   G0STA Espdjg-Andersen
firmed; during virtually every electoral campaign, including the 1988 campaign, the three nonsocialist parties have been preoccupied more with combating one another than with attacking the Social Democrats. In contrast, both in Denmark and Norway bourgeois coalitions fared well in the late 1960s and in the 1970s and 1980s.
It would misrepresent reality, however, to argue that SAP's political dominance occurred by default. Social Democracy has played an active role in nurturing rightist divisions, and its dominance has been possible because of its own coalition strategies. Social Democracy could not have held power for half a century on the basis of rightist splits alone.
A common Scandinavian characteristic is the historical legacy of cross-class alliances for democracy in the nineteenth century and for an anticrisis policy package again in the 1930s. The postwar evolution of this alliance took different forms in the Nordic countries. In Denmark, the Social Democrats failed to gain dominance precisely because the Agrarian party (Ven-stre) was too strong; in Norway, the Social Democrats became preeminent, but only under the condition that their policies remained favorable to the needs of the rural and fishery-based interests in the peripheries. In Sweden, in contrast, the SAP found itself compelled to reinvigorate the red-green alliance during the 1950s as the only means of staying in power. The SAP was the overwhelmingly dominant partner in the alliance (unlike Denmark), but the Agrarian party nevertheless came to dictate the limits of permissible reform. This constraint, paradoxically, favored Swedish Social Democracy's future.
The SAP emerged from the war with a fairly radical and ambitious welfare state-cum-full-employment program that clearly superseded the limited accord made with the Farmers' party in the 1930s. The SAP's electoral success in the 1940s would have seemed to warrant a dashing and aggressive program. Yet postwar circumstances swiftly eroded these conditions, and the party slipped back to its 46 percent level. The necessity of a continued alliance with the farmers meant that the SAP's reformist zeal was seriously dampened in the 1950s, which seems to have weakened its electoral fortunes.
It was the interrelation of class structural change and the increasingly burdensome alliance with the fanners that reversed SAP's fortunes. Political realignment crystallized with the conflict over pension reform in 1957 to 1959. In part, the realignment occurred because the farmer and peasant population shrank dramatically during the 1940s and 1950s as Sweden embarked on a wave of urbanization and because the Farmers' party refused to support the LO-SAP pension reform bill. Most important was SAP's understanding that long-term power would have to rest on its ability to mobilize the growing white-collar strata. The pension issue was uniquely tailored for a political realignment that entailed a break with the farmers and an opening toward the new middle classes. The SAP, unlike its Nordic
Single-Party Dominance in Sweden   I   47
brethren, was thus in a position to transform itself from the old "peoples" party to the new "wage earner" party model. A parallel realignment was not feasible in the other two countries, in Norway because a confrontation with the periphery interests would seriously harm any party, be it left or right and in Denmark because the fanners' interests remained politically dominant well into the 1960s.
A peculiar characteristic of the SAP's shift from a farmer to a middle-class coalition is that the rising white-collar class (although increasingly unionized) was largely unaffiliated with any given party. Aside from upper-echelon professionals and civil servants, who traditionally cast their loyalties with the conservatives, and the lowest ranks of routine white-collar occupations, who had a tradition of supporting Social Democracy, the new class was politically up for grabs. Only the small centrist Liberal party could make any real claim to representing the middle classes. In this context, the coalition-building strategy would turn from the tradition of parliamentary alliances between parties to that of mobilizing the middle-class vote internally.
Although the SAP never made a clear decision on these two alternatives, subsequent events were decisive. The pension conflict, combined with SAP's highly successful attempt to appeal to the middle strata, resulted in a major rise of electoral support and party membership during the 1960s. Indeed, the SAP came to enjoy single-party majority status in 1968. As a reciprocal effect, SAP's success meant that the nonsocialist parties' chances to mobilize the critical new middle-class electorate were stalled.
The new middle-class realignment of Social Democracy was evident not only in party electoral fortunes but also in the country's organizational matrix. Although retaining their formal political independence, the basically white-collar TCO unions were moving closer and closer to the LO as well as to the SAP's policies. And the heavily public sector bias in the growth of white-collar jobs strengthened the ties of that group to Social Democracy. Yet the middle-class coalition that appeared so successful in the 1960s proved to be somewhat fragile. A number of factors conspired to stall its full institutionalization. First, issues such as aggressive income and status equalization became an Achilles' heel for Social Democracy because they were strongly demanded by SAP's traditional base yet faced widespread opposition among the middle strata. Second, New Left issues such as the environment and nuclear power (which the SAP promoted) pulled the middle strata toward the reconstituted and "intellectualized" small Communist party and the Center party. Third, the economic democracy plan for collective wage earner funds that SAP promoted in the 1970s was far from popular among white-collar groups (or even workers).
The result was that SAP's potential for mobilization eroded in the 1970s; the party's capacity to forge a sufficiently viable middle-class alliance internally was thwarted. One can view the SAP's decision to incorporate the
48   /   G0STA Esping-Andersen
Communists as an unofficial alliance partner after 1973 as a functional equivalent to single-party rule on die basis of a worker-middle-ciass mix; yet as long as the nuclear power issue remained, SAP's ability to govern continued to erode. The relegation of the nuclear power question to a popular referendum in 1980 helped remove it as a Social Democratic Achilles' heel.
It would be incorrect to say that the SAP's middle-class coalition collapsed during its opposition years, 1976—82. The party's tosses to the right included both workers and white-collar groups (although the latter were more likely to shift); more to the point, the electoral shift was by no means a landslide. The return of both workers and the middle classes in the 1980s suggests that the long-run coalitional status of white-collar groups remains uncertain. Their return to Social Democracy can be interpreted as a disenchantment with the failed right. This suggests a weak case for SAP's mobili-zational capacity, or it can be interpreted as a new enthusiasm for Social Democracy in light of its convincing performance in bringing the Swedish economy out of its severe difficulties in the 1980s. It is still too early to judge. Yet a scrutiny of SAP's policies and their electoral impact should aid us considerably in understanding the party's short- and long-term mobiliza-tional capacity.
The Institutionalization of Power through Policies
In the final analysis, for several reasons, party dominance is a function of successful policies. First, voters may be persuaded by ideological promises for a period—but not in the long haul. Labor parties, in particular, are expected to satisfy demands for social and economic betterment, security, and equalization. They must be able to improve material conditions.
Second, social reforms are never neutral but have long-term implications for social stratification and the balance of political power. Policies may be motivated by immediately pressing problems, and their implementation helps institutionalize interests. Thus social policies may solidify class and status differences; or they may stigmatize and foster dualisms; or they may cultivate universalism and broad solidarity. As we shall see, a precondition for Social Democratic dominance is its ability to institutionalize universalism. Its success in this area helps secure broad popular solidarity behind the welfare state and, as a consequence, shelters it against tax resistance.
Third, the coalitional necessity that Social Democracy continuously faces means that it must be capable of presenting positive-sum policy packages, that is, a policy mix that benefits both workers and the potentially allied classes in a Pareto-optimal way.
Swedish Social Democratic dominance rests ultimately on its capacity both to launch and to institutionalize a social democratic society via re-
Single-Farty Dominance in Sweden   I   49
forms. The SAP succeeded, as nowhere else, in creating a universalistic welfare state, and it managed to install a general consensus behind full employment in large part because that goal was secured without incurring substantial economic costs. On one hand, class structural and coalitional conditions were favorable for full employment; on the other, the causality is reversed: SAP represents a case in which policies have facilitated electoral mobilization and have reproduced Social Democracy's power. A survey of its policy record since the 1930s will illustrate this point. The discussion will hinge on SAP's combined social and economic policy achievements.
As we have seen, the 1930s provided the stepping-stone for social democratic power in the Nordic countries. But as events in Denmark and Norway suggest, this alone offered no guarantees for the long run. Was there anything unique about SAP's performance? The answer is both yes and no. A number of conditions existed that were especially advantageous for Swedish social democracy. First, in comparison to Denmark and Norway, the issues of democratic and social stability were less settled in Sweden between the wars. Social democracy's consolidation of power in the 1930s helped put an end to industrial strife and labor market unrest, reaffirmed social democracy's image as the ultimate guarantor of democracy, and, above all, the red-green alliance affirmed the party's status as a party of the people, not of a class. To quote the party leader, Per Albin Hansson, the party became the carrier of the "Peoples Home" (Folkhemmet).13 As such, SAP was a consensus builder, a party committed to the national interest, a party capable of rising above sectionalism and narrow class egoism.
Second, the SAP was privileged by inheriting a clean slate with regard to policies. The legacy of social reform from previous administrations was extraordinarily thin. Indeed, if Sweden today ranks as the world leader in welfare statism, its status in the 1930s was that of an international laggard. Unlike in other nations, the Swedish conservatives and liberals had not embarked on the strategy of preempting and stalling labor militancy with "social pacifist" reforms. In Sweden, therefore, a political platform based on social policy and full employment was thus a response to exceptionally dire need, expecially in light of the 1930s Depression. The promotion of unemployment insurance and employment creation programs in combination with agricultural relief became a rallying point of substantial electoral promise.
Third, Swedish Social Democracy profited tremendously from pure and simple good fortune. With the formation of the Red-Green alliance, the party succeeded in implementing its chief reform objectives: unemployment insurance, employment creation, farm subsidies, and its internationally vanguard countercyclical stimulation policies. The SAP adopted the Keynesian
13. A. L. Berkling, Fraan From till Folkhemmet: Per Albin Hansson som Tidningsman och Taláre {Falkoping: Metodica, 1982).
SO    /   G0STA Esping-Andersen
prescription before the General Theory had been published. In reality, the deficit-financed expansionary program was modest and cannot account for Sweden's rapid economic recovery in the 1930s.14 On one hand, economic conditions were favorable even before SAP came to power because Sweden had gone off the gold standard in 1931 and the currency remained undervalued. On the other hand, the rapid decline in unemployment after 1933 occurred mainly because of the rapid rise of Swedish exports, which were predominantly targeted to Hitler's vast rearmaments drive. The paradox is that Swedish Social Democracy benefited from the Nazi dictatorship.
Nevertheless, SAP reaped electoral benefits from its policies. It was spared the trauma of cutting back existing welfare measures, as the German socialists and the British Labour party were forced by the Depression to do, and it was instead in the position of implementing vast improvements. The resurgent economic growth improved industrial profits, farm prices, employment, and incomes. Social Democracy had proven itself capable of governing and, with rising popular support (it gained more than 4 percent in the 1936 election), had confirmed its status as a party likely to remain in office. Thus industry was forced to move from its traditional oppositional stance toward one of accommodation.
Thus emerged another key element in SAP's subsequent dominance, the social accord between capital and labor in the 1930s. This accord consisted of two more or less officially sanctified agreements: the famous Salts-jobaden accords between the LO and the Employers' Association, which de facto put an end to industrial strife and secured a long era of regularized and peaceful collective bargaining; and the results of negotiations between labor leaders and industrialists, which fostered mutual trust.15 The bourgeoisie came to accept the idea of equalization via the welfare state; labor promised to honor the sanctity of private property rights.
Swedish Social Democracy's performance in the 1930s was therefore more auspicious than elsewhere from the point of view of harvesting long-term political dividends. Yet there is nothing globally unique in this success; the legacy of Franklin D. Roosevelt in the United States appears in many respects even more spectacular. The question turns to how the SAP succeeded in reproducing the position it had gained in the 1930s.
