University of Aarhus

Using a quantitative approach this paper deals with the relations between government and opposition in the Danish Folketing. A central question concerns the applicability of quantitative research methods developed in the United States. To what degree must the methods be modified when used in a context sharply differing from that of the Congress or American state legislatures? This question is important today in the Nordic countries where the behavioral approach has spread only slowly to the fields of political science outside the earlier established and more profound research in voting behavior.¹

Since the pioneering studies of A. L Lowell and Stuart A. Rice, much American research in legislative behavior has used the roll calls as data.³ The predilection for this kind of research is understandable: roll call data are easily accessible, highly standardized, well-suited to analysis by various quantitative methods, and deal with the most central phase in the legislative process — the pivotal moment of decision.

At the risk of oversimplification, we may say that most roll call-studies have centered around two problems. (1) Party cohesion in the legislatures has been studied. To what degree do representatives follow the party majority? What factors explain consistent and/or deviant behavior? (2) Much interest has been given to identification of groupings or blocs among representatives across and within party-boundaries. Variables, such as differences in constituencies, regional differences etc., have been used as explanatory factors.

Apart from some recent studies, which use relatively complicated techniques, most of the research has used a few rather simple indices, most of which were introduced by early researchers in this area. Examples include the concept of "party-vote" (Lowell, 1901); the "index of cohesion" and the "index of likeness" (Rice, 1928); the "index of party loyalty" (Turner, 1951); and especially the method devised by Rice for detection of blocs in small political bodies (Rice, 1928). These and related techniques - all of which use roll calls as data - differ in their aims.

---

* Parts of this paper were presented to the first Scandinavian Conference of Political Science in Oslo, June 1966. The manuscript was finished before the election of November 22, 1966.

Some describe the roll calls. Other describe behavior of individual representatives.
Others identify aggregates of legislators. And finally a group of indices
describes relations between legislators and aggregates of legislators.3

A Danish political scientist specializing in legislative behavior does not draw much inspiration from the American roll call studies, neither concerning problems studied nor techniques used. The results simply are not interesting when you ask questions about cohesion in the Folketing or about tendencies toward blocformation across party-boundaries. Section 2, below will show that cohesion is extremely high in all parties, and party loyalty with only few exceptions is the dominating norm.

Use of the above-mentioned quantitative techniques is out of the question because roll calls are literally absent in the Danish legislative process.⁴ Thus differences in the character of the legislative process make direct application impossible. Also some techniques have been coined for two-party systems of the Anglo-Saxon type, and their logical structures do not fit a multiparty-system.⁵ Finally, difficulties arise because the procedural rules of the Folketing permit the representative to choose among three different ways of casting his ballot, contrasted to the "Yeas and nays" of the Congress.

It is tempting to conclude that Danish legislative behavior research is unable to utilize these rigorous techniques, however, this would not be wholly accurate. We cannot, of course, use the vote of the representative as our analytical unit in legislative research, but it is in fact possible to find other units on which to apply the techniques or similarly — derived techniques.

In fact, one of the purposes of the following provisional survey of some aspects
of Danish post-war politics is to show that such quantitative methods can be
applied to an analysis of divisions in the Folketing.

1. The Problem of Cooperation and Conflict

In the sparse literature ol Danish political science one issue has a central position: relations between government and opposition and the degree of cooperation in the legislative process.⁶ Long periods with minority governments, interrupted by shorter periods with majority coalitions, have generated endless debate as to the forms and prospects of cooperation in the Folketing. This discussion has gone on among post-war politicians and slowly has attracted scholarly interest.

Scientists have turned primarily to "traditional" historical data, e.g. material about the formation or governments, statements about the prospects of cooperation by leading politicians and material stemming from greater legislative conflicts. Danish Parliamentary Government has been described - at least during long periods — as being "two-bloc-parliamentarism", in which two parties, Socialdemokratiet (Social Democratic Party) and Radikale Venstre (Radical Liberals) have formed one cooperating constellation of parties opposed by two

other cooperating parties, Venstre (Liberals) and Konservative Folkeparti (Conservatives). And around these two stable constellations are some small parties but none normally has been able to exert significant influence upon the formation of governments and day-to-day legislation. Relations between the two blocs are considered analogous to the relations between the two great parties in Great Britain.⁷

One type of data has not been used systematically in the establishment of this view. This is data from the divisions in the Folketing, especially the final division, which brings to an end the third reading of the bills. At this time the political parties indicate whether they will be co-responsible for the governmental bill,⁸ wash their hands of it by abstaining, or vote against the bill and thus establish a basis for attacks on the government.

