University of Stockholm

There is no country among the western democracies which can point to such stability in the composition of its government as Sweden. Despite the fact that five parties have long been represented in the Swedish riksdag, one party has dominated the government decade after decade. The five parties are the Conservatives, Liberals, Agrarians (since 1957 the Center party), Social Democrats and Communists. The predominant party is the Social Democrats. With the exception of 100 days during the summer of 1936, the party has without interruption been in power since the autumn of 1932 to the present date. Several of these years the party has governed alone; however, during the 1930's for three years (1936-1939) a coalition was formed with the Agrarians; during the war years (1939-1945) a national coalition was formed with the Agrarians, the Conservatives and the Liberals; lastly during the 1950's the Social Democrats and Agrarians again formed a coalition for six years (1951-1957). Now since the autumn of 1957 Sweden has had a government composed solely of Social Democrats.

In view of this stability, it might be expected that the question of the composition of government would not have attracted any considerable attention in Sweden during the postwar period. However, a public debate on the question, with two different dimensions, has actually occurred. First a discussion has been conducted actually aiming at far-reaching changes in the existing political system. Secondly, there has been a discussion concerning various alternative governments within the framework of the existing system.

On the basis of these interrelated debates in Swedish postwar politics, two types of analyses will be made in the following pages. First, I intend to classify various alternative governments in relation to one another. Subsequently I will try to explain the parties' attitudes toward various alternative governments on the basis of goals which the parties are assumed to have for their action. In all discussions on various alternative governments, the attitude of the political parties is decisive. They determine, through their actions, which alternatives will be realized and which will not.*

* This article is a summary of a part of my book, Mellan samlingsregering och tvåpartisystem. Den svenska regeringsfrågan 1945-1960, which was published in the autumn of 1968.

Two Polar Alternatives

The postwar discussion in Sweden, which could be said to deal with major changes in the existing system, has included two alternatives. The first can be designated as a system of permanent all-party government in a multiparty structure, the other as a system of majority rule in a permanent two-party structure. A convenient abbreviation for the former system is PA, for the latter P2. My thesis is that these two systems are opposite poles and that Sweden's present political system — as well as Norway's - combines features of both PA and P2.¹ In other words, Sweden is located between PA and P2.2

Classification of political systems is currently a field attracting much attention in political science. Previous attempts of this type were usually based on the position and internal relations of the major institutions of government. At present there is an effort to utilize several other criteria: the number of parties, the degree of consensus in society, the level of economic development, the degree of willingness to compromise in decision-making, etc.³ That PA and P2 are regarded as two opposite systems is based on a line of reasoning derived from two such criteria: the number of parties and the degree of proportional party representation at various stages of the decision-making process.

These two criteria for classifying political systems could, of course, be differentiated and specified to a varying extent. In the case of party structure it is sufficient, however, for my purposes to make a distinction between two parties and more than two parties.4 In the case of proportional party representation it is likewise sufficient to make a distinction between two extremes: a situation when all parties - two or more - are represented in exact proportion to their electoral strength at all stages of the decision-making process and a situation when one of these parties is singly represented or strongly overrepresented at all these stages.

A connection is assumed to exist - if other factors such as the degree of consensus, the level of economic development etc. remain constant - between party structure and the degree of proportional representation at various levels. When there are only two parties - under the condition that they are not of exactly equal strength - there are possibilities to adopt majority decisions without cooperation between the parties. On the other hand, when there are more than two parties - under the condition that none of the three or more parties alone has a majority - cooperation between two or more parties must occur in order to attain a majority decision. Cooperation, in turn, may be assumed to be made more difficult, or conversely facilitated, by the degree of proportional representation at the various stages of the decision-making process.

In this way, one arrives at two ideal types of political systems: on the one hand a multiparty structure where all parties are proportionally represented at all stages of the decision-making process, including the government, on the other hand a twoparty structure where the party which has won a majority is singly represented in the government and overrepresented or singly represented at the remaining stages of the decision-making process.

These two ideal types are closely related to the two alternatives in the Swedish

debate designated PA and P2 respectively. In terms of actually existing political systems, these two ideal types most closely resemble the Swiss and the British systems respectively, even if they have on various occasions deviated quite substantially from the archtypes sketched here. In the Swiss system, for example, the political parties have not always been - but are now — proportionally represented in the cabinet, the Federal Council.s In the British system, for example, there has not always in reality been — but are now — only two parties of which only one forms the government.

The Swedish political system thus combines aspects of PA and P2, of the Swiss and the British systems respectively. The PA aspect is found in the simultaneous existence of a multiparty structure and of a high degree of proportional party representation at all stages of the policy-making process with the exception of the cabinet. The political parties are not only proportionally represented in both chambers of the riksdag but also in the various committees of the riksdag. The political parties are additionally well represented, although not in strict proportionality to their electoral strength, in the commissions appointed by the government to investigate special questions; an article in this volume of SPS is devoted to the Swedish system of investigatory commissions. The political parties - excluding the Communists - are also informed about or involved in limited areas of government decisions through intermittent deliberations of the party leaders. They all are additionally represented in the administration through the appointment of individual party members to key administrative posts. Finally, at the local government level all parties are represented not only in various representative councils but also in executive bodies. On the other hand, the P2 aspect of the Swedish political system is found in the existence of one party having received nearly 50 per cent of the electoral vote during several decades and in the existence of majority rule at the cabinet level.