The period that followed invited serious problems. During the war, Swedish neutrality meant that the labor movement could not (as in Norway) harvest sympathy and loyalty on the basis of heroic resistance to Nazism. Indeed, the SAP coalition government compromised its ideals by allowing German troop transports across Swedish territory.
14.  Lars Jorberg, "Industrialization in Sweden," in Carlo Cippola, ed., The Economic History of Europe (London: Fontana, 1976); Erik Lundberg, Instability and Economic Grotvth (New Haven: Yale University Press, 1968); Assar Lindbeck, Svensk Ekonomisk Politik (Stockholm: Bonniers, 1975).
15.  Sven Anders SÖderpalm, Direktörsklubben (Stockholm: Raben Si Sjögren, 1976).
Single-Party Dominance in Sweden   I   51
In addition, in 1944 the party launched a major program revision that consolidated its status as a reformist party focused on welfare reform and full employment while, concomitantly, promoting a fairly radical model of economic planning. The latter had to be shelved soon after the war.
Yet the welfare state—cum—full-employment promise soon presented the party with difficulties. Although most of the ambitious welfare reforms (especially pensions) could be financed by reducing defense spending, the commitment to full employment soon confronted the classical Phillips curve dilemma; by 1948—49, wage pressures and inflation severely threatened the balance of payments and economic growth. SAP's response, as in Britain, was to impose a two-year wage freeze on the trade unions. This freeze helped stabilize the economy, but it severely jeopardized political stability. The trade unions were unwilling to accept further political measures that influenced income and feared a likely future scenario of stop-go policies. In other words, the SAP government faced a zero-sum problem.
The solution came from rhe trade union movement in the form of the now world-famous Rehn model for an active labor market policy. The adoption of this model ensured Social Democratic capacity to govern over the next decades by sparing the party (government) from having to intervene frequently in wage negotiations and thereby avoided serious tensions between unions and the party.
The Rehn model, named for its architect, Gösta Rehn, a trade union economist, was a brilliant positive-sum solution to the Phillips curve problem: based on the assumption that the unions pursue an aggressive and equalizing wage policy, the model called for government-sponsored programs of mobility and retraining so as to absorb labor made redundant in declining industries and transfer it to dynamic growth sectors. In brief, the model assured both equality and efficiency: it promoted equality and full employment, and it helped promote industrial rationalization and restructuring without costs to individual workers. Even skeptics of the social democratic model acknowledge the success of the Rehn model in promoting balanced economic growth during the postwar decades.10
The active labor market policy became Swedish Social Democracy's answer to the perennial efficiency-equality dilemma that confronts labor governments. The party could present itself as an effective guarantor of both ideals.17 The political opposition, in consequence, was prevented from claiming that socialist policies would destroy the economy.
Nevertheless, the SAP's electoral performance began to slip during the 1950s. It suffered a marked loss in the 1956 elections. The reasons were not immediately identifiable, but, in hindsight, the problem was SAP's continued reliance on its farmer allies, which prevented it from pursuing some
16. Lindbeck, Svensk Ekonomisk Politik.
1~. Theodore Geiger and Frances Geiger, Welfare and Efficiency (Washington, D.C.: National Planning Commission, 1978).
52   /    G0STA Esping-Andersen
of its more ambitious social reforms (health reform was delayed by many years, for example) and forced upon SAP a noticeable ideological and programmatic caution and even lethargy. To its supporters, the SAP would have appeared increasingly bland, technocratic, and tired. The turning point came with the proposal for a major pension reform. If the active labor market policy was the cornerstone for SAP's power in the economic arena, the pension policy came to occupy the same position in the social arena.
At first, the issue would appear relatively uncontroversial; it involved the introduction of an earnings-related second-tier pension system. Yet it soon became one of the most conflictual issues in postwar Sweden. Urged by the trade unions, SAP presented in 1957 a bill for a supplementary pension (usually referred to as the ATP pensions). This bill would hardly have provoked conflict were it not for its design and radical implications. First, in line with SAP's traditional social policy model, it called for compulsory and universal membership—a principle explicitly designed to foster broad national solidarity and equality of status. Second, in contrast to the employers' counterproposal, it was to be legislated (and thus not a voluntary, private occupational plan) and financed with publicly controlled pension funds. The method of funding was undoubtedly the key issue because the employers as well as the nonsocialist parties correctly saw the hugely funded system as a shift of capital control from private to public sector authority.
The lines of conflict over the ATP pension issue were drawn very sharply between left and right. But the pension issue invoked far more than a question of pensions and credit markets; it became a question of Sweden's future political power balance. The issue forced the SAP to break with its erstwhile farmer allies but offered it a unique opportunity to mobilize the rising white-collar strata, which increasingly came to favor the SAP's pension proposal as the conflict dragged on in the late 1950s. The ATP pension issue was a political turning point for Social Democracy.
This is not the place to review the long pension conflict. It involved both the resignation of a government and a national referendum and was finally resolved with a one-vote majority in favor of the SAP in 1959.1S The important point is that it fostered a coalitional realignment at the most opportune time, just when the historical importance of the farmers had decayed and the new middle classes began to assert their electoral clout. Not surprisingly, the SAP underwent a substantial programmatic and ideological reorientation. Its traditional populist Peoples Home image was shelved in favor of a new image as a wage earner party. Its electoral success during the 1960s bears witness to its new-found ability to attract white-collar votes.
The SAP was invigorated by its policy achievements. It had proven itself capable of synthesizing economic efficiency and social equality concerns; it had succeeded in coalescing the working class and white-collar groups; and it presided over an economy that became internationally celebrated for its
18. Björn Molin, Tjänstpensionsfragan (Lund: Akademisk Forlag, 1967).
Single-Party Dominance in Sweden   I   53
ability to ensure sustained full employment, welfare, equality, social peace, international competitiveness, and material growth.
This positive policy record was lacking in both Danish and Norwegian Social Democracy; Danish Social Democracy was the least capable of securing a similar positive-sum solution. Although Danish Social Democrats proposed policies almost identical to those of their Swedish brethren, Denmark's political conditions were such that they were systematically stalled. Jn economic management, there was nothing equivalent to the active labor market policy; instead, the Social Democrats were continuously compelled io intervene with incomes policies, which alienated the trade unions. On the social policy front, they were unable to pass a supplementary pension; the result was proliferating and highly inegalitarian private pensions. The net effect in Denmark was that Social Democracy had little capacity to win over the new middle classes and thus even less capacity to succeed with its reform intentions.
An irony of Swedish Social Democracy is that its celebrated model began to unravel exactly when the SAP harvested its electoral dividends. By the early 1970s, the positive-sum formula was gradually turning negative-sum. The signals appeared everywhere: industry faced declining profits, and investment levels began to fall, jeopardizing the commitment to full em-pioyment; workers were increasingly unwilling to participate in the active manpower policy system because it often meant unwanted geographical relocation; the equalizing wage-bargaining system faced growing resistance from the worker "aristocracies," who saw that their wages were slipping relatively; and with a slowdown of economic growth, higher inflation, and welfare state consolidation, tax rates grew very rapidly in the early 1970s. The formula that had so successfully carried Social Democracy through the postwar era had become contradictory. And it is this turnabout that accounts for SAP's losses during the 1970s.
The SAP's immediate responses actually aggravated the problems. Escalating wage costs and demand pressures began to jeopardize exports and competitiveness; SAP's response was to manufacture a relatively severe recession from 1971 to 1973, which, in turn, led to acute overloading of the active manpower programs, which had to absorb rising unemployment; and ii led to a subsequent wage explosion. By 1974-75, the economy was in acute imbalance; public expenditures grew tremendously as did wages and, of course, taxation. And with the onset of the international economic crisis after 1973, the government's chief concern came to be focused on maintaining full employment.
The means available for maintaining full employment were limited to the absorptive capacity of the active manpower programs, expansion of public sector employment (in welfare state services), and industrial subsidies. The government did succeed in retaining full employment but at enormous public budgetary costs.
Sodal Democracy's lack of a convincing policy alternative in the face of
54    /   G0STA Esping-Andersen
these accumulated problems may not directly explain its substantial electoral losses in 1976. But it undoubtedly furnishes the backdrop. There were two ad hoc issues that clearly weakened the party in the 1970s. One was the controversy over nuclear energy. The trade unions and the SAP had been strong supporters of a nuclear power buildup in Sweden and were unprepared for the rising popular backlash. "When it came (led by the leftist Communist party and the erstwhile Agrarian party, now the Center party), SAP was vulnerable and lost considerable votes on the issue in the 1976 elections. Second, the Swedish LO had developed a plan, written by Rudolf Meidner, for economic democracy through collective wage earner funds as a means to salvage its commitment to wage equalization and as a new thrust toward democratization of the economy. The economic democracy issue came to a head in the 1976 election because of its controversial nature and the abjectly clumsy way it was handled. Because economic democracy became the key issue of the 1970s and 1980s, that issue requires some explanation.
Economic democracy was of double importance to the trade unions. First, the unions had, since the 1960s, demanded greater control over investments and greater democratic control of the economy in general. The Meidner plan intended to tax a percentage of profits to be reallocated in investment funds collectively owned by workers; indeed, it was designed so that, over four or five decades, workers would control a majority share of Swedish enterprise. In other words, the plan called for a creeping socialization of the economy. Second, the economic democracy proposal served labor's more immediately pressing problems of collective bargaining, especially the resistance to continued wage equalization among large sections of the labor force, who saw that their wage moderation fueled excessive profits. LO's intention was to preserve wage equality while granting workers, as a collectivity, an indirect compensation in the form of share ownership.
It is not difficult to see that the economic democracy plan explicitly broke with the social accord of the 1930s. Its incursion into the prerogatives of private ownership implied that labor demanded a renegotiation—one that clearly invoked terms unacceptable to employers. The immediate response of the employers was, of course, extremely hostile. The LO passed the plan on to the SAP for legislative consideration in the middle of the 1976 electoral campaign, and the political focus was naturally shifted toward the Social Democratic party. This tactical clumsiness accentuated SAP's weakness in the elections, and the issue was expertly exploited by the nonsocialist parties, which could successfully rally on the slogan of "socialism or freedom."
The 1976 elections gave the SAP only 42.8 percent of the vote, in Swedish terms a painful loss. The loss cannot be ascribed to one single issue— electoral survey data suggest that party shifts were caused by a variety of factors, among which nuclear power and economic democracy certainly
Single-Party Dominance in Sweden   I   55
n'gured prominently.19 Growing skepticism about Social Democracy's capacity to manage the economy was clearly evident, and a majority of voters began to believe that the nonsocialist parties would be better able to guarantee continued full employment.20
At first, the 1976 elections may have appeared to terminate Swedish Social Democratic dominance, especially when, in the 1979 elections, the SAP was not returned to power. The 1976 elections were the first in which the bourgeois bloc was able to unify and take office. But the two terms the SAP spent in opposition did little to alter its de facto political dominance. The nonsocialist coalition proved itself incapable of sustained unity; it broke apart both in its first and its second terms, resulting in weak minority cabinets. After the coalition's first term, 1976-79, SAP's electoral share increased moderately; following its second term, SAP returned to its normal 45 to 46 percent share, and it retained this level in the 1985 elections. In the 1988 elections, SAP declined by 1 percent, but with implicit Communist party support and the bourgeois parties as divided as ever (and further emasculated by the emergence of the Green party), SAP's grip on government power remains as solid as ever. Hence, viewed in electoral terms, the six years in opposition would appear as a temporary and transitory experience. What is more important is the nonsocialist coalition's policy performance while in office.