With some qualifications, it seems evident that the attitude of the Danish parties toward the work of the government becomes most clear in the divisions. And since it is possible to delineate the voting behavior of every party in every division since the end of World War 11, it is possible to contribute to the discussion of cooperation and conflict by analyzing this series of division-results. The author states emphatically at the outset that this analysis can never supersede other types of analysis already made on the subject. The different approaches are complementary.

Section 2 will show that party groups in the Folketing possess so much cohesion that it is possible to look upon every single division as a "piling of bricks" (Meyer, 1965, p. 95), by which a constellation of parties is forming a majority in favor of the bill against an abstaining and/or outright opposing minority.

If we want to study long-term development and possible trends, it is necessary to compress a mass of data into a few relevant and longitudinally comparable indices. This may tell us something about the tendencies of conflict and cooperation in the legislative process. Since 1945, 2600 governmental bills have been voted upon in the Folketing.⁹

At this point the Anglo-Saxon roll call-analysis comes to mind. With parties as coherent as the Danish parties, it is possible to treat the votes of a party's representatives as an aggregate and the aggregates analogous with individual votes.

Instead of asking how many representatives agreed/disagreed with the government's bill, we ask how many parties agreed/disagreed. With techniques related to those of the roll call studies, we can ask meaningful questions about the degree of agreement among the parties in the Folketing and about the tendency towards bloc-formation of parties in the divisions.

Let us use an analogy. International politics as an academic discipline often treats international relations as a special kind of interaction between "grouppersons", i.a. states (Carr. 1964, p. 149). In the same way, the analysis of legislative behavior under certain conditions can find it appropriate to treat relations between representatives as relations between "group-persons", i.a. the parlia

mentary groups of the parties. We can use concepts and analytical tools
suggesting those used in behavioral research treating individual behavior.

2. Party Cohesion In the Folketing

It is impossible to form any precise picture of the occurrence of breaks in the
party lines of the Folketing. Aside from roll calls, which occur seldom, the
divisions are anonymous divisions in which individual members vote by rising.

Aside from the fact that this method of division can easily result in counting mistakes — and perhaps raise doubt about the result of a division, it prevents a precise identification of the noting of individual members. If we disregard the numerous cases where the "dissenters" indicate they intend to vote differently from the majority of their group and if we disregard the instances where it is possible to see through the anonymity in other ways,¹⁰ we are prevented from stating precisely which parliamentary members dissented.¹¹

For reasons irrelevant to tho legislative process,12 it seems impossible to make
an even partial survey of dissention voting for the first part of the post-war period.
However, in this respect the situation improves for 1953—65.

For this 12-year period it is possible to determine the number of divisions in
which breaks occurred and v.o determine, in part, how many and which party
group members dissented from the group majority. These dissentions are

registered in Table 1. In light of the qualifications, the table should be read with caution. Also one should be aware that it registers both the dissenting yes and no votes, as well as cases where dissention meant one or more members abstained from voting.¹³

Table 1 shows that breaks in the cohesion of the party groups have occurred in about 12 per cent of all divisions. If one compares the sum totals with the figures for individual parties, it is seen that many of the breaks occur simultaneously, i.a. in the same division. In the majority of cases only one or two members of a group broke party lines, while much less frequently several members at the same time voted differently from the majority of the group. There do not seem to have been such extensive cleavages during this period that it has been impossible for a spokesman of the group to indicate a majority position during the debates. Thus, there actually is a basis for considering the party groups as "group persons".

Although no connection between the parliamentary strength of the governments and the tendency towards cohesion within the individual parties can be demonstrated, we can point out some interesting general differences in the degree of cohesion within the individual party groups.