In the Swedish debate, proponents of a modification of the Swedish political system in the direction of PA have conceived of attaining this system through a guarantee from the political parties to attempt always to govern jointly in the future. This permanent all-party government would possibly also be complemented by a compulsory referendum based on the Swiss model. Spokesmen for P2, on the other hand, have concentrated on attaining a consolidation of the non-socialist parties - the Conservatives, Liberals, and the Center party - into one large party on a par with the Social Democrats. The existence of two major parties, in turn, has usually been viewed as paving the way for a more clear-cut majority rule than that which Sweden has had so far. The Communist party, due to its smallness, has been largely bypassed by advocates of PA and P2.

The debates concerning PA and P2 have, however, been conducted rather sporadically. The debate on PA was primarily concentrated to the 1940's and the 1950'5. The driving force behind the debate was the country's largest morning paper, the Liberal Dagens Nyheter under the editorship of Herbert Tingsten; immediately after the second world war Tingsten had left his professorship in political science at the University of Stockholm to assume the post of chief editor of Dagens Nyheter. He and those who shared his opinions have developed a variety of - often contradictor)' - arguments for their standpoint. The debate on P2 has occurred during the entire

postwar period but with particular intensity during the 1960'5. Unlike the debate on PA, it has not been dominated by a single highly articulate person; nor has it been conducted on precisely the same level of abstraction. At the same time this debate, contrary to the debate on PA, has been followed by some limited attempts to convert the discussed ideas into practical action. Individuals from the three non-socialist parties have discussed a merger of the parties. In a constituency in southern Sweden - comprising the cities of Hälsingborg, Landskrona, Lund and Malmö - a list of candidates including representatives of all three non-socialist parties was placed on the ballot during the general election of 1964.

Various Alternatives within the Swedish System

The Swedish postwar discussion concerning various alternative governments within the framework of the existing system has had the two aspects typical of such discussions in parliamentary systems with more than two parties: 1) what proportion of the parliament is a government to be based upon and 2) which party or parties are to form a government.

With regard to the parliamentary basis of a government, three main types can be
discerned: a government based on

1. the entire parliament, i.e. 100 per cent or nearly 100 per cent
2. not the entire parliament, but a majority or half of the parliament, i.e. more

than 50 per cent or 50 per cent

3. a minority of the parliament, i.e. less than 50 per cent.

A government of type 1 is closely related to PA: in both cases the government is in principle based on the entire parliament. The definition of PA also stipulated that the all-party representation at the cabinet level was to be permanent. Here it is only a matter of a government which is temporarily based on the entire parliament. It can then be appropriately called a temporary all-party government in a multiparty structure (TA).

A government of type 2, in turn, is related to some degree to P2: in both cases the government is based on a majority of the parliament. The definition of P2 also stipulated the existence of a two-party structure. Here, by contrast, it is a matter of a multiparty structure. Within the framework of a multiparty structure, there are at least three conceivable variants of governments based on a majority of the parliament. The first is a government composed of only one party. The second is a government comprised of a unified bloc of parties. The third is a government which consists of neither one party nor a unified bloc of parties but of a coalition of a more loosely knit nature. The first of these three variants comes closest to P2. It can then be called a system of majority rule in a temporary two-party structure (T2). On the other hand, a majority government composed of several parties - whether they form a unified bloc or not — can be labelled simply a coalition government (C).

A government of type 3 - one that is based on less than 50 per cent of the parliamerit

- has not any characteristics in common with either a government in a system
of PA or with P2. Such a government - irrespective of whether it is composed of one
or more parties — can be designated a minority government (Mi).

These four types of government - T A, T2, C and Mi - thus fall within the framework of a system which in turn is conceived as being located between the poles PA and P2. It remains then to rank these four types according to their relative position to these two poles. PA and P2 were conceived as polar in so far as the conditions for cooperation between the parties were assumed to be maximal in the one system, and minimal in the other. A line connecting the two, thus, ought to indicate - depending upon which end is regarded as the starting point - a diminishing, or increasing, degree of cooperation between the parties. As mentioned, cooperation in turn is presumed to be positively correlated with the degree of proportional party representation.

It is evident from the preceding remarks that TA is most closely related to PA, and T2 to P2. More diffucult, however, is the ranking of the two intermediate types - C and Mi. It may be argued that C ought to be placed immediately after TA since a coalition government, as defined here, at any rate means that a majority of the parties represented in the parliament — although not all as in PA and TA — are also represented at the cabinet level and there they have opportunities to cooperate. On the other hand, it may be argued that G ought to be placed immediately before T2 since this type of coalition government is related to the system with a government in the majority and an opposition in the minority as presupposed in T2 and P2.