The bourgeois coalition was clearly captive to the hegemony of the Social Democratic agenda. It took office in the midst of the economic crisis and was forced to expend the lion's share of its attention on the maintenance of full employment. To this end, it continued—by and large—Social Democracy's previous, albeit questionable, policies; the accent was on massive subsidization of industries and jobs. Concomitantly, it was both unwilling and unable to cut welfare state programs. The result was, predictably, escalating budget deficits and growing imbalances in the economy. Thus it was Social Democracy which, in the end, could win elections by persuading the electorate that it was the only political force capable of responsible economic management. After the nonsocialist bloc's second term in office, the voters faced an odd political choice between efficient and inefficient social democracy; that they opted for the latter is not surprising.
Conclusion
The account given thus far in this chapter seems to suggest that SAP's return to power in 1*982 was largely by default. In a superficial sense, there is some truth to this argument: its opponents were unable to present an
19.  Olof Petersson, Valjarna och Valet 1976 (Stockholm: liber, 1977).
20.  Ibid.; Esping-Andersen, Politics against Markets.
56   /   G0STA Esping-Andersen
attractive or even convincing alternative. And SAP's return in 1982 was not founded on a workable and consensual new political formula with which to build a viable class coalition and thus launch itself on a new era of undisputed party dominance.
The SAP was brought back in office because its capacity to manage the economic difficulties was more credible, not because the party had anything new to offer. Granted, the 1982 elections included the controversial economic democracy proposal (in a significantly watered-down version), but this is certainly not what accounts for its reinvigoration. All accounts indicate that the proposal remains fairly unpopular among the party's working-class base and much more so among the white-collar strata.21 It distinctly alienates the employers.
The considerably modified economic democracy plan was passed in parliament with support from the small Communist party in December 1982. In the new version, the plan has a built-in maximum limit to capital accumulation in the wage earner fund, and it stipulates that the fund's investments take place in the open, private market. The past years' experience with economic democracy does not suggest that it has had a significant electoral or coalitional dividend; nor has it brought major changes to the Swedish economic system.
Instead, SAP's return to political dominance rests by and large on its respectable economic policy performance along conventional policy parameters: it has been able to calm wages and inflation, and Swedish international competitiveness has been restored. The huge deficits have been transformed into surpluses.
Why was Swedish Social Democracy able to continue to rule even under a bourgeois flag, discredit the opposition, and then return seemingly unchallenged? The real answer has not to do with marginal electoral shifts or party credibility but with how Sweden became thoroughly social democratized over a period of many decades. The answer must, in the end, be found in the interplay of the three factors examined in this chapter: Sweden's social structure, the mobilization of citizens and political coalitions, and SAP's ability to institutionalize its power by enacting reforms. These three factors have come to interact in such a way that Social Democratic power is virtually impossible to break.
Social Democracy's policies became the nexus around which this favorable interaction emerged. The critical question is not SAP's day-to-day, or year-to-year, political management of the social economy but its policies' lasting influence or- shaping social structural evolution and on coalition building. The key is SAP's consistent adherence to the principle of universalist inclusion. Its welfare state empire is one in which virtually every single program embraces and benefits more or less the entire populace. The wel-
21. Sören Holmberg, Svcnska Valfara (Stockholm: Publica, 1981).
Single-Party Dominance in Sweden   I   57
fare state thus actively constructs a social structure in the image of social democracy: it overrides cleavages, dualisms, and differentials by cementing equal status of citizenship across a very broad array of vital human needs, lne welfare state, in turn, is assured optimal support because virtually everyone benefits directly and because the welfare of the majority is directly dependent on it. This explains why Sweden, the most extreme case of welfare state expenditure and taxation, has yet to experience even a modest tax revolt. Even a marginal tampering with welfare programs in Sweden will im mediately unleash mass furor, as the bourgeois coalition experienced in its last term.
The Swedish social democratic welfare state is, unlike welfare states almost everywhere else, a consensus-building institution. It not only supplements the citizenship of equal legal and political rights; it is viewed as actively reinforcing them. In this sense, it has become an indispensable collective good and a definitional part of Swedish nationhood. It is the Swedish equivalent of the Italian mamma.
Nevertheless, the spread of universalism alone would clearly not suffice to explain Social Democratic dominance. Over the past two or three decades, and for many to come, SAP's position depends on middle-class support. Thus the public benefits must be attractive to the discriminating tastes of high-income groups as well as to workers. It was this issue that came to the fore with the conflictual ATP pension controversy in the late 1950s and which Social Democracy successfully solved. The SAP managed, in brief, to build a welfare state that, at once, pleased workers and the middle classes. It did this by synthesizing the principle of adequate benefits {through earnings-graduated benefits and high-quality services) and equality (through universal and equal rights and status). Powerful middle-class solidarity behind the welfare state and social democracy is not surprising when we realize that its distributional consequences are advantageous to the better-off.22
In summary, Swedish Social Democracy's extraordinary political dominance rests on its capacity to count on a very broad social solidarity; and it is able to do so because the SAP itself largely constructed it. Accordingly, we reach a conclusion that sounds tautological: Swedish Social Democracy is dominant because Sweden has become thoroughly social democratized. But the tautology disappears when we disentangle historical forces of social democratization; these do not refer just to party power and election victories but to the interaction effect of structural properties inherent in Swedish society, the organizational features .of political mobilization, especially coalition building, and the institutional characteristics of policy development.
22. A. C. Stahlberg, ATP Gynnar HoginnkomsHagarna (Stockholm: Institute for Social Research Reprint, 1981).
2. Conditions of Labor-Party Dominance: Sweden and Britain Compared
Jonas Pontusson
Among reformist labor parties operating in advanced capitalist societies, Sweden's Social Democratic party clearly stands out as the most successful. From 1932 to 1976, Sweden had but three prime ministers, and they were all Social Democrats. The party's share of the popular vote slipped in the 1970s, but only slightly. Following six years of coalition government by the "bourgeois" parties, the Social Democrats restored their claim to be the "natural party of government" by winning three consecutive elections in the 1980s (1982, 1985, and 1988). Their share of the popular vote in these elections ranged between 44 and 46 percent, as compared to an average of 46.7 percent for nine parliamentary elections from 1944 to 1970.1
Arguably, Sweden represents a limiting case of labor reformism in advanced capitalism. A great deal has been written about what the Swedish Social Democracy has or has not achieved in income redistribution, welfare provision, employment, control of investment, and other areas. In this chapter, I pursue a different question: how and why did Social Democracy come to assume such a dominant role in Swedish politics? The existing literature tends to answer this question by describing the political ascendancy of Swedish Social Democracy. A more adequate answer presupposes a system-
I rewrote this chapter while on a grant from die German Marshall Fund of the United States, and I wish to acknowledge its support. For comments on earlier drafts, 1 thank Miriam Golden, Geoffrey Garrett, Peter Hall, Peggy Kahn, Peter Katzenstein, Leo Panitch, T. J. Pempel, Michael Shalev, Sven Steinmo, Sidney Tarrow, Vincent Wright, and anonymous reviewers.
1. The Social Democrats' best postwar election was 1968, when they polled 50.1 percent of the vote, and their worst postwar election was 1976, when they polled 42.7 percent. {The election of 1968 was exceptional, occurring less than two weeks after the Soviet invasion of Czechoslovakia in 1968.) Following Swedish convention, the term bourgeois parties is here used as a collective label for the three parties to the right of the Social Democrats: the Center party, the Liberals, and the Conservatives. Also, it should be noted that the Social Democratic tenure in government was interrupted for about three months in 1936.
58
Conditions of Labor-Party Dominance   I   59
atic comparison with countries in which reformist labor parties have been less successful. In the preceding chapter, G0sta Esping-Andersen tries to specify the determinants of Social Democratic dominance by focusing on the features that distinguish the Swedish case from the Norwegian and Danish cases. The following analysis conforms to a similar logic but instead uses Britain as the foil shedding light on the political success of Swedish Social Democracy.
From a strictly social structural point of view, the British Labour party has enjoyed more favorable circumstances than any other party purporting to represent the interests of workers as a class. For no other country ever became as thoroughly proletarianized as Britain, the first nation to industrialize. Whereas manual workers accounted for 69 percent of the gainfully employed population of Britain at the time of the 1951 census, the corresponding figures for Sweden were 41 percent in 1930 and 39 percent in 1965.2
If Swedish Social Democracy could become the "natural party of government," then surely British Labour should have been able to assume this role as well. Britain seemed slated for one-party dominance in the wake of Labour's landslide victory in the election of 1945, but the Tories returned to power in 1951. The Labour governments of 1964-70 and 1974-79 also failed to create the conditions for durable Labour rule. The Labour party's share of the popular vote dropped precipitously from 1966 (48.1 percent) to 1983 (27.6 percent) and recovered only slightly in 1987 (31.5 percent). After three consecutive defeats at the hands of Margaret Thatcher, it is an open question whether Labour will ever again be able to form a majority government.
There are two possible interpretations of the divergent fortunes of Swedish Social Democracy and British Labour. One interpretation would be that the divergence dates to the 1970s (or late 1960s) and derives from more favorable economic circumstances and/or more competent leadership in Sweden. According to this view, a similar welfarist/Keynesian consensus was established in both countries after the war. In the face of economic adversity, the consensus broke down in Britain but survived in Sweden. Alternatively, one might argue that Sweden's "postwar consensus" was quite different from Britain's and that this accounts for its greater endurance. Britain's commitment to an overvalued pound precluded a consistently Keynesian approach to economic management; more important, postwar Swedish governments went beyond the liberalism of Keynes and Beveridge to promote economic efficiency and social equality.
2. Figures from Perry Anderson, "Problems of Socialist Strategy," in Perry Anderson and Robin Blackburn, eds., Towards Socialism (London: Fontana, 1965), p. 248, and Goran Therborn, "The Swedish Class Structure," in Richard Scase, ed., Readings in the Swedish Class Structure (Oxford: Pergamon, 1976), p. 155. The British figure includes those self-employed engaged in manual labor.
60   /   Jonas Pontusson
Table 2.1. Voting support for left parties among different categories of wage earners: Sweden and Britain compared, 1964 and 1979 (in percent)
	Britain	i	Sweden*	
	1964	1979	1964	1979
Sal aria t Routine nonmanual Manual	20 31 70	22 36 55	8   (8) 47 {AS) 77 (72)	27 (20) 41 (37) 71 (66)
Sources: Anthony Heath, Roger Jowell, and John Curtice, How Britain Votes (Oxford: Pergamon Press, 1985), pp. 32-33; and Walter Korpi, The Democratic Class Struggle (London: Roudedge & Kegan Paul, 1983), pp. 241-42.
*The first Swedish figures refer to votes cast for the Social Democratic and Communist parties; the figures in parentheses refer to votes cast for the Social Democratic party.