The fact that DKP at no time during its representation in the Danish Folketing suffered a break in cohesion conforms to the observation of voting behavior in other Communist parties.¹⁴ Of the remaining "left wing" parties, Socialdemokratiet also has been marked by a noticeable cohesion, particularly when one considers that dissentions occurring in recent years almost without exception have been caused by one single member of the parliamentary group. On the other hand, from the beginning the newly created SF has shown less cohesion — a fact which manifested itself in a complete break up (in two stages) of the group during the session of 1963/64 (Folketingsårbogen 1963//64, pp. 10 ff).

The small parties display no less cohesion than the large ones. Only one of the small parties deserves a special comment, namely, Danmarks Retsforbund (DR). This is because two of the breaks in the party unity occurring during the 1957—60 period while the party was represented in the government were caused by ministers who took a stand towards the government bills, different from the remaining members of the group.¹⁵

Just as noteworthy is the considerably smaller degree of cohesion in the parties Venstre (V) and Det konservative Folkeparti (KF). Breaks here reveal themselves not only in the form of the individual member's dissention but also in a number of cases where a more comprehensive break in party lines has occurred.

This is not the place for a more thorough analysis. But on the basis of the investigations involved in Table 1, it is natural to mention that a further analysis undoubtedly would identify four main groups of instances where breaks in cohesion occur.

First, an appreciable number of dissentions are caused by some individual
members of parliament.™ These dissentions have appeared mainly in the passage

of financial bills over which conflicts have arisen or which were the results of
"kriseforlig".

Beyond these persons, other could be named who singly have taken a dissenting stand at various times. They can be characterized by having a current or an earlier association with the large interest groups. Also, their dissenting votes appeared in connection with passage of a bill opposed by the organizations concerned.¹⁷ These instances have been noticed particularly with bills which meant an interference in the recognized autonomy of labor market organizations.

The third typical group of bills often giving rise to dissentions is not easily described in general terms. When bills concerning religious, traffic, sexual and some other matters are passed, cohesion is slight, especially in V, KF and U. In summary, these dissentions concern questions which give rise to strong emotionally toned public debate.

The fourth group consists of bills about conditions in delimited geographic localities. During the reading of a bill considered especially relevant in a part of the country, the members of parliament representing that particular part sometimes take the same position irrespective of the divergent views of the parties.¹⁸

In the enumeration of these four types, no conclusion has been reached as to
the extent one can distribute numerically the breaks in cohesion on these types.
The phenomenon of dissentions deserves further scrutiny.

The conclusion from this brief description is that even if breaks in the unity of the Folketing groups occur more frequently than is generally assumed, especially by critics of party discipline in Danish parliamentary groups, they are quite clearly the exception. Thus, we can consider the groups as "group persons". Further, we have shown that the degree of cohesion of the groups in the Folketing cannot adequately be described by quantitative methods developed in American research. And if it had been possible to apply them, they would not have been sensitive enough to uncover the differences among parties.

3. The Extent of Agreement

On the basis of his studies of roll calls in several American states, Rice demonstrated
that the extent of agreement in divisions was appreciable:

"In fact, it is safe to generalize: In every public "deliberative" body
disagreement upon the final disposition of a question
is the exception rather than the rule." (Rice, 1928, p. 118)

The Danish Folketing is no exception. The majority of the divisions at the third reading are unanimous.™ And if we reduce the requirements on total agreement, we can show that the majority of bills proposed by the government are passed with the consent of at least the four so-called "old" parties, SO, RV, KF and V. These four parties in the period after 1945 have held about 85—90 % of the

seats. This explains why agreement among these parties is often described as
"broad agreement" in the political debate. From 1945 to 1965, the 2600 divisions
have been distributed as follows:20

I "Unanimity"

58.4 %

II "Broad agreement" 26.3%
111 "No agreement" 15.3%

It is natural to ask to what extent deviation from this simple average has occurred in the period. Is it possible to find variations in the degree of "unanimity" and the degree of "broad agreement"? Is there a connection between the movement in these variables and the shifting parliamentary constellations in the postwar period?