On the scale below, Mi is placed immediately after TA, and G immediately before T2. The explanation is the following: In a situation when a government is based on a minority of the parliament, the center of gravity in the decision-making process is assumed to shift from the cabinet to the legislature. In a situation, however, when a government is based on a majority coalition in the parliament, as in the case of C, such a shift does not occur. The characteristic feature is rather the emergence of a stable cleavage between the parties of the majority coalition represented in the government and the minority not represented. In other words, when the center of gravity shifts to the parliament as a result of minority governments, there are greater opportunities for cooperation among all parties than if a majority government of a coalition character exists. Thus - when opportunities for cooperation is the basis of classification — Mi is nearer PA and TA, whereas G is nearer T2 and P2.

The number of theoretically possible alternatives concerning the party composition of a government is obviously determined by the number of existing parties. (Governments not composed of representatives of political parties but of high officials and experts are excluded.) In a five-party structure, such as the Swedish one, the number of theoretically possible alternative governments is 31. If calculations are limited to the four largest parties, the total of possible alternatives is 15. As mentioned,

the four largest parties are the Conservatives (moderata samlingspartiet), the Liberals (folkpartiet), the Center Party (centerpartiet), and the Social Democrats (socialdemokraterna). The usual Swedish abbreviations are ms, fp, cp, and s. The 15 alternatives, listed according to the number of parties included, are:

No necessary, logical connection exists between any of these fifteen alternatives and TA, Mi, C, and T2. Since the four parties' actual number of representatives in the parliament fluctuates, changes occur in the character of the; various combinations of all-party, majority or minority government. The closest approximation of a necessary connection is between alternative 1 (ms + fp + eps) +s) and an all_-party government, inasmuch as it has virtually never been proposed that an all-party government, transplanted in Swedish conditions, would be composed of all five parties but only of the four largest parties.

As pointed out, Sweden displays no major variations during the postwar period with regard to the proportion of the riksdag upon which the government de facto has been based, nor with regard to the number of parties composing the government. The wartime four-party coalition was a typical TA government; the two-party coalition between the Social Democrats and the Agrarians during the 1950's was a typical C government. The Social Democratic one-party government which existed before and after has mainly been a T2 government. This government for a few years after the dissolution of the C government, however, had virtually the character of a Mi government. In Sweden, it is sometimes difficult to determine whether a government is a majority or minority government. The riksdag is divided into two chambers with equal competence, and divergent opinions exist concerning whether the classification of governments is to be based on the numerical strength in each chamber separately, in only the directly elected lower chamber, or in both chambers together.

The public debate, however, has dealt with considerably more alternative governments than those which have been formed. This debate has flared up primarily in two contexts: first in connection with negotiations on the composition of government between two or more parties, and secondly during election campaigns. On five occasions during the postwar period, direct negotiations have been conducted between parties concerning the government: (1) the summer of 1945 when it was decided to dissolve the wartime TA government; (2) the autumn of 1948 when the Social Democrats and the Agrarians unsuccessfully negotiated to form a C government; (3) the autumn of 1951 when these negotiations were resumed and now met with success; (4) the autumn of 1956 when the two governing parties, which both had suffered an election defeat, negotiated on a short-term continuation of the coalition; (5) and the autumn of 1957 when the only real cabinet crisis in Swedish postwar politics occurred in connection with the dissolution of the C government. In several elections during the postwar years, the future composition of government has appeared

uncertain. In many quarters, it has been felt that there was reason to expect a reversal in the parties' political strength. Prior to the general election to the lower chamber of the riksdag in the autumn of 1968 - against the background of sizable Social Democratic losses in the 1966 local elections - there was an intensive discussion of alternatives to the one-party ministry of the Social Democrats. Inasmuch as the Social Democrats received over 50 per cent of the votes cast in the 1968 election, the situation remained as before.

Party Attitudes towards Three Alternatives within the Swedish System

The attitude of the political parties toward various alternative governments can be elucidated and explained by diverse factors. In this attempt to explain the parties' attitude toward some alternatives within the framework of the existing system, the analysis, as indicated in the introduction, focuses on a number of goals assumed to determine the activity of the parties.6 These goals are assumed to be:

1. to enter the government or to remain in office

2. to implement the party program; program is conceived as the sum of a
party's views and desires

3. to receive the largest possible number of votes

In a two-party system, these three goals for party action coincide. The party which in such a system wins the most votes usually also gains office and maximal opportunities to introduce its program. The situation is quite different in a multiparty system as Sweden. The possibilities of a party entering the government need not always increase with a mounting number of votes. The possibilities of a party to implement points of its program need not always increase with its transition from an opposition party to a governing party. In other words, the goals — holding office, vote maximization, program implementation - can, and often do, come in conflict with one another in a multiparty system.