Proponents of the latter interpretation might justifiably speak of Sweden as an instance of labor "hegemony" in capitalist society. The notion of labor hegemony is a slippery one, however, I shall not try to nail it down here. Elsewhere, I have tried to demonstrate that the Swedish labor movement has been able to determine the issues on the political agenda and set the terms of public policy debate to a far greater extent than its Brirish counterpart.3 Table 2.1 brings out a related point: it shows that Sweden's two socialist parties have together mobilized a significantly larger share of the votes cast by both blue-collar and white-collar voters than the British Labour party.4 For manual (blue-collar) and routine nonmanual voters, this was true in 1964 as well as 1979.
The British Labour party relies more heavily on the votes of manual workers than does the Swedish left and has been more adversely affected by the decline of the manual working class relative to the electorate as a whole. Nevertheless, the Labour party's share of the blue-collar vote dropped sharply in the 1970s. The British Labour party seems to have been caught in the dilemma stipulated by Adam Przeworski and John Sprague: reformist socialist parties must appeal to nonworkers to gain an electoral majority, yet such appeals undermine their support among workers.5 As Table 2.1
3.  Jonas Pontusson, Swedish Social Democracy and British Labour [Ithaca: Western So cieties Occasional Papers, Cornell University, 1988), chap. 2.
4.  It seems appropriate to include votes cast for the Swedish Communist party in r:-.:;-comparison. The party is considered part of die labor movement and has invariably voted for legislation proposed by Social Democratic governments whenever a unified bourgeois opposition would otherwise have prevailed. The Communists' share of the total vote averaged 5.2 percent from 1944 to 1970.
5.  Adam Przeworski and John Sprague, Paper Stones (Chicago: University of Chicago Pn-w, 1986).
Conditions of Labor-Party Dominance   I   61
indicates, the Swedish left also gained white-collar votes at the expense of bluŕ-collar votes in the period from 1964 to 1979, but the trade-off appears to have been less steep. At any rate, it should be clear from the table that the combination of strong working-class support and cross-class appeal constitutes the key to the electoral strength of Swedish Social Democracy.
The electoral mobilization capacity of Swedish Social Democracy can be treated as a cause as weil as a consequence of its control of the political agenda. In postwar Sweden, electoral competition has primarily revolved around the support of the white-collar strata. "Whereas the Social Democrats' hold on the blue-collar vote has never been seriously challenged, the bourgeois parties have, had to respond to Social Democratic initiatives to defend their core constituencies. In Britain, by contrast, the Labour party has been on the defensive by virtue of its weaker hold on the blue-collar vote.
Esping-Andersen's analysis emphasizes the role of government policy in the "virtuous cycle" of Social Democratic dominance in Sweden. Essentially, Esping-Andersen argues that the reform policies implemented by the Swedish Social Democrats in the postwar era enabled them to avoid the electoral "decomposition" experienced by the Danish Social Democrats as well as the British Labour party in the 1970s.6 Depending on how it is organized, the welfare state may accentuate or attenuate conflicts of interest within the electoral base of Social Democracy. The potential for conflict between taxpayers and welfare recipients is reduced to the extent that the welfare state benefits al! citizens by providing a high level of coverage on a universal basis. Furthermore, the welfare state promotes working-class solidarity to the extent that it decommodifies labor power (i.e., to the extent that it provides services rather than cash benefits and/or severs the link between consumption and income from employment). The contrast between the Swedish and British welfare states on these scores would seem to be even sharper than the Swedish-Danish contrast presented by Esping-Andersen.7
In a similar vein, one might argue that the policies of Social Democratic governments have enhanced the organizational resources of the Swedish labor movement. "While Social Democratic governments have directly and indirectly promoted unionization, the unions affiliated with LO have played a crucial role in mobilizing the blue-collar vote for the Social Democrats, and there is a definite correlation between union membership and socialist voting among white-collar as well as blue-collar strata. More subtly, the
6.  Casta Esping-Andersen, Politics against Markets: The Social Democratic Road to Power (Princeton: Princeton University Press, 1985), and Chapter 1 in this volume.
7.  Cf. Andrew Martin, "Is Democratic Control of Capitalist Economies Possible?" in Leon I.i.-idberg et al., eds-, Stress and Contradiction in Modern Capitalism (Lexington, Mass.: D. C. Heath, 1975); Norman Furniss and Timothy Tilton, The Case for the Welfare State (Blooming-ton: Indiana University Press, 1979); and Mary Ruggie, The State and Working Women (Princeton; Princeton University Press, 1984).
62   /   Jonas Pontusson
system of press subsidies introduced by the Social Democrats in 1969 has largely (though not exclusively) served to support daily newspapers owned by the labor movement.8
The political dominance of Social Democracy in Sweden might thus be conceived as a virtuous cycle involving three essential features: capacity to mobilize voters, control of the political agenda, and control of the government. Each feature can be said to explain the others, and beyond that, it is difficult to say anything definitive about the causal relations among them. Accepting the notion of one-party dominance as a phenomenon that tends to reproduce itself, the following discussion explores external conditions that might explain how this virtuous cycle is established and sustained over time. Esping-Andersen's analysis of the feedback effects of government policy invites the obvious question, Why have some labor parties been able to wield government power in such a way as to reproduce their political dominance and others have not? Esping-Andersen argues that the economic and political strength of the petit bourgeoisie has constrained the exercise of political power by labor in Denmark, but social structure clearly does not explain the success of Swedish Social Democracy relative to British Labour.
Economic circumstances would seem to provide a more compelling explanation. Arguably, rapid postwar growth enabled the Swedish Soda! Democrats to implement reforms that would pay off politically and to avoid the problems associated with incomes policy in Britain. Sustained economic growth was undoubtedly a necessary condition for the political success of Swedish Social Democracy, but the postwar growth rate of the Swedish economy was by no means exceptional, and the difference between Sweden and Britain on this score was not so great. The Swedish economy grew at an average annual rate of 3.0 percent and the British at 2.4 percent in the 1951—73 period, as compared with 4.4 percent for France, 4.8 percent for Germany, and 5.5 percent for Italy.9 An economistic account of the contrast between our two cases would presumably focus on the competitiveness of export industries and the peculiar vulnerability of the pound sterling by virtue of its traditional role as a reserve currency (rather than aggregate growth rates). Without denying the importance of such factors at any given conjuncture, the following discussion emphasizes political variables. As Stephen Blank argues, Britain's industrial decline should not be seen as a predetermined, inexorable process; the politics of economic policy has mattered throughout.10 We should not presume that more vigorous efforts to
8.  See Pontusson, Swedish Soda! Democracy, pp. 26—27, for a more extensive discussion. With union members accounting for roughly 85 percent of the work force, Sweden is by far the most thoroughly unionized of die capitalist countries.
9.  Figures from Peter Hall, Governing the Economy (New York: Oxford University Press, 1986), p. 26.
10. Stephen Blank, "Britain," in Peter Katzenstein, ed., BeUveen Power and Plenty (Madison: University of Wisconsin Press, 1978).
Conditions of Labor-Party Dominance   I    63
reform British capitalism by postwar Labour governments would necessarily have failed.
I shall first address the significance of different electoral systems as an explanatory variable and then consider the nature of the opposition faced by labor parties in the two countries. The interests and organization of capital as well as the parties of the right will be discussed. In the course of mis discussion, I shall try to build a case for looking at the politics of labor. The Swedish labor movement has enjoyed more favorable circumstances than the British in several respects, but it has also been more able to take advantage of favorable circumstances. In other words, I shall argue that strategic choices by labor matter. When such an argument is made, strategic choices are commonly treated as discrete and more or less fortuitous. By contrast, I shall argue that the Swedish labor movement possesses a certain strategic capacity that the British labor movement has always lacked and that this capacity is embedded in the organizational structures of the labor movement.
The oft-noted contrast between the fragmented structure of British unions and the centralized structure of Swedish unions is of secondary importance in this context. My discussion of the politics of labor will instead emphasize the integration of union and political activities in Sweden and the essentially political orientation of the Swedish union movement, that is, its willingness to subordinate short-term economic gains to long-term political goals. These features, rooted in the formative experiences of the Swedish labor movement, stand in sharp contrast to the instrumental relationship between the Labour party and the unions in Britain and the "economism" of British unions.
Comparing the Swedish experience to. that of the other Scandinavian countries, Esping-Andersen argues that "the key to Swedish Social Democracy's ascent to power . . . lies not in the labor movement but in the class structural mix." My comparison of Swedish Social Democracy and British Labour yields the opposite conclusion. The choice of comparative reference points clearly matters a great deal to our understanding of the Swedish case. But Esping-Andersen's and my own analyses are by no means mutually exclusive; quite the contrary, they could easily be integrated within a broader comparative discussion of conditions of labor-party dominance.
Electoral Systems
From 1944 to 1970, the Swedish Social Democrats averaged 46.7 percent of the popular vote in nine parliamentary elections, and the British Labour party averaged 46.0 percent in eight elections. The Swedish Social Democrats controlled the government for the duration of this period, but the
64    /   Jonas Pontusson
British Labour party held office less than half the time. It is tempting to conclude that the contrast between these two cases hinges on the effects of different electoral systems, proportional representation in Sweden versus winner-take-all in Britain. The comparison of aggregate voting shares is somewhat misleading, however, for it ignores the Communist party's contribution to the electoral strength of the Swedish labor movement as well as Britain's more favorable social structure. The only reason the British Labour party did relatively well in aggregate terms was that manual workers constituted a greater proportion of the British electorate. As indicated in Table 2.1, the British party's ability to mobilize the blue-collar vote was distinctly inferior.
Nonetheless, the argument that proportional representation explains the ability of the Swedish Social Democrats to stay in power and to reproduce their political dominance deserves attention. One version of the argument would hold that control of the government matters more than the particular policies a party pursues in government. In a slightly different vein, Sven Steinmo argues that the Swedish Social Democrats have been able to approach policy making in a different manner from the British Labour party because they have been virtually assured of at least a share of government power. Above all, they have been more able to pursue long-term policy objectives.11
On the assumption that the distribution of votes between socialist and nonsocialist parties would remain the same, an electoral system modeled on the British would have enhanced the political power of the Swedish Social Democrats, but Steinmo challenges this assumption. The electoral strength of the Swedish left is itself a product of proportional representation, he argues. Whereas the British system of winner-take-ali elections forces left critics of Labourism to remain within the Labour party, proportional representation has enabled the left wing of the Swedish labor movement to survive as a separate political force. According to Steinmo, the absence of a strong left wing within the party has enabled the Swedish Social Democrats to appeal more effectively to swing voters in the center of the political spectrum.
I am persuaded by Steinmo's argument that electoral systems have a profound impact on patterns of policy making, but I am not persuaded by his argument that electoral systems explain the political dominance of Swedish Social Democracy and the failure of British Labour to become equally dominant. For one thing, Steinmo speculates about the effects of electoral systems on the size of the left electorate without any reference to their effects on the size of the right electorate. One could just as well argue that a winner-take-ail system would undermine the ability of Sweden's
11. Sven Steinmo, "Taxes, Institutions and the Mobilization of Bias" {Ph.D. dissertation, University of California, Berkeley, 1986).