We can see in Figure 1 how general acceptance of the government's bills immediately following the war was replaced, even before the beginning of the 1950's by strong opposition among the small parties, DR and DKP. At the same time, the first eight post-war years were marked by a strong tendency toward cooperation among the four "old" parties in the divisions. With the formation of the Hans Hedtoft cabinet in 1953, this situation changed noticeably. Under this weak minority government, which at the time of its succession had not received any definite support from the other parties, opposition increased, mainly because the two great parties, V and KF dissociated themselves increasingly from the government's bills. It is interesting to note that while this part of the opposition raised stronger objections, there was in the years 1953—57 a slight tendency toward greater agreement between the government and the two small opposition parties.

Besides "long" trends, Figure 1 also reveals some tendencies which manifest
themselves from year to year.

First, neither do the larger ror the smaller opposition parties change their attitude towards the government's bills in the parliamentary year just before an election. The survey reveals nothing about an unusual increase or decrease in their voting against or abstaining from voting.

Second, one notes the predominantly parallel course of the two curves. The voting of the four "old" parties and the smaller parties seems related in such a way that an increased tendency toward voting for or against the government bills in the "old" opposition parties follows a similar tendency in the DR and DKP parties for the 1945—60 period and in the SF and U parties in more recent years.

We speak here only about the extent of agreement in the final divisions. It is tempting from all of our material to make inferences about the extent of agreement in the Danish political system and in the legislative process as a whole, or, in other words, to use our data as "surface indications" (Rice, 1928, p. 121). In the last section of this paper, I will indirectly dicuss the possibilities of making such inferences. The problem is very intricate, and a full analysis would require a thorough examination of the whole legislative process, as well as scrutinizing all the pivotal foci, in which conflicts and consensus are revealed. What I will do here only is mention the most apparent significant source of bias in inferences of this type.

A fallacious inference may occur if we overlook the fact that opposition parties do not alone determine the extent of agreement and disagreement. It is not only parties outside government that exert influence on whether a government bill is passed unanimously or with generally "broad" approval in the Folketing. The government itself influences the handling of its own bills, largely because it decides which bills should be submitted how and when, but also because it controls the bill's treatment in the parliamentary committee. Despite the provision in the Folketingets Forretningsorden "that the committee chairman should facilitate the business of the committee as much as possible", the chairman has power to hold a bill back. Thus, since the chairman usually belongs to the government party, it is possible for the government to prevent a bill for which it is impossible to collect the necessary majority from being brought to a division in the Folketing within the parliamentary session.

Against this background, a justified objection to inferences from the material in Figure 1, would be that it only contains a registration of the parties' stands on the bills which reached the final vote. Therefore, conclusions about the extent of agreement on government bills in general are possibly encumbered by a serious bias, namely, the possibility that governments whose bills meet strong objections from opposition parties allow controversial bills to be "buried" in the parliamentary committee and only allow bills to reach the final divisions when agreement has been reached.

If it were possible to prove such a tendency, the inferential value of the yardsticks for agreement in this section would be low. To be able to justify their usefulness as a basis for wider conclusions, it is necessary to investigate more carefully to what extent and under which parliamentary constellations government bills do not reach the third reading. An answer to these questions is hinted at in Table 2, which shows eight "percentages of accomplishment".

Before it is possible to judge the strength of the objection stated, this calculation of "percentages of accomplishment" must be combined with an investigation of the reason why every detained bill did not reach the final division. We also then have to examine to what extent a bill that was not passed was reintroduced and accepted in the following parliamentary session. However, it already seems clear from figures in Table 2 (compared with Figure 1) that the objections cannot be totally damaging.

The year 7953 again stands as a demarcation line in Danish post-war politics. Before this year, a high "percentage of accomplishment" was connected to high agreement among the four "old" parties in the divisions. After 1953, this agreement generally has been much smaller. At the same time a connection between the size of "percentages of accomplishment" and the government's basis in parliament seems to reveal itself. Majority governments have been characterized by a high rate of accomplishment, minority governments by a considerably smaller.