The parties' attitude toward three alternative governments within the framework of the existing system will be described in this section. The alternatives most frequently discussed during the postwar period were: a TA government of the wartime model, a G government composed of the three non-socialist parties, and a G government consisting of one non-socialist party and the Social Democrats. A new TA government was discussed primarily during the end of the 1940's and during the 1950'5. A C government composed of the three non-socialist parties has been discussed during the entire postwar period but with increased intensity during the 1960'5. A C government consisting of a non-socialist party and the Social Democrats was discussed mainly at the end of the 1940's and during the 1950's when it became a reality through Agrarian and Social Democratic cooperation in the government for six years.

An all-party government. General assumptions can be made about the parties' attitude toward a TA government: The party already in government or viewing itself as having substantial possibilities of entering the government by other means than a TA government is less interested in an all-party government than the parties not in this situation. Similarly, the party which feels that membership in a TA government would additionally pose difficulties in vote maximization and would also jeopardize vital program points is less interested in an all-party government than the parties not in this situation.

The Social Democrats were the party most negative to an all-party government. The party pressed for the dissolution of the wartime government of this type. During the postwar years it has consistently dimissed all proposals for a new all-party government. A TA government, it has been declared, ought to be formed only in an acute international situation. The three non-socialist parties, on the other hand, adopted a somewhat more positive attitude even if this positiveness varied with the party and the situation.

The consistently negative attitude of the Social Democrats; toward a TA government corresponds quite well with the general assumptions above. The party, which by virtue of its size was to remain in government the entire postwar period, would be compelled to share the advantages of holding office with the three non-socialist parties if a four-party government were formed. Furthermore, as a member of such a government, the party would run particular risks in losing votes inasmuch as a foremost rival for the working class vote - the Communists - would not be included in this government. The Communists would remain in opposition with promising possibilities to attract voters dissatisfied with the Social Democrats' cooperation with the non-socialists in the government. Lastly, the Social Democrats, through acceptance of a TA government, could possibly risk obstruction of reforms regarded as essential by the party. When the party urged a dissolution of the wartime TA government, one argument was that a continued coalition with the non-socialists would impair implementation of the party's postwar program. When the party during the cabinet crisis in the autumn of 1957 refused to participate in a new four-party government, it was partially because of the fear that the party's stand-point on the major political issue of the moment - the question of supplementary pensions - would be bargained away.7 In general, a four-party government - requiring compromises between the parties - has appeared as an unsuitable instrument for a party intent upon reforming and changing society.

Among the non-socialist parties, the Agrarians initially during the postwar period exhibited greatest interest in a TA government. This interest may seem surprising inasmuch as this non-socialist party, as events were to show, was in no way limited to a four-party cabinet if it wanted to enter the government. Instead this interest may be explained by the party, as a member of a TA government, not having to risk losing votes and primarily by the fact that such a government was regarded as an advantage in introducing parts of the party's program. The key area of interest to the party - agricultural policy - was considered to be favorably affected by the

pacified political atmosphere resulting from a TA government, at the same time as it was considered easier to gain sympathy for the farmers' interests within the narrow circle constituted by a cabinet as opposed to the large assembly formed by a parliament.

It should be added that the Agrarians' sympathies for aTA government seem to have cooled somewhat toward the close of the 1950'5, although the party still recommended a four-party government in connection with the cabinet crisis in 1957. This decline in interest corresponds with the general assumption made above concerning a TA government, since the change of heart occurred at a time when the party sought the support of new groups of voters by adopting a new name and by putting forward a distinctive proposal during the political controversy over the pension question. The party's major base of electoral support - the farmers and rural population - continued to shrink, and the party recently christened the Center Party attempted to gain a footing among small businessmen. New votes are won more easily if a party is not a member of a TA government.

The Conservatives also displayed a certain interest in a four-party government. Thus in the spring of 1945 the party was sympathetic toward a continuation of the wartime TA government, and the party suggested such a government during the cabinet crisis in 1957. This favorable attitude, however, was coupled with reservations caused by anxiety that the party would be compelled to accept a policy dominated and dictated by the Social Democrats. At the end of the war, this anxiety concerned the postwar program of the Social Democrats; in the autumn of 1957, it concerned the Social Democrats' proposal for a compulsory supplementary pension system. On the other hand, the Conservatives did not have any fear of losing votes as a result of membership in a four-party government, and they were also well aware of the advantages of being represented in the cabinet. There are not many alternative governments to which the Conservatives have access; sympathies within the party in the spring of 1945 were thus combined with a definite anxiety that the other two non-socialist parties could conceivably form a government with the Social Democrats and leave the Conservatives, it was said, to wander in the wilderness together with the Communists.