Conditions of Labor-Party Dominance   I   65
bourgeois parties to appeal to swing voters in the center of the political Spectrum. The fragmentation made possible by proportional representation has enabled these parties to appeal to a wide range of voter preferences. In the 1970s, the Conservatives appealed to antitax, antiwelfare sentiments, the Center party appealed to environmentalist concerns, and the Liberals sought to distinguish themselves by advocating increased aid to developing countries and equal rights for women and immigrants.
Steinmo's argument about the connection between electoral arrangements and the electoral appeal of the left raises two distinct issues: factionalism and radicalism. Internal party unity is indeed an outstanding feature of Swedish Social Democracy from a comparative perspective and clearly distinguishes Swedish Social Democracy from British Labour. But is it true, as Steinmo implies, that British Labour has been more leftist than Swedish Social Democracy? It is clear that a significant proportion of British voters came to perceive the Labour party as extremist in the early 1980s, but polling data do not provide an answer to the question here, for voters' perceptions and preferences may well vary independently of party policy. Under some circumstances, voters may perceive nationalization measures, for instance, as extremist; under others, they may not.
The idea of British Labour as more leftist than Swedish Social Democracy has some validity if we compare party rhetoric in the early 1980s, but it becomes untenable if we extend the time frame of the comparison and consider not only what these parties have said they would do but also what they have done when in power. One would be hard put to argue that the electoral promises made by Harold Wilson in 1964,1966,1970, and 1974 were, on balance, more radical than those of the Swedish Social Democrats in the corresponding period (and the same is true for the electoral promises of James Callaghan in 1979). In power, the Labour party has each time abandoned its reformist ambitions in the face of balance-of-payments difficulties, demonstrating a marked penchant for "calling upon its keenest supporters for the greatest sacrifices."12 Meanwhile, the Swedish Social Democrats have conceived and implemented a series of far-reaching reforms, pertaining to industrial democracy as well as public welfare provision and education. And the so-called Meidner Plan adopted by LO in 1976 would have entailed a more far-reaching collectivization of ownership than any proposal put forth by the left wing of the British Labour party in the 1970s.13
The absence of factionalism within the Swedish Social Democratic party might be linked to Social Democratic control of the government as well as
12.  James Ctonin, Labour and Society in Britain, 1918-1979 (London: Batsford, 1984), p.
13.  The Meidner Plan became a source of intense ideological controversy in 1978—82 and was never implemented. The debate over wage earner funds illustrates the limits of Social Democratic control of the political agenda in postwar Sweden. See Jonas Pontusson, "Radi-calization and Retreat in Swedish Social Democracy," New Left Review no. 165 (1987): 5—33.
66   I   Jonas Pontusson
proportional representation. At the leadership level, the party has largely become a party of careerists, and the promise of power has served to imc-grate potential dissidents. By the same token, I would argue that factional strife within the British Labour party should be seen as a symptom of the crisis of Labourism rather than its cause. It is surely not an accident that the most intense factional struggles have occurred when the Labour party has been out of office.
From a broader comparative perspective, proportional representation clearly does not provide a sufficient explanation for the political success of Social Democracy in Sweden. After all, several West European countries have proportional representation, and in only a few has Social Democracy come to assume a dominant position. Wedding proportional representation to rapid economic growth does not seem to yield much greater explanatory power.
Bourgeois Parties
With Britain as his principal reference point, Francis Castles argues that the dominance of Social Democracy in Sweden (and the other Scandinav inr. countries) is a product of the fragmentation of the right.14 At the same time as the Swedish Social Democrats have been able to mobilize working-class support by invoking the ideology of class struggle, they have been able to play the bourgeois parties against each other and to strike compromises with either or both parties of the center, the Agrarian/Center party and the Liberals.
The differences between the party-political constellations opposed to Libor in our two cases are clearly related to the differences between their electoral systems. Whereas the electoral ascendancy of the Labour party served to marginalize the Liberals in Britain, proportional representation enabled several bourgeois parties to survive in Sweden. Indeed, one might argue that proportional representation encouraged these parties to maximize electoral support in the short run, at the expense of a long-term hegemonic project. (This argument is quite different from Steinmo's.)
Though reproduced by proportional representation, the fragmentation of the right must, as Castles argues, ultimately be explained in terms of prcin-dustrial class relations.15 Swedish feudalism was characterized by a relatively weak nobility and an independent peasantry. The commercialization
14,  Francis Castles, The Social Democratic Image of Society (London: Routledge &C Kegan Paul, 1978). Esping-Andersen develops essentially the same argument in Politics against Markets and in his contribution to this volume.
15.  Cf. also Timothy Tilton, "The Social Origins of Liberal Democracy," American Pohucsl Science Review 68 (1974): 561-71. Tilton and Castles both approach the Swedish case from the comparative perspective of Barrington Moore, Jr., Social Origins of Dictatorship aid Democracy (Boston: Beacon Press, 1966).
Conditioľis of Labor-Party Dominance   í    67
of agriculture diverged from the modal patterns of England and Prussia. While the nobility gradually became transformed into a royal urban bureaucracy, the enclosure movement in the first half of the nineteenth century consolidated a distinct class of farmers with small-to-medium-sized holdings. The politics of the bicameral parliament established in 1866 initially revolved around conflicts of interest between farmers, whose representatives (Agrarians) dominated the Second Chamber, and the urbanized nobility, whose representatives (Conservatives) dominated the First Chamber. By contrast, the commercialization of agriculture in England, which occurred much earlier, transformed the nobility into a capitalist landowning class, and the reconstituted Conservative party of the late nineteenth century spanned the urban-rural divide.
The political unity of Swedish farmers disintegrated over the issue of democratic reform, but the tradition of independent farmer politics remained. The Agrarian party was formed in the wake of World War I. Initially leaning toward the right, the party was distinguished by its concern with the immediate interests of a well-defined constituency and its lack of a itror.g ideological profile. This pragmatism made possible the "cow trade" ■jí 1933, whereby the Agrarian party supported the Social Democratic government's public works program in return for higher agricultural tariffs, md subsequent Red-Green coalition governments (1936—40 and 1951-57).
'I he pension reform of 1959 might be characterized as a second breakthrough for Swedish Social Democracy. As Esping-Andersen has emphasized, the reform enabled the Social Democrats to mobilize new support imong white-collar strata and thus weather the defection of the Agrarian jany.16 Divisions among the bourgeois parties again played a critical role in this realignment. The Agrarian party objected to the pension reform proposal put forth by the Social Democrats (known as ATP) but refused to back :he alternative advocated by organized business and the opposition parties. Instead, it presented its own proposal for pension reform in the advisory referendum of 1957. Although a majority of voters supported the other two proposals, the ATP proposal received a large plurality (45.8 percent).
The Agrarian party left the government shortly after the referendum, and the extraordinary elections of 1958 resulted in a tie between the proponents and opponents of ATP in the Second Chamber (i.e., between the socialist and the bourgeois parties). Having lost badly in the elections, the Liberals began to reevaluate their opposition to the ATP proposal. The Social Democrats and the Liberals failed to reach a compromise, but a Liberal member of parliament abstained when the ATP proposal was finally put to a parlia-
16. C£. also John Stephens, The Transition from Capitalism to Socialism (London! Mac-miihr., 1979). See Hugh Heclo, Modern Social Politics in Britain and Sweden (New Haven: Yale University Press, 1974), for the most detailed treatment of the politics of the ATP reform available in English.
68   /   Jonas Pontusson
mentary vote in 1959. The ATP system, commonly considered the centerpiece of the modern Swedish welfare state, was thus introduced by a one-vote majority. The Liberal party endorsed the reform once passed, and by the late 1970s the ATP system had become a sacred cow with which no party dared to tamper.
The fragmentation-of-the-right thesis is quite apposite, but several qualifications are in order. To begin with, the critical importance of the coalition with the Agrarian party for the breakthrough of Swedish Social Democrats in the 1930s does not reveal why the British Labour government of 1929 failed to adopt a Keynesian response to the Depression. The Red-Green coalition in Sweden was necessary because of the continued importance of small farmers in Sweden, and it entailed significant concessions by the labor movement. Britain's greater proletarianization should, in principle, have made it easier for labor to break with the orthodoxy of deflation. Moreover, the Labour government did not confront a united bourgeois front in 1929-31. Quite the contrary, the idea of expansionary spending was the centerpiece of the Liberal party's platform in the election of 1929, and Ramsay MacDonald could have chosen to ally with the Lloyd George Liberals rather than the Tories.17
With respect to the Swedish case, we should not lose sight of the limits within which the Social Democrats have been able to play the bourgeois parties off against one another. Whenever the Social Democrats have broached issues pertaining to the fundamentals of capitalism, as in the debate over planning in the immediate postwar period and the recent debate over wage earner funds, the bourgeois parties have joined forces in opposition.
Also, the Agrarian party and the other bourgeois parties together held a majority in the Second Chamber of parliament on only one occasion between 1937 and 1976, briefly in 1957-58, and even then the socialist parties held a majority in joint voting by the two chambers. Plausibly, the Social Democrats could have stayed in power without any compromises with the bourgeois parties. To be sure, greater reliance on Communist support might have entailed electoral risks, but the point here is simply that the Social Democrats have been able to exploit divisions among the bourgeois parties because they have bargained with them from a position of strength.
The fragmentation of the right can be seen as a consequence as well as a cause of Social Democratic dominance. As suggested earlier, the cross-class appeal of the Social Democrats has placed the bourgeois parties on the
17. See Peter Gourevkch, Politics in Hard Times (Ithaca: Cornell University Press, 1986), chap. 4; and Margaret Weir and Theda Skocpol, "State Structures and the Possibilities for 'Keynesian' Responses to the Great Depression," in Peter Evans, Dietrich Rueschemeyer, and Theda Skocpol, eds.5 Bringing the State Back In (Cambridge: Cambridge University Press,
Conditions of Labor-Party Dominance   I   69
defensive: these parties have had to respond to Social Democratic initiatives to defend their core constituencies. This was most obviously the case with the Liberals in the pension reform debate.
The Organization and Interests of Capital
It would be wrong to conclude from the preceding discussion that the opposition to labor has been weaker in Sweden than in Britain, for party politics constitute but one dimension of the balance of class forces, and there is, in both cases, a curious asymmetry between party politics and industrial relations. In the arena of party politics, the Swedish case is characterized by Social Democratic dominance and fragmentation of the right, but in the arena of industrial relations, it is characterized by exceptionally strong employer as well as labor organizations. British industrial relations are, by contrast, characterized by organizational fragmentation on both sides. It is instructive to consider how the organization and interests of capital have affected the fortunes of Swedish Social Democracy and British Labour.
It is not necessary, for our purposes, to describe the differences between Swedish and British business organizations in any detail. Suffice it to note that the collective bargaining and lobbying activities of Swedish business have been coordinated by two comprehensive and centralized peak organizations, the Swedish Employers' Association (SAP) and the Federation of Industry (the latter being a trade association). The Employers' Association was created in 1902 and from the beginning sought to bargain direcdy with unions. Indeed, the institutionalization of industrywide bargaining in the early 1900s and the economywide bargaining in the 1950s was largely imposed upon the unions by the employers. The Confederation of British Industry (CBI), established in 1965, is far less representative and far less centralized than its Swedish counterparts and has always avoided direct involvement in collective bargaining.