Consequently, the post-war governments in general have not resorted to "burying" bills in the committees when these bills met with strong objections from the opposition parties. A closer scrutiny would certainly show us that bills which could risk a majority against them were left in the committees, while many bills were forwarded, which could only expect a small majority.

These reflections show clearly that inferences based on the time series in Figure 1 require caution. Further analysis has to be made before setting up and testing hypotheses about the conditions of cooperation under the changing parliamentary situations in the years since 1945. In the next section I will pursue one of the directions such an analysis must take, and in Section 6 I will comment further on the factors which produce the high agreement in the final divisions.

4. The Attitude of Individual Parties toward Government Bills

At the conclusion of the tiird reading a party can vote for or against, or can abstain from voting. A survey that permits a comparison of the attitudes of the various parties towards the government bills at different times requires technical aids in the forms of indices, i.a. calculations which compress the large amounts of data and neutralize irrelevant factors. If we want to set up an index, the alternative possibilities for voting behavior must be considered. Also, we must realize that the number of government bills passed shows considerable fluctuation from year to year during the period. Thus, to be able to make longitudinal comparisons, we must neutralize movements in that variable.

With these two conditions in mind, I have set up an index which has as its purpose the registering of the extent to which various parties vote differently from the government party (parties) and which, therefore, can be called appropriately an "index of distance".

This index is based upon comparison in pairs of the parties' voting behavior - This comparison registers th9 differences and similarities in the behavior of the two parties concerned.²¹ The procedure compares the behavior of the parties in every division in a given period (generally the parliamentary session, cf. note 13). Every time the two parties have voted the same way, it has been indicated by O point. Each time one party has voted for a bill, while the other voted against, the point value has been 1 point; and every time one party has abstained, while the other has either voted for or against, the point value has been 1/2. This scale:

of course assumes that a qualitative difference among the three forms of voting allows itself to be expressed in quantitative terms and that the arbitrary point scale expresses a realistic judgement of the parties' own evaluation of the differences in position.

By summing point values o^f all the votes compared in the session considered, we get a point total in absolute terms. Placing this in relation to the number of divisions in the session, we get an index number which, combined with similar numbers from other sessions, produces a time series of the "distance" between the two parties, that is, their tendency towards heterogeneous voting. The index assumes the value 0 if the parties voted completely uniformly during the period, and the value 100 if they took diametrically opposed stands on the bills.²²

When index values have beien computed for every party-combination in every
parliamentary session since 1945, these values can be diagrammed to permit an
easier survey of the longitudinal development.

Figure 2 illustrates the distances between Socialdemokratiet and every other party represented in the Folketing. If I had followed strictly the general approach in this paper, the diagram should have contained a comparison of the government party (parties) and the other parties. As such, a comparison would be difficult to survey in diagrammatic form because of the changing government formations. I have chosen to compare Socialdemokratiet with the other parties instead. Because SD has formed governments throughout the post-war period (aside from the years 1945—47 and 1950—53) and the distance between SD and the government parties during the

1950-53 period was minimal, it is possible to form an approximation of the
distances between the government parties and the other parties in the period
1947-65.

Before we fully discuss the relationship between the parties in the Folketing, we should point out that Figure 2 supports and elucidates assertions brought out in Section 3, especially the point that 1953 forms a demarcation line in the postwar development.

5. The Danish Oppositions

The Government-Opposition dichotomy, inherited from the English Parliament, is politically well established in all Western democracies. Politicians and political scientists are accustomed to dichotomize the political parties under these two labels. Upon closer scrutiny, however, it is seen that in most countries this twinconcept is a rather ambiguous descriptive term. Denmark is no exception. In Denmark the ambiguity indeed seems overwhelming. One is tempted to follow Robert A. Dahl's description of the American scene: "To say where the government leaves off and the opposition begins is an exercise in methaphysics" (Dahl, 1966, p. 34). Thus, while we have no difficulties with the identification of "the government", it seems impossible to put under one formula the behavior of the parties outside government.²³