Finally, the Liberal attitude toward a TA government in peacetime has shifted among various groups within the party and at different points in time. At the end of the war, a majority within the party seem to have been uninterested in a continuation of the four-party government. Contrary to the Conservatives, the Liberals were not worried about becoming politically isolated if the four-party coalition was terminated. Nor did the party, as the Agrarians, view a continuation as being advantageous from the point of view of its program. And the Liberals believed, as the Social Democrats, that it would be easier to win votes if the four-party government was wound up. During the latter part of the 1940's the Liberal party expanded in a spectacular way for Swedish conditions: the party received 12.9 per cent of the votes in the 1944 election, 15.6 per cent in the 1946 election and 22.8 per cent in the 1948 election. The aversion, which subsequently lingered within the Liberal party, toward a new TA government stemmed, as in the case of the Conservatives,

from an apprehension for Social Democratic dominance in such a government. In addition, a feeling existed that there were many other alternative governments open to the Liberals as a party in the center of the political spectrum. But during the cabinet crisis in 1957 the Liberals were the non-socialist party which most vigorously pleaded for a new four-party government, a change in attitude which primarily appears to have been conditioned by a desire to secure a temporary truce to party conflict. The pressing issue of the moment, the pension question, created severe strains on the Liberal party — whose electoral support consisted of businessmen, white-collar workers and laborers. However, since the cabinet crisis resulted in the formation of a Social Democratic minority government, the pension question was not removed from the thick of the battle. Instead the question came to dominate the two election campaigns during 1958 and caused the Liberal party spectacular losses, almost equal to the gains it had won ten years earlier.

A non-socialist "bourgeois" three-party government. Very general assumptions can also be made about the three non-socialist parties' attitude toward a non-socialist three-party government: The non-socialist party regarding itself as having substantial possibilities of entering the government by some other means than by this alternative or a TA government, and which further feels that membership in a non-socialist three-party government would involve the risk of losing votes and finally would impede the introduction of fundamental points of its program is less interested in a non-socialist three-party government than the parties not in this situation.

The Conservatives are undoubtedly the party which during the entire postwar period to the present day has most enthusiastically propounded that the three nonsocialist parties ought to be prepared to form a non-socialist three-party government in the event that the Social Democrats were put in a clear minority. The two other non-socialist parties have adopted a considerably more cautious attitude. In fact, the Agrarians were negative toward non-socialist three-party cooperation during the 1940's and 1950'5. The party's refusal in connection with the cabinet crisis in the autumn of 1957 prevented the formation of a non-socialist three-party government which would have commanded a parliamentary majority in the lower chamber. The Liberals were not quite so negative during the 1940's and 19505; in the autumn of 1957 the party leadership was also prepared to investigate the possibilities of forming a non-socialist three-party government. During the 1960's these two centrist parties, which have established fairly close cooperation, have also adopted roughly the same stand toward possible government cooperation with the Conservatives. The parties declare themselves ready in the event of a Social Democratic election defeat to form a government with the Conservatives, but the parties have not yet gone so far as to sit down before an election and draft a future government program on a three-party basis.

The Conservatives, parallel with their announced readiness to form a three-party bourgeois government, have emphasized rather strongly the mutual affinity of the non-socialist parties. The three parties were said to be united by something more than merely the common label "bourgeois" and the common desire to oust the Social

Democratic party from its dominant position. The other two parties were appreciably more cautious in their declarations about affinities between the three parties. Within the two parties, the very word "bourgeois" had not the same positive overtones as in Conservative quarters. "It reminds me of old, plush sofas and such", declared once Bertil Ohlin, Liberal party leader from 1944 to 1967.⁸

The three non-socialist parties' different attitudes toward a non-socialist threeparty government correspond rather well with the above assumptions. The Conservatives, located at one end of the political spectrum, actually had no other real alternatives than this and a TA government. For the other two parties with their common centrist position, other alternatives still remained. Nor did the Conservatives need to fear any substantial losses in votes through participation in a three-party government. However, the other two non-socialist parties felt they would have to reckon with such losses. Both parties appeal to groups of voters who could conceivably vote Social Democratic if the two parties established too intimate cooperation with the Conservatives. For the Conservatives, a possible drawback with membership in a non-socialist three-party government was of course the risk that they could be forced to make larger programmatic concessions than the other two parties. But at the same time the Conservatives would of course have better opportunities to gain sympathy for their views via a non-socialist three-party government than via a government in which they were not represented at all.

The Social Democratic party, not directly involved in a non-socialist three-party government, has been the highly interested bystander. In their comments, the Social Democrats have vacillated between portraying the three non-socialist parties as unified on all essential points and as fundamentally disunited. Depicting them as united and ready to form a three-party goverment could, as feared within the Liberal party and Center party, strengthen sympathies for the Social Democrats among groups of voters who were reluctant to see the formation of a bourgeois bloc. Depicting them as divided and incapable of forming a joint cabinet could, on the other hand, create a general lack of confidence in these parties and make it easier for the Social Democrats to approach one of them with the aim of collaboration. A leading Social Democratic politician, Per Edvin Sköld, at the end of the war formulated Social Democratic strategy vis-a-vis the non-socialist parties in this manner: "We have to choose the old axiom of divide and rule. And I believe that our position would be better if we could split the bourgeois front".⁹ The Social Democratic leadership has — often with success - applied this strategy during the past decades.