The organizational cohesion of capital as well as labor has facilitated the institutionalization of class compromise, and this in turn appears to have facilitated Social Democratic dominance. Specifically, employer organization has facilitated the exercise of wage restraint by the unions and enabled the Swedish labor movement to avoid the internal tensions associated with incomes policy in Britain. Though essential to the success of labor reformism, wage restraint becomes an intractable problem for unions if employers offer wage increases that exceed what the unions have secured through collective bargaining. To a greater extent than in Britain, employers' organizations in Sweden have shared responsibility for enforcing wage restraint. . At the same time, corporatist relations between business organizations and the government have weakened the position of the opposition parties in
70   /   Jonas Pontusson
Sweden. During the 1950s and 1960s, government and business representatives met regularly to discuss matters of economic policy. Such meetings seldom entailed direct bargaining, but they did legitimize government policies, making it difficult for the bourgeois parties to claim that Social Democratic rule was bad for the country.
Geoffrey Ingham argues persuasively that the pattern of industrialization explains the organizational cohesion of Swedish employers and the lack of such cohesion in Britain.15 Sweden industrialized much later and more rapidly than Britain. Being latecomers and lacking a significant domestic market, Swedish industrialists from the very beginning had to contend wuh world competition. Most firms pursued niche-orientated marketplace strategies based on a narrow range of products. Consequently, they seldom competed with each other, but their dependence on export markets made it imperative to resist wage increases. At the same time as the commonality or interests among employers has been greater in Sweden than in Britain, tlít relatively small number of employers in each industry, because of the country's small size as well as a more concentrated industrial structure, lias facilitated employer coordination.
The Swedish business community is distinguished from its British counterpart not only by a higher level of organization and organizational centralization but also by the hegemony of export-oriented manufacturing capital, as opposed to finance capital. Arguably, the failure of the Labour party to impose its own policy priorities when in government is related to the prominent role of the City of London (i.e., the financial community) in the British political economy.
It is a commonplace that the first Wilson government's commitment to maintain an overvalued pound effectively torpedoed its attempt to break with the pattern of stop-go policy and to institutionalize tripartite collaboration to promote industrial development. As deflationary measures were imposed in response to the sterling crisis of 1965, planning turned out to be little more than a rhetorical exercise, and the government came to rel> on wage restraint as the centerpiece of its strategy to restore international competitiveness.
Clearly, it was the City and its clients that benefited most from the policy of defending the value of the pound. Stephen Blank argues that it is nonetheless false to see the failure to devalue as an expression of the City's political influence. According to Blank, Wilson and most, if not all, members of his cabinet very much believed in the policy of defending the pound. But why was there such a strong consensus around the defense of the pound? hi his
18. Geoffrey Ingham, Strikes and Industrial Conflict (London: Macmillan, 1974). i'or a critique of Ingham, see Peter Jackson and Keith Sisson, "Employers Confederations in S-* eden and the U.K. and the Significance of Industrial Structure," British Journal of Industrial Relations 14 (1976): 306-23. Jackson and Sisson rightly criticize Ingham for ignoring the significance of the political realignment of die 1930s for the decline of strike activity, but explaining variations in strike activity is not my concern here.
Conditions of Labor-Party Dominance   I   71
ffort to demonstrate that direct pressures from the City did not alter the
;overnment's policy in the 1960s, Blank misses the hegemony of the City in
he arena of foreign economic policy, a hegemony that was institutionalized
) the nineteenth century and confirmed by the return to the gold standard
n 1925.19 In view of the close links between the City, the Bank of England,
nd the Treasury, it seems problematic to treat the City as a pressure group
Landing outside the state and then to inquire whether it has influenced state
■olicy.
To characterize the contrast between Sweden and Britain in terms of the
ominance of manufacturing as opposed to finance capital is somewhat
misleading, however, for it is really the merger of industrial and finance
capital that distinguishes the Swedish case from the British. Again, the
differences between the two cases derive from the timing and pattern of
industrialization.
By comparison with later industrializers, Britain industrialized relatively slowly, and early industrial entrepreneurs enjoyed very high rates of profit. Consequently, British industry has traditionally relied on retained profits and equity markets to finance new investment. The City of London never engaged in industrial finance; rather, it financed commercial transactions and became a conduit for the export of British capital. By contrast, Sweden and other late industrializers had to mobilize large amounts of capital quickly, and banks assumed a key role in the finance of early industry.20 The banks became actively involved in the affairs of their corporate clients, and the intimate relationship between finance and industry was cemented by overlapping ownership. Each of the big commercial banks remains the organizational nexus of a distinctive constellation of industrial firms and capitalist families.
These differences in the relationship between industrial and financial capital correspond to basic differences in the structures of the British and Swedish economies. Imported goods account for a very large proportion of gross national income in both cases. But Sweden has paid for its imports with revenues derived from manufacturing exports, whereas Britain has relied on the return on investment abroad and an invisible trade surplus (income from international banking and shipping services) to make up for a chronic deficit in the balance of visible trade. The commonality of interests between labor and capital has been greater in Sweden than in Britain by virtue of the manufacturing basis of the economy as a whole and the export
19.  See Frank Longstreth, "The City, Industry, and the State," in Colin Crouch, ed., State and Economy in Contemporary Capitalism (New York: St. Martin's, 1979); and Geoffrey Ingham, Capitalism Divided (London: Macmillan, 1984).
20.  It is hardly necessary to point out that this argument follows the logic of Alexander Gerschenkron, Economic Backwardness in Historical Perspective (Cambridge, Mass.: Harvard University Press, 1962). The structure and organization of capital is an area in which the existing literature (including the Swedish-language literature) on the Swedish political economy is very weak.
72   /   Jonas Pontusson
dependence of manufacturing industry. From this perspective, centralized union and business organizations may be necessary but do not represent a sufficient condition for the institutionalization of class compromise.
Policy Innovation
The Swedish Social Democrats have enjoyed more favorable economic and political circumstances than the British Labour party, but they also seem to have been more able to take advantage of favorable circumstances. Policy innovation would appear to be an important consideration here. In Andrew Shonfielďs judgment, "Sweden offers the solitary instance in the postwar capitalist world of a trade union movement which has made a significant intellectual contribution to the development of the system." By contrast, the British Labour party displays, in Goran Therbom's words, "a striking lack of initiative, imagination and capacity."21 Perhaps the innovative capacity of the Swedish labor movement accounts for its political dominance. If so, how is this innovative capacity to be explained? Let us briefly examine the two most famous instances of Social Democratic policy innovation: the adoption of an active employment policy in the 1930s and the postwar strategy to reconcile full employment and price stability, known as the Rehn-Meidner model (the latter being the contribution to which Shonfield refers).
Economic historians have shown that the public works program introduced as part of the "cow trade" of 1933 played only a minor role in the economic recovery of the 1930s. Going a step further, Nils Unga challenges the conventional wisdom that the recovery program on which the Social Democrats campaigned in 1932 was informed by the writings of Keynes and that the Social Democrats, once in power, proceeded to apply the principles of Keynesian economics in a conscious and consistent manner.22 Unga demonstrates that the Social Democratic leadership perceived deficit spending as an extraordinary crisis measure, a necessary evil, rather than a means to stimulate economic expansion. Although unemployment among union members did not drop below 10 percent until the onset of the war, the government made no attempt to combat unemployment through deficit spending after 1933. In fact, it strove to repay the crisis loans o£ 1933 ahead of schedule on the grounds that social reforms presupposed the return to a balanced budget.
Unga argues that the Social Democrats' break with the orthodoxy of deflation was a response to the concrete concerns of organized labor rather
21.  Andrew Shonfield, Modern Capitalism (Oxford: Oxford University Press, 1965), p. 2O0; Goran Therborn, "Britain Left Out," in James Curran, ed., The Future of the Left (London: Polity Press, 1984), p. 123.
22.  Nils Unga, Socialdemokratin och arbetslöshetfrägan, 1912-34 (Lund: Arkiv, 1976)-
Conditions of Labor-Party Dominance   I    73
than the result of new theoretical insights. A system of public relief works, paying well below market wages, had been established in the 1920s. At the outset, the unions supported the policy of below-market wages, for it meant that more unemployed could be put to work. The competition for jobs in the open market could thus be reduced, but this presupposed that relief works did not reduce the level of regular employment in the public sector. LO's support for the policy of below-market wages became untenable as the pool of truly extraordinary public works was exhausted and the definition of relief works was extended in the course of the 1920s. In response to pressure from relief workers as well as public sector unions, LO changed its position in 1928 and called for the replacement of the system of relief works with public works of a productive nature, paid at market wages. The Social Democratic party in turn adopted this position.
Unga's analysis suggests that innovative thinking by party leaders is not the key to the different responses to the Depression of the labor parties in Sweden and Britain. Rather, the divergence seems to hinge on the responsiveness of party leaders to union concerns. In proposing to cut unemployment benefits, Ramsay MacDonald and Philip Snowden completely disregarded the views of the Trades Union Congress (TUC). But why were Swedish party leaders more responsive to union concerns? The strategic choices made in 1929—31 must be understood against developments in the preceding decade.
In Sweden, union membership grew steadily in the 1920s. The membership of unions affiliated with LO nearly doubled, increasing from 280,000 in 1920 to 553,000 in 1930, and the rate of unionization among nonagrarian blue-collar workers increased from 31 percent to 52 percent. The Social Democrats made substantial gains in the elections of the early 1920s, but their share of the popular vote dropped from 41.1 percent in 1924 to 37.0 percent in 1928. In Britain, by contrast, electoral support for the Labour party continued to increase through the 1920s, peaking at 37.1 percent in 1929. Yet the membership of unions affiliated with the TUC dropped from 6.4 million in 1920 to 3.7 million in 1930, and the rate of unionization among blue-collar workers dropped from 51 percent to 26 percent. Most of this decline in union membership actually occurred before the disastrous general strike of 1926 (the rate of unionization was 31 percent in 1925).23
Swedish unions entered the Depression in a much stronger position than the British unions, and their leverage vis-a-vis the Social Democratic party leadership was undoubtedly strengthened by the party's electoral difficulties. But to explain divergent strategic choices in terms of the balance of power between the two wings of the labor movement is not entirely satisfactory. For reasons to which I shall return, party-union relations in the two
23. Unionization figures are from Anders Kjellberg, Facklig organisering i tolv lander (Lund: Arkiv, 1983), pp. 269, 299.
74   /   Jonas Pontusson
cases have always differed qualitatively. As Lewis Minkin argues, the Labour party is distinguished from Social Democratic parties by two seemingly contradictory features: it is more purely a trade-union party, yet the party leadership from the very beginning enjoyed a greater degree of autonomy vis-ä-vis the "movement."24 Arguably, the coincidence of electoral growth and union decline in the 1920s, and the outcome of the general strike, enhanced this autonomy.
The second case of policy innovation also brings out the importance of the organizational makeup and internal politics of the Swedish labor movement. The basic elements of what came to be known in the 1960s as the Rehn-Meidner model were first presented in a report to the LO congress of 1951 written by Gösta Rehn and Rudolf Meidner, two academic economists hired by LO when it established a research department a few years earlier.25 The model was conceived as an alternative to the wage freeze of 1948—50. Against the background of this experience, Rehn and Meidner argued that unions could not prevent wage increases that employers were able and willing to pay and that the unity and strength of the labor movement would be undermined if the unions tried to assume this role. Instead, the government had to assume responsibility for price stability.