By way of introduction let us take the most significant although not the most typical pattern of "governrrient-opposition"-relation, that has emerged in the post-war period. In 1950 a Social Democratic minority government resigned a few months after an election/The resignation was released by a division in an interpellation debate in which the government was defeated by a majority consisting of Venstre, Konservative Folkeparti and Danmarks Retsforbund (Rigsdagsårbogen 1950/51, pp. 571—57(5). The new government was formed by Venstre and KF.²⁴ Under this government, which lasted until 1953, the defeated Socialdemokrati, the largest party in "the Opposition", never voted against the government during three full years (cf. Figure 2).²⁵ With the existing distribution of seats, the Social Democrats actually had no possibilities of forcing the government to resign. Nevertheless, the Social Democratic attitude toward government bills departs markedly from the prevailing conceptions of the "opposition role" and stands in sharp contrast to the patterns of opposition in the remaining part of the post-war period.

We have seen that the Danish party system is a multiparty system with highly cohesive parties. The electoral law favors a strictly proportional representation for those parties which have overcome the threshold of representation, and several new parties have appeared in the post-war period.²⁶ Under these conditions the elections have beem highly competitive.²⁷

But what about the legislative work? I have sketched above the remarkable
relation between the government and Socialdemokratiet in the years from 1950—

1953. This relationship was unique, and the factors producing the situation were numerous and much too difficult to evaluate here. Instead let me pose another more general question: Are the parties not-in-government competitive vis-å-vis the holders of governmental power? Does the behavior of these parties justify the label "The Opposition", or do they differ so much in behavior that such a label loses its descriptive value? To get an answer, we might inspect the patterns of party behavior which the index of distance has revealed.

Three distinct patterns of opposition emerge in the Danish Folketing in the later years while Socialdemokratiet has been in government, alone or in coalition. These patterns range from close collaboration with the government in the divisions to a certain though not considerable degree of competition.

First, let us treat Radikale Venstre as an opposition party sui generis. Since its formation in 1905, it has built and almost without interruption maintained a tradition for close collaboration with Socialdemokratiet. As can be seen from Figure 2, this collaboration emerges in the divisions. Since 1945 the party has been in governmental coalition with SD in two periods (1957—60 and 1960—64). In the remaining time it has always favored maintenance of the Social Democratic government, i.e. a change in the personnel of government has never been among its goals. The party, which has no significant relations with the large Danish interest organizations, has — when not in government — found its primary channel of influence in the legislative work, or, more exactly, bargaining in committees and corridors of the Folketing. Because of the peculiar distribution of seats (cf. below, Table 4), its legislative manoeuvering has been very successful, marking the party as the holder of the balance in Danish politics.²⁸

The role as mediator has been a strategy employed not only to advance specific policy goals but has become a goal itself. The mediator-role has been played not only when the Social Democrats were in government but also in the periods 1950—53 and 1945—47. Nothing can express the party's view better than this passage from a recent speech by a leading Radical politician, Hilmar Baungaard:

"Cooperation is a simple necessity under a minority government, and as no party alone is likely to command a majority in the foreseeable future, it is no overstatement that cooperation among the parties is the foundation on which Danish Parliamentary Government and Danish politics must in general build their existence. This will be the attitude of the Radikale Venstre, irrespective of what government we have. Surely it will cost us varying degrees of effort to cooperate, but we will do so." (Folketingstidende 1965/66, col. 2917.)

If it is questionable to refer to the Radikale Venstre as opposition,²⁹ the same cannot be said about the other small parties — until 1960 Danmarks kommunistiske Parti and Danmarks Retsforbund, and since the 1960-election Socialistisk Folkeparti and De Uafhængige. Figure 2 shows these parties — all belonging to the type which Maurice Duverger has christened "permanent minority parties" (Duverger, 1959, p. 290) — have been relatively most competitive among the Danish parties. Several factors may help explain this.