A non-socialist — Social Democratic two-party government. General assumptions can also be made in conjunction with this alternative government. The non-socialist party which to the greatest extent regards itself as being able to reckon with still another alternative government, not yet mentioned, and which further runs the largest risk of losing votes through such cooperation and which finally has the most to lose - or the least to win-in implementing its programs is of course least interested in a coalition with the Social Democrats. The Social Democratic party, for its part, is

least interested in a coalition with the non-socialist party which to the greatest extent can be conceived as limiting the advantages of the Social Democrats' being in office, and which to the greatest degree can be expected to cause indignation among Social Democratic voters and compel modifications of Social Democratic policy.

At the end of the second world war, there was a tradition in Swedish politics of government cooperation between the Liberals and the Social Democrats and between the Agrarians and Social Democrats. The Liberals and Social Democrats had formed a cabinet during the end of the 1910's, the Agrarians and Social Democrats, as mentioned, during the 1930'5. No tradition of government cooperation between the Conservatives and the Social Democrats, however, existed. Nor has this possibility been discussed during the postwar period. Despite the existence of a tradition of cooperation between the other two parties and the Social Democrats, primarily the Agrarian party — and not the Liberals — was mentioned as; a conceivable coalition partner. The Agrarians also showed interest, the Liberals did not. The Social Democrats persistently attacked the Liberal party and reserved mild comments for the Agrarian party. The C government, formed in the autumn of 1951, included the Agrarians, not the Liberals. Several years after the dissolution of this government in 1957, the Social Democrats and Agrarians still seemed to have especially good relations. Only now during the latter part of the 1960's does it no longer appear equally clear that the non-socialist partner in a possible non-socialist - Social Democratic government would be the offshoot of the former Agrarian party, the Center party. But this eventuality of a non-socialist - Social Democratic two party government is now virtually never mentioned.

The Liberals' lack of interest and the Agrarians' interest at the end of the 1940's in forming a coalition with the Social Democrats correspond fairly well with the assumptions made above. At this point, due to its rapid advances, the Liberal party entertained greater ambitions than the Agrarians. The Liberals, in contrast to the Agrarians, clearly indicated that they felt themselves large enough, in the event of a Social Democratic set-back, to bring about a fourth alternative government, a Mi government comprised of the Liberals alone. Both the Liberals and the Agrarians had reason to fear a loss of votes if they entered a coalition with the Social Democrats. The Agrarians, who actually received the bid, expressed the greatest anxiety about losing voters. The Liberals were well aware that the party, if it had entered a coalition with the Social Democrats, would have surely lost votes to the other two non-socialist parties at the same time as it would have no longer been possible for the Liberals, which was their aim, to expand at the expense of the Social Democrats. However, it was judgments concerning the implementation of the party program which constituted the most essential difference between the Liberals and the Agrarians in their attitude toward a coalition with the Social Democrats. The Liberals considered themselves as not having much to gain, but much to lose. The Agrarian party with its concentrated interest in agricultural policy felt, however, that it had much to win. Agriculture was to an especially large degree dependent upon government aid. The decision of the Agrarians at the beginning of the 1950's to form a coalition with the Social Democrats was ultimately determined by anxiety

about the hard-pressed situation of agriculture and by a hope that this situation
could be improved through government cooperation with the party representing the
largest number of consumers.

The attitude of the Social Democrats - a predisposition toward the Agrarians and disinclination toward the Liberals — also corresponds fairly well with the assumptions above. The Agrarian party was the smaller of the two. Consequently it could also be less expected to encroach upon the Social Democrats' dominant position in a coalition cabinet. The Agrarians, by virtue of their homogeneous membership and concentrated interest in a limited sphere of policy, constituted an easier bargaining partner. By accommodating the Agrarians' demands in the field of agriculture, other parts of the Social Democratic government program could be left reasonably intact. Nor did the Agrarians hold the same ambitions as the Liberals to win votes among the core groups of the Social Democrats. A coalition with the Agrarians had an indisputable advantage for being a coalition in a multiparty system: a program could be rather easily agreed upon at the same time as the parties had their main sources of support in two separate sections of the electorate, the workers and the farmers respectively. Similarities in party programs are usually connected with appeals to the same groups of voters and consequently with a contest for the same votes. It may be added that the Center party no longer has the same advantages from the Social Democrats' point of view that its predecessor the Agrarian party was conceived to have as a coalition partner from the 1930's through the 1950'5. The party has grown; at present it is the largest non-socialist opposition party. The electoral support of the party has become more heterogeneous than before; in this way, inter alia, the party has much more than previously concentrated on winning votes among the same groups of voters as the Social Democrats. Parallel with this development, the party's interest in agricultural questions is not as predominant as earlier.

As already intimated, a fourth alternative has also loomed in the background during these somewhat parallel discussions about the three alternative governments. This alternative is a Mi government. In a multiparty structure such as Sweden it is in itself fairly natural that also this alternative would be drawn into the debate; for a few years of the postwar period the actually existing Social Democratic one-party government can also, as noted earlier, be characterized as a Mi government.