Rejecting the "pure" Keynesianism of the immediate postwar period, Rehn and Meidner argued that selective measures to sustain full employment should be combined with a restrictive fiscal policy that would keep the lid on corporate profits and thereby enforce employers' resistance to inflationary wage increases. For their part, the unions should pursue a solidaristic wage policy, coordinating their wage bargaining with the objective of eliminating wage differentials caused by variations in corporate profitability and imposing a wage structure based on the principle of equal pay for equal work. This wage structure would be more acceptable to workers and thereby reduce inflationary wage rivalries. It would also promote productivity growth through a selective profits squeeze and thus make it possible to reconcile high wages with price stability.
Although it restrained the wage demands of workers in the most efficient firms or sectors, solidaristic wage bargaining would raise the wage of the least efficient firms or sectors and force them to rationalize production or go out of business. To deal with the unemployment generated by the squeeze on the profits of inefficient firms or sectors, the government should develop an active labor market policy designed to promote labor's adjustment to structural changes in the economy. In short, the Rehn-Meidner model pre-
24.  Lewis Minkin, "The British Labour Party and the Trade Unions," Industrial and Labor Relations Review 28 (1974): 7-37.
25.  Subsequent reports by the LO research department elaborated further on various components of the Rehn-Meidner model. For a comprehensive treatment of the evolution of LÖ's strategy and the relationship between theory and practice, see Andrew Martin, "Trade Unions in Sweden," in Peter Gourevitch et al., eds-, Unions and Economic Crisis (London: Allen & Unwin, 1984).
Conditions of Labor-Party Dominance   I   75
scribed a policy package consisting of a tight fiscal policy, active labor market policy, and solidaristic wage policy.
The Rehn-Meidner model was never completely implemented, but it did provide the labor movement with a coherent approach to wage bargaining and economic policy. Although its significance as the intellectual foundation of labor's postwar strategy can hardly be exaggerated, the Rehn-Meidner model has all too often been treated as a more or less fortuitous, and essentially theoretical, achievement by a couple of individuals. The elaboration and implementation of the Rehn-Meidner model must be understood in the context of internal politics and external circumstances of the Swedish labor movement.
Taken separately, the basic elements of the Rehn-Meidner model were not particularly original. What was innovative about this intellectual construction was the way it combined a number of ideas that had long been discussed and were broadly accepted within the labor movement. Invoking the egalitarian ideals of the labor movement, low-wage unions put forth the idea that LO should collectively pursue a "solidaristic" or socialist wage policy at several LO congresses in the interwar period. At the same time, most union leaders were predisposed in favor of rationalization and structural change by virtue of the economy's exposure to international competition. The idea that real wage increases depended on productivity increases prevailed among union leaders long before the Rehn-Meidner model was elaborated.26 In addition, several public commissions of inquiry in the 1930s and 1940s advocated government intervention to promote labor mobility.
While combining preexisting ideas, the Rehn-Meidner model addressed the immediate concerns of the labor movement, and did so in a way that obviously suited the unions. For the model treated inflation as a result of capital being inefficient rather than wages being too high and shifted the political burden of fighting inflation from the unions to the government.
Significantly, the report presented to the 1951 congress did not immediately convince the LO unions to begin practicing solidaristic wage bargaining. Rather, LO's pursuit of a solidaristic wage policy evolved in piecemeal fashion, in response to external pressures from the employers and internal pressures from the iow-wage unions.27 Though solidaristic wage policy implied some form of economywide coordination of wage bargaining, LO only reluctantly agreed to direct negotiations with the Employers' Associa-
26.  See Axel Hadenius, Facklig organtsationsutveckling (Stockholm: Raben & Sjogren, 3976).
27.  See Peter Swenson, Fair Shares (Ithaca: Cornell University Press, 1988), and Hugh Heclo and Hendrik Madsen, Policy and Politics in Sweden (Philadelphia: Temple University Press, 1986), chap. 3. Arguing that the "Swedish model" represents an ex post facto justification for policies that the labor movement adopted in response to immediate problems, Heclo and Madsen provide an important corrective to conventional interpretations, but they go too far in the opposite direction.
76   I   Jonas Pontusson
tion in the second half of the 1950s. For the employers, such negotiations represented a means to enforce wage restraint. The centralization of wage bargaining in turn strengthened the position of low-wage unions within LO. These unions in effect refused to go along with central agreements unless they contained provisions to reduce wage differentials.
Arguably, LO would have had to pursue a solidaristic wage policy even if the Rehn-Meidner model had never been conceived. But the mode! served to facilitate this outcome by identifying the long-term interests of the labor movement as a whole with the "selfish," short-term interests of the low-wage unions. In the end, the Rehn-Meidner model prevailed as the unifying strategy of the labor movement because the low-wage unions, in alliance with the LO research department, prevailed within LO, and because LO prevailed within the labor movement.28
Again, the innovative capacity of the Swedish labor movement seems to turn on the interaction between its different components rather than the intellectual qualities of labor leaders or policy advisers. This discussion introduces a new perspective on the contrast between Swedish Social Democracy and British Labour, for it suggests that we are not simply comparing two more or less similar labor movements operating in different circumstances. The labor movements themselves are different.
Labor Movements
As indicated above, LO has played a very important role in the success story of Swedish Social Democracy. But it is not adequate to characterize the differences between the Swedish and British labor movements—the dominance of the unions in one case and the dominance of the party in the other. After all, the unions control the vast majority of votes at Labour party conferences and can effectively dictate party policy if they agree among themselves. At least formally, the LO unions have much less leverage over "their" party. What truly distinguishes the Swedish labor movement from the British is a more political, class-oriented unionism and closer integration of party and union activities.29
As many authors have argued, the formation of the Labour party in 1906 did not represent a break with the essentially economistic orientation of British unions.30 For the unions, the decision to organize a third party was,
28.  The Rehn-Meidner model became a source of conflict between LO and the government in the 1950s; see Gösta Rehn, "Finansministrarna, LO-ekonomerna, och arbetsmarknadspolit-iken," in Jan Herin and Lars "Wenn, eds-, Ekonomisk debate och ckonomisk politik (Stockholm: Norstedts, 1977).
29.  Cf. Winton Higgins, "Political Unionism and the Corporatist Thesis," Economic and Industrial Democracy 6 (1985): 349-81.
30.  See Tom Nairn, "The Nature of the Labour Party," in Anderson and Blackburn, eds., Towards Socialism; James Hinton, Labour and Socialism (Brighton: Harvester, 1983); and Leo Panitch, Working Class Politics in Crisis (London: Verso, 1986), chap. 1.
Conditions of Labor-Party Dominance   I   77
in the first instance, a response to a series of court decisions (Taff Vale in particular) that threatened to curtail their immunity from tort actions and hence their ability to engage in strikes. The unions' dissatisfaction with the Uberal party's inaction on this score undoubtedly extended to other issues as well, but the crucial point is that the unions turned to independent political action primarily as a means to secure legislation that would protect or enhance their ability to pursue the interests of their membership through collective bargaining. The unions' conversion to the idea of political action as an alternative to industrial action, that is, as a means to satisfy working-class interests directly, was slow and incomplete.
A fairly clear-cut division of labor between union and party leaders became institutionalized in the Labour party's formative period. So long as iheir activities as unions were not directly affected, the unions allowed the party leadership great autonomy in formulating party policy and setting political priorities. As Perry Anderson puts it, "the organizational nexus tying the trade-union and parliamentary institutions of labourism together did not mean any practical unification of industrial and political struggles. Quite the reverse: they were rigidly separated in the ideology of the movement, each wing defining its arena of activity as off-limits to the other—in the familiar duo of economism and electoralism that came to define Labourism."'31 Relations between the Labour party and its union affiliates fjave changed significandy since the early 1960s, of course, but the separation of union and party spheres remains more pronounced than in any other West European country.
The way collective affiliation has been organized illustrates the different nature of union-party relations in Britain and Sweden. In the British case, union members are affiliated with the Labour party through their national unions, and each union casts the votes of its affiliated members as a single bloc vote at party conferences. National union leaders stand between the party and the bulk of its membership. Indeed, the party headquarters does not even have the names and addresses of collectively affiliated members. In the Swedish case, by contrast, union branches affiliated with party organizations at the local level until the party abolished collective affiliation in 1987. Collectively affiliated members had the same status as individual members, and their votes were cast by the representatives of local party organizations at party congresses. Leadership positions in union and party organizations at the local level frequently overlapped, and this pattern will undoubtedly continue. Whereas the formal-organizational link between party and unions occurs at the level of the national party convention in Britain, the link occurred at the local level in Sweden, and relations between the unions and the party at the national level have been almost entirely informal.32
31.  Perry Anderson, "The Figures of Descent,'1 New Left Review no. 161 (1987): 52. Cf. Minkin, "The British Labour Party and the Trade Unions."
32.  Since collective affiliation was not an instrument of union influence over party policy,
78   /  Jonas Pontusson
The differences in the structure of party-union relations are closely related to differences in the strategic outlook of the unions. Swedish unions have been much more inclined to think in terms of political exchange; that is, they have been willing to restrain the exercise of power in the marketplace to achieve political objectives perceived to benefit the long-term interests of the working class as a whole. Equally important, Swedish unions have been less willing to let government policy become the exclusive purview of party leaders. Herein lies the significance of the Rehn-Meidner model. At the same time as it assigned pivotal importance to wage bargaining as a means to promote income equalization and structural change, the model specified very clearly the kinds of policies that the government had to pursue for the unions to be able to assume this role.
Virtually all the major policy initiatives of the Social Democrats in the postwar period were proposed—or at least very actively supported—by LO. This is true, most notably, of the supplementary pension reform of 1959, the development of an active labor market policy from the late 1950s onward, the "industrial policy offensive" of the late 1960s, the industrial democracy reforms of the early 1970s, and, most recently, the idea of wage earner funds. As Winton Higgins and Nixon Apple stress, it is LO rather than the Social Democratic party that has kept up the momentum of labor reformism.33
The political orientation of Swedish unionism manifests itself in rank-and-file attitudes as well as the statements and actions of union leaders. In a 1971 interview survey of workers in two comparable British and Swedish factories, Richard Scase registered markedly different responses to the question, "What do you think should be the major aim of trade unions?" He found that 42.6 percent of the Swedish respondents suggested improved social justice or socialism, 39.4 percent mentioned economic factors, and 9.8 percent wanted improved working conditions; the corresponding figures for the British respondents were 1.6 percent, S5.9 percent, and 23.6 percent.34
Swedish unions became more politically active in the postwar period, but I argue that this development represents a natural outgrowth of earlier experiences. The dominance of industrial unionism helps explain the políti-
the decision of the 1987 Social Democratic party congress to abolish collective affiliation is unlikely to alter the relationship between the party and die unions. Curiously, a thorough study of the organization and internal politics of the Swedish Social Democratic party has yet ro be undertaken. For a brief, journalistic account, see Martin Linton, The Swedish Road to Socialism (London: Fabian Tract no. 503, 1985).