First, having escaped the threshold of representation, the small parties meet the next barrier to legislative influence: only parties with at least ten representatives are able to get members elected to the 17-member-committees, which form the nucleus in the law-making process. Thus, DR, 1945—50 and 1953—60, DKP, 1947—60, and U, from 1960, to mention only the most significant small parties, have been kept from the pivot of the legislature. They have had to be satisfied with talking and taking part in the divisions.³⁰ It is no exaggeration to say that these institutional restrictions on the small parties create a hostile attitude toward the larger parties and a frustrated attitude towards the bills dealt with.³¹

Second, these parties have been, so to speak, in the wings. Except for Danmarks Retsforbund, they have been regarded as splinter parties by the "old" parties. Thus, in spite of deliberate attempts, they have never been able to receive due recognition. Especially between Socialdemokratiet and DKP, and later SF, a deep cleavage has emerged, perhaps the deepest in the Danish party system. Until recently the Social Democratic leaders have outrightly rejected considering the party on the left wing a supporter in government formation and passing of government bills.

Finally, a special kind of frustration among the small parties has resulted because of the total exclusion from the top-level negotiations around "kriseforlig" and other major complexes of legislation. The small parties often are confronted with a iait accompli when the time for parliamentary debate and decision comes. "Broad agreement" has been reached by the leaders of the four "old" parties without consultation of the small parties.

For all the similarities in the position of the Danish "permanent minority
parties", the obvious differences must not be neglected. The four parties, which
we consider here, can be dichotomized in several dimensions.

First, their competitive behavior occurs over different issues. To be sure, all have heavily centered their opposition on bills advanced by the Minister of Finance, including the budget bill. Thus, while the four "old" parties always support the budget, DR, U and SF abstained at the third reading and DKP always voted against. In addition, DKP and SF have opposed bills put forth by the Minister of Deferense, while DR especially has opposed bills of the Minister of Commerce and U seemingly has spread its opposition on a wider range of ministers. We get a more reliable measure of this tendency to oppose on different issues from Table 3. a. which shows it is relatively rare to see the two small parties at a given moment vote against a government proposal

simultaneously. The table indicates that the permanent minority parties have
been situated at opposite wings in Danish politics, especially in the later years.

Second, it is a plausible assumption that the legislature has been regarded differently as a site for encounters with the government by the two sets of parties. On one side, Danmarks Retsforbund and De Uafhængige, parties without well established contacts with powerful interest organizations, without a press of their own and without access to the committees, have been forced to follow a strategy that would create a distinctive party image in the voters' minds. The site of opposition for these parties has been the open debate in the Folketing, and their behavior can be viewed partly as a means of gaining votes in the next election to overcome the thresholds set up by the four "old" parties.

Observers of Danish politics have argued convincingly that Danmarks Retsforbund in particular collected the votes of those who were discontented with one or all of the "old" parties or with "the system", — i.e. those voters who disapproved of the compromises of Danish politics.³² True or not, the entering of a governmental coalition for the first time in the life of the party — shifting from a strategy of legislative competition to one of governmental collaboration (cf. Figure 2) — cost the party all 9 of its mandates, a loss from which it has not recovered.

It is not yet possible to judge the similarities or differences between De
Uafhængige and DR in this respect, but it seems plausible that U also attracts
voters discontented with the strategies of Venstre and Konservative Folkeparti.

In contrast to DR and U, the two parties on the left wing, and especially the Communists, have been able to oppose the government through another channel, trade unions. Only a handful of unions are and have been dominated by Communists, but in many unions they form an active minority accused by the Social Democratic majority of creating discontent and collecting the discontented. It is impossible to tell whether the legislature or the unions was perceived as the most important site of encounter by the Communists in the 1950'5.

The last two parties outside the government, Venstre and Konservative Folkeparti, often label themselves and are labelled by the government "The Opposition". They are considered the only parties able to take over the responsibilities of government. It has generally been assumed that a shift in government will take both of them into office, a view which the two parties' leaders have confirmed on several occasions, especially when the possibility of a Venstre government has been mentioned.³³

In the 19th century Højre (the predecessor of KF) and Venstre were the great antagonists in Danish politics. Recently it has been suggested that the two parties should merge into one large party. The fact is that in this century a slow and sometimes interrupted movement has taken place, bringing them close to each other in political outlook. A significant step was taken in 1950 when the two parties jointly formed the first "bourgeois"-government since the 1920'5. From this year a close cooperation has characterized their relations. In Figure 2 we find similar trends in their overt voting behavior, and if we supplement this

observation with a view of the distance between the two parties, the similarities become still clearer (Table 3. b). Except for Radikale Venstre and SD, no Danish parties have displayed greater similarities in parliamentary voting than Venstre and KF.