The allusions made during the postwar years to governments based on a minority in the riksdag have either been accompanied by aggressive condemnations or rather embarrassed recommendations. The aggressiveness has been shown primarily by the Social Democrats. A government based on a minority in the riksdag, it was held, could not generate adequate strength and effectiveness. The furtive recommendations have been mainly voiced by the Liberals. During the 1920's Swedish politics was characterized by minority governments based on a fairly small portion of the riksdag. The party which held the parliamentary balance during this decade was precisely the Liberals, while the Social Democrats and the other two non-socialist parties felt themselves frustrated. Thus - and quite naturally so - the Liberal party recalls with some nostalgia this decade and its system of government, while the other parties — although with varying degree of rancor — condemn it.

«»

Party Attitudes towards PA and P2

In a number of political systems, it is possible, as e.g. Giovanni Sartori has done in a frequently quoted essay, to distinguish between pro-system parties and anti-system parties.10 Contrary to pro-system parties, anti-system parties are assumed to be intent upon overthrowing the prevailing system. Such a distinction is not relevant to Sweden, nor to any other Scandinavian country. All Swedish parties — including the Communists - accept the existing system. The parties' sense of well-being within the system, however, can vary.

The postwar Swedish debate concerning PA and P2 implies a discussion of alternatives, which, according to the classification employed here, may be said to entail a modification of the existing system. Typically enough, the party spokesmen have as a rule characterized this discussion as rather meaningless and superfluous. Certain differences in attitudes can, however, be observed in the relatively unengaged comments of party spokesmen. Some are completely indifferent, while others show some sympathy. Concerning the variations in these attitudes, a rather elementary assumption can be made: spokesmen for parties which have reason to feel some dissatisfaction within the existing system are more predisposed toward PA and P2 than spokesmen for relatively satisfied parties. Dissatisfaction among political parties is chiefly presumed to arise through a prolonged position in the opposition and dim prospects of entering the government in the future.

The Swedish Conservatives would seem to have most reason to feel dissatisfied within the existing system. Not only have they not been in the government for more than two decades. They also have, as previously pointed out, a position in the political spectrum which markedly restricts the number of government combinations in which the party can conceivably be included. The Conservatives are also consequently the only party which has revealed sympathies for both PA and P2.

The Conservatives would, both with a PA system and a P2 system, have greater chances of being represented in the government than those afforded by the present system. With a PA system the party would be guaranteed permanent government representation; with a P2 system the section of the consolidated non-socialist party roughly corresponding to the Conservatives of today would enter office as soon as the Social Democrats were put in the minority. It ought to be added that PA and P2 further appeal to each of the two somewhat opposed tendencies usually found represented in conservative parties. One is a desire to depoliticize public life, to conduct a policy elevated above various groups and classes, to make room for various categories of experts in decision-making at the expense of the politicians. A PA system can be expected to have effects of this type. The other tendency is a desire to rally and coordinate all forces in society opposed to socialism, to offer a clear alternative to that represented by the labor movement. A P2 system would operate in this direction through a merger of all non-socialist parties.

The other two non-socialist parties, despite the common centrist position they
claim to have established, would seem to have rather varying reasons to sense satisfaction.
The Liberals, with the exception of the wartime TA government, have been

constantly in opposition, while the Agrarians (now the Center party), on the other hand, really have lived up to a classic pivotal position, inasmuch as the party has sometimes been in opposition, sometimes has governed with the Social Democrats. In spite of this, it was the Agrarians who indicated some interest in at least one of the two alternatives outside the system, while the Liberals adopted a chilly attitude toward both of them.

The alternative which both the Liberals and Agrarians repudiated was P2; this attitude was stronger during the 1940's and 1950's than during the 1960's when these parties, as mentioned earlier, declared themselves prepared to enter a nonsocialist three-party government. The negative attitude toward P2 among the Agrarians and Liberals seems primarily to have stemmed from an anxiety that a consolidated non-socialist party could conceivably estrange many voters who had previously voted for the two centrist non-socialist parties. In other words, it was felt that if the three parties were to merge it would be more difficult, not easier, to attain the common objective of the non-socialist parties, to put the Social Democrats in a minority position. In addition, within the two centrist non-socialist parties, a mutual affinity between the three parties was not felt so strongly as within the Conservative party, and the Agrarians and Liberals generally viewed the division of society into two halves through the existence of two major parties as an unhappy arrangement. The guarantee for the formation of homogeneous majority governments provided by a P2 system was not so highly esteemed since the two parties - above all the Liberals - were not in principle opposed to the existence of minority governments.