33.  Winton Higgins and Nixon Apple, "How Limited Is Reformism?" Theory and Society 12 (1983): 603—30. The TUC did, in LO-Hke fashion, develop its own economic policy agenda in the 1970s; see Stephen Bomstem and Peter Gourevitch, "Unions in a Declining Economy," in Gourevitch et al., eds., Unions and Economic Crisis,
34.  Richard Scase, Social Democracy in Capitalist Society (London: Croom Helm, 1977), p.
Conditions of Labor-Party Dominance   í   79
cal orientation of Swedish unions, and industrial unionism is in turn a related consequence of the pattern of industrialization. Being a late indus-trializer, Sweden skipped over the early, craft stages of industrial production, and Swedish industry quickly adopted advanced forms of production technology and organization. Except in a few industrial sectors, organizing skilled workers in defense of craft privileges was a much less viable strategy in Sweden than in Britain. But industrial unionism must also be seen, in part, as a political choice that early union organizers made, a choice informed by socialist ideology. Whereas the unions organized the Labour party in Britain, the Social Democratic party organized the unions in Sweden.
Related to the question of the timing of economic and political development, the struggle for equal and universal suffrage played a critical role in the formation of the Swedish labor movement and helps explain the features that distinguish the Swedish labor movement from the British. As in imperial Germany (and other Continental countries), the struggle for democratic reform served to politicize union activities and thus produced what might be termed class unionism. The demand for universal suffrage figured prominently in early union organizing, and the LO unions staged a three-day general strike in support of universal suffrage in 1902. In response to employers' efforts to cut wages, the unions' attention shifted to the industrial arena in the following years. As collective bargaining became institutionalized, union and party activities became more differentiated, but this differentiation occurred within the framework of a common political project.35 In the face of a cohesive and aggressive employer movement, the weakness of the unions' marketplace power, as revealed most notably in the disastrous general strike of 1909, curtailed the development of economistic thinking among the unions.
While the Social Democrats and the Liberals allied at the parliamentary level, the struggle for democratic reform came to serve as a point of convergence among the "popular movements" {folkrörelser) that had emerged in the late nineteenth century.36 The major popular movements were the labor, temperance, and nonconformist movements. Like the labor movement, the temperance and nonconformist movements mobilized the lower strata of society against hierarchy and privilege. They appealed primarily to farm laborers and poor farmers but also gained support among workers in the mill towns that provided the social setting of early industrialization. Overlapping membership between the labor movement and other popular movements was common.
35,  See Donald Blake, "Swedish Trade Unions and the Social Democratic Party," Scandinavian Economic History Review 8 (1960): 19-44.
36.  See Sven Lundkvist, "Popular Movements and Reforms, 1900-1920," in Steven Koblik, ed., Sweden's Development from Poverty to Affluence, 1750-1970 (Minneapolis: University of Minnesota Press, 1975).
80   /   Jonas Pontusson
The alliance with the Liberals and the broad popular coalition behind the demand for universal suffrage undoubtedly enhanced the legitimacy of the Social Democratic party and of the labor movement more generally. Perhaps more important, this experience in coalition politics shaped the strategic outlook of the labor movement. As Steven Koblik suggests, its interaction with other popular movements enabled the Swedish labor movement to escape the "class ghetto" that isolated the German and Austrian labor movements from the petit bourgeoisie and from society at large.37
The British labor movement was never quite so insular, and it certainly interacted with nonconformist movements. But other popular movements were less significant in Britain, and there did not exist any overarching issue, like universal suffrage, around which such movements could converge with the labor movement. Whereas the Swedish Social Democrats proclaimed themselves a "party of the nation" in 1910 and explicitly sought to appeal to new voting groups in the election of 1911, the British Labour party did not open its ranks to "workers by brain" until 1918. (It actually rejected a proposal to this effect in 1912.)
I do not mean to imply that the Swedish Social Democrats opted to become a national party and the British Labour party opted to remain a class-pure party. Leo Panitch argues persuasively that the political strategy of the Labour party was from the beginning premised on the idea of serving the nation.38 But the Labour party accepted the national interest as essentially given, and the unions affiliated to it remained representatives of sectional interests. In a sense, the key to the hegemony of Social Democracy in Sweden is its ability to avoid this juxtaposition of national and sectional interests. On one hand, the Swedish labor movement has subordinated sectional interest to class interest; on the other, it has wedded class interest to national interest. The way the Rehn-Meidner model combines considerations of equality and efficiency represents the most obvious illustration of the latter point, but the earlier, vaguer, and more all-encompassing idea of the "people's home" (folkbemmet) linked class interest and national interest in a similar fashion.39
If my analysis is correct, the ability of the Swedish labor movement to avoid becoming caught in the juxtaposition of sectional and national interests is not simply a matter of fortuitous circumstances; it is, in large measure, a product of the organization and politics of the labor movement itself and, in particular, a product of what might be termed class unionism.
Furthermore, the coincidence of the struggle for democratic reform with the formation of the labor movement helps explain the electoral strength of
37.  Steven Koblik, in ibid., p. 466.
38.  Leo Panitch, Social Democracy and Industrial Militancy (Cambridge; Cambridge University Press, 1976).
39.  See Seppo Hentilä, "The Origins of Folkhem Ideology in Swedish Social Democracy," Scandinavian Journal of History 3 (1978): 323—45.
Conditions of Labor-Party Dominance   I   81
Swedish Social Democracy. In Britain, most male workers gained the right to vote in 1867, and the Liberals, in particular, had established a strong electoral base within the working class before the formation of the Labour party. In Sweden, by contrast, the Social Democratic party predated the Liberal party by a decade and could appeal to first-time working-class voters on the basis of its vanguard role in the suffrage struggle. Whereas the Social Democrats effectively controlled the working-class vote from the party's inception, the Labour party had to win over working-class electoral support from the other parties. Furthermore, the struggle for democratic reform enabled the Social Democrats to project themselves as a national party, and to appeal to lower-class strata other than industrial workers.
Conclusion
The Swedish labor movement emerged from the struggle for universal suffrage and parliamentary government with a hegemonic disposition, by which I mean that it was inclined to set priorities for long-term political objectives and had learned to consider building alliances with other social and political forces. The British labor movement never acquired a hegemonic disposition in this sense.
By now, at least a few readers will already have formulated the following objection: the early dominance of industrial unionism and the formative role of the struggle for political democracy clearly distinguish the Swedish labor movement from the British, but the formative experiences of the Swedish labor movement are hardly exceptional from a broader comparative perspective. The struggle for political democracy played a pivotal role in the formation of most labor movements on the Continent; yet German, French, or Italian labor parties, to mention only the most obvious examples, have never assumed a dominant position comparable to that of Swedish Social Democracy. In other words, the argument that I have developed is subject to the same objection I raised with reference to the economists explanation of the contrast between the two cases: it does not represent a sufficient explanation of the political dominance of Social Democracy in postwar Sweden.
It thus becomes necessary to distinguish between labor's ability to generate a hegemonic project and its ability to realize such a project. Arguably, the German, French, and Italian labor movements also developed a hegemonic disposition in their formative period, but they failed to become dominant because of unfavorable circumstances. Two basic background conditions set the Swedish case apart from late-industrializing countries on the European continent. First, Sweden industrialized more thoroughly than other late industrialize«: self-employment and small business became more marginalized, and the industrial working class assumed greater numerical
82   /   Jonas Pontusson
importance.40 Second, there were no significant religious, ethnic, or It guistic cleavages in Sweden.
On both these counts, the British labor movement also enjoyed mo favorable circumstances than its Continental counterparts. At this level generalization, one could perhaps characterize the British case as havii favorable circumstances but no hegemonic disposition, the Italian case one of hegemonic disposition but unfavorable circumstances, and the Swe ish case as hegemonic disposition and favorable circumstances. (The Unm States is the most obvious case of no hegemonic disposition and unfavorab circumstances.)
As we have seen, however, the Swedish labor movement has also enjoyi more favorable circumstances than the British labor movement in sei er respects—second-order favorable circumstances, as it were. First, its opp nents in the realm of party politics have been divided, for reasons that ha to do with preindustrial class cleavages and proportional representatio Second, the manufacturing basis and organizational cohesion of Swcdr capital, rooted in the timing and pattern of industrialization, have fadhtau the institutionalization of class compromise. Third, the economic margi-for social reforms have been wider in Sweden than in Britain by virtue of tl greater international competitiveness of the Swedish economy.
To avoid the charge of hístorícísm, let me emphasize that I do not bchcr that the origins of a phenomenon constitute an explanation of its coniinin existence. I hope that it is clear from my analysis in the last section thai tl hegemonic disposition of the Swedish labor movement has persisted becau it has been embedded in the organizational structures and political practic of the labor movement. By virtue of their comprehensive and centrr.li/.c character, Swedish unions have been more able to pursue a coherent set of wage-bargaining objectives and have effectively been forced to consider political exchange.
Similarly, the feedback effects of government policy identified by Esping-Andersen are crucial to understanding how Social Democratic hegemony has been reproduced over time. Although welfare reforms have provided ?. material foundation for the ideology of solidarity, government policies hj\ c enhanced the mobilizational capacity of the labor movement by promoting unionizaton and subsidizing the labor movement press.
40. Cf. Przeworski and Sprague, Paper Stones, p. 35.
The Political Economy of Labor-Party Dominance and Decline in Israel
Michael Shalev
Mapai: A Dominant Party
Simplistically defined, dominant parties are those that head governments ■er long periods (on a scale of the number of decades rather than of bincts), continuously or with only ephemeral interruption. They are not cessarily majority parties, but they always enjoy a plurality of the popular jte. Until 1977, Mapai (the Hebrew acronym for Palestine Workers' party) d its successor the Israel Labor party, which since 1969 has run in electa! ''alignment" with smaller labor parties, eminently fits this description.
In race, Maurice Duverger's notion of one-party dominance was for many
ye.jrs a leading paradigm of Israeli political sociology.1 Mapai was formed in January 1930 by a merger of the two leading labor
1 he rhemes of this chapter and a number of related issues that could not be dealt with here ;re jlaborated in my forthcoming volume Labour and the Political Economy in Israel (Oxford: Oxio-d University Press). Readers curious about my intellectual debts will find them described :n :!iar book. I wish to record here my gratitude to the editor of the present collection for his r'rc'.,s. patience and assistance under trying circumstances and for arranging my participation in two stimulating and convivial conferences. I am also very grateful to Lev Grinberg and Gcts.io:i Shafir for contributing several ideas developed in the chapter and to Michael Barnetr, Jotv Bcinin, Adam Seligman, Gershon Shafir, and Sidney Tarrow for commenting on draft versions.
1. See Maurice Duverger, Political Parties: Their Organization and Activity in the Modern Stare 'London: Merhuen, 1964), pp. 307-15,417-18. Several prominent Israeli scholars have sinployjd the notion of dominance as a party system to analyze Israel under Mapai. In particular, ".cc the work of Yonathan Shapiro, The Party System and Democracy in Israel, unpub-lislifd English version of the author's Democracy in Israel (in Hebrew) (Ramat Gan: Massada, 1978,; and Asher (Alan) Arian and Samuel H. Barnes, "The Dominant Party System: A Neglected Model of Democratic Stability,' * journal of Politics 36 (1974): 592-614. The inter-aaliz2r:on of this paradigm by other observers is evident in an article by Yaakov Reuveni, "The Po!:-_icaI Economy of the Likud, 1977-1984: A Diagnostic Examination" (in Hebrew), Rivon L'Kaikala 126 (1985): 237-47.
83