The alternative which the Agrarians, unlike the Liberals, were sympathetic toward was PA. The Agrarians were sympathetic toward a permanent all-party government for largely the same reasons as they as a rule were positively inclined toward a temporary all-party government. Within a PAs ystem the Agrarians, who have assigned high priority to being represented in the government, would be assured cabinet representation on a permanent basis. The Liberals - who on some occasions have and on other occasions have not evinced sympathy for a TA government — have been reluctant to make any clear recommendations for a PA system. The possibilities of various alternative governments offered by the existing system - if only the Social Democrats were put in the minority — pleased the Liberals. Simultaneously, the Liberals professed that in principle they had misgivings about the absence of an organized opposition in a PA system; in Swedish politics the Liberal party had once taken the initiative in proposing the introduction to Sweden of classical British parliamentarism with the division of government and opposition. A leading liberal M.P., once, in a polemic statement against Herbert Tingsten and his pleas for a PA system, emphasized that working with alternatives and not committing oneself to a single solution was elementary in practical politics: "We think therefore that the question of a four-party government ought to be retained as one alternative among several. That a liberal party, however, with its centrist political position would make the demand for joint government its only alternative would, in our opinion, be the most advanced political stupidity which generally can be accomplished in this connection."¹¹

The Social Democratic party stands out undeniably as the Swedish party which

would seem to have greatest reason to feel satisfaction within the existing system inasmuch as it has dominated the political scene for nearly four decades. Nor does PA seem to have offered any attraction. The arguments levelled against a TA government recurred with, if possible, still greater unequivocalness concerning the question of making such a government permanent. On the other hand, P2 appears to have offered some attraction. These sympathies for P2 which have existed — and still exist - within the Social Democratic party can however hardly be explained by a sense of dissatisfaction. Rather they coincide with the party's fundamental view on cabinet government: a government ought to be based on a homogeneous majority in parliament in order to be able to achieve a desirable effectiveness in its work. In a P2 system there are - whatever the outcome of an election - maximal guarantees for obtaining majority governments formed by a single party. At the same time as a P2 system is in this way connected to the Social Democratic fundamental of view on cabinet government, it is contrary, however, to the Swedish conditions in which the Social Democrats with indisputable success have operated, and with the strategy the party has employed: to drive, if possible, a wedge between the non-socialist parties, inter alia by offering close cooperation with one of them. Through its attitude toward P2 the party to some degree has long been confronted - and still is - with a conflict between principles and actual behavior.

Another matter is that the Social Democrats in spite of everything feel a certain dissatisfaction within the existing social or economic system. This was especially evident during 1966-1968 when there appeared to be great risks of a Social Democratic defeat in a general election. In a broader perspective the party, despite its lasting majority position in the riksdag, actually feels that it is in an inferior position. It lacks the large power resources which the non-socialists are considered to have at their command in the form of support from the press, business, conservatively inclined officials and experts. This feeling of weakness has, in turn, been regarded as constituting grounds against a modification of the system in the direction of PA and P2. A permanent all-party government would further strengthen the non-socialist parties' already strong position in society. A P2 system, if the non-socialist party won the majority and thus could form the government alone, would mean a harmful accumulation of power for the other party, thereby tipping the balance in society.

For the Social Democrats, the ideal state of affairs - within a social and economic system that they feel somewhat dissatisfied with - has instead appeared to be that which has now existed for approximately four decades: a system of majority rule in a multiparty structure where the majority is represented by a single party and where the division into more than two parties has facilitated a large measure of cooperation at other levels than the cabinet level.

---

NOTES

1 Gf. S. Rokkan, "Norway: Numerical Democracy and Corporate Pluralism" in R. A. Dahl (ed.) Political Oppositions in Western Democracies, New Haven: Yale University Press, 1966.

2 Cf. concerning this view of polarity, N. Stjernquist, "Sweden: Stability or Deadlock?" in Ibid.

---

3 Cf. various papers presented at the International Political Science Association's seventh congress in Brussels 1967, the main theme of the congress being "Recent research on typologies of political regimes and political development".

4 Several of the papers presented at the IPSA congress mentioned in note 3 contained different suggestions for classification basedon the number of parties and the relative strength of the parties.

5 A good account of the Swiss system and the composition of the Federal Council even during the postwar period is provided in Gh. Hughes, The Parliament of Switzerland, London: Cassell & Gompagny Ltd. 1962.

6 Cf. A. Downs, An Economic Theory of Democracy, New York: Harper & Row, 1957 and G. Sjöblom, Party Strategies in a Multiparty System, Lund: Studentlitteratur, 1968.

7 For a discussion of the supplementary pension question in Sweden, see B. Molin, "Swedish Party Politics: A Case Study" in Scandinavian Political Studies, Vol. I, 1966.

8 Version of B. Ohlin's speech in Eskilstuna 20/9 1957 in Stockholms-Tidningen 21/9 1957.

9 Statement by Per Edvin Sköld made during a meeting of Social Democratic party board, Socialdemokratiska arbetarepartiets styrelse 5/11 1944.

10 G. Sartori, "European Political Parties: the Case of Polarized Pluralism" in J. LaPalombara & M. Weiner (eds.) Political Parties and Political Development, Princeton: Princeton University Press, 1966.

11 W. Svensson i Ljungskile in Frisinnad Tidskrift, 1951, p. 136.