This close similarity, at least since 1950, has been founded partly upon agreements between the parties. The significant change in behavior after 1953 was accompanied in 1954 by discussions between the party-leaders on the strategy toward the Social Democratic government (Dahlgaard, 1964, p. 131). Later the collaborative efforts came into the open when the two parties in 1959 at a joint meeting of their parliamentary groups issued a joint program concerning economic policy. This brought Danish politics nearer the Anglo-Saxon model than ever before — with three parties in governmental coalition, two large parties forming the alternative, and only the Communists outside.

Close collaboration between V and KF continued in the following years. In the election campaigns they refrained from fighting each other, and in the Folketing they took the same stand on almost every government bill. But in the elections of 1960 and 1964, Venstre lost votes while KF gained. The result was a displacement of the balance between the parties as measured in seats. In 1957 Venstre had 45 seats and Konservative Folkeparti 30, in 1964 the numbers were 38 and 36.

In the spring of 1965 twc newly-elected members of Venstre suddenly and without notice entered discussions with the government, resulting in an agreement. In a subsequent third reading of a series of government bills about a hotly disputed increase in some indirect taxes, the two members, now retired from the parliamentary group of Venstre, voted for the bills, while the rest of Venstre plus KF voted against. The government had produced the necessary majority, and Venstre was disrupted.

This is interesting in many respects, but our only concern here is that it announced a conflict within Venstre between those who wanted to see a still closer collaboration with KF, eventually fusing the two parties into a large opposition party, and those who, aii the two rebels, thought Venstre should loosen the ties with KF. The conflict culminated in the summer of 1965 with the retirement of the leader of Venstre who had favored a fusion.

These events have been mentioned because the movements in Venstre in the first half of 1965 led to a new constellation of the Danish parties, a loosening of the interactions between V and KF, thereby opening up the possibility of new patterns of opposition in Denmark.³⁴

Since the retirement of the 'V-K-government" in 1953, relations between V and KF, on one side, and SD — and thereby RV — have been characterized by more competition than before. This change soon opened a debate about the role of the opposition. The two large parties were accused of obstruction. It was said that they pursued a "policy of folded arms", and the ideal of "a cooperating democracy" was set up as an alternative, especially by Radikale Venstre (Meyer, 1965, chapter 12).

Debates concerning the principal problems of Danish parliamentary government seldom take a systematic and fundamental form. Most of the contributions in the post-war period bear the mark of the specific context in which they were set forth,³⁵ and the stable part in the arguments of the different parties lend themselves to presentation in an apothegmatic form. If we try to reconstruct the arguments of Venstre and Konservative Folkeparti in the years since 1953, we get the following rather ambiguous statement: the two parties have always in general stated their willingness to cooperate with the government and try to reach agreement. "When it is possible to agree, we naturally ought to agree," stated a Conservative leader once. But in the same breath it has always been stated that "where principal antagonisms exist, we neither can nor ought to blur those antagonisms by compromises, which only compromise the participants".³⁶ Of course, this statement of the principal arguments is crude. First, it is notable to see how the arguments of KF in particular have stressed the Anglo-Saxon model of opposition, while the spokesmen for Venstre have never attached the same degree of importance to the controlling and alternative—presenting functions of the opposition.

And* second, a provisional survey of the principal debates in the Folketing seems to support an assumed correlation between the character of the arguments of V — KF and the strength of the government. Even with these modifications, one may dare to say that the quotation above indicates the stable main features of the two parties' arguments.

6. The Formation of Consensus

We have now briefly identified the three dominant patterns of oppositions in
Danish politics and have discussed some of the factors, both ideological and
institutional, which can help to explain them.

Our examination of the distances between the parties in the final division has revealed considerable agreement among all of them, especially among the four "old" parties which, in more than one respect, form a nucleus inside the party system.

We must conclude that although we are able to identify some patterns of
opposition by an analysis of the divisions, these patterns only emerge in a few
instances during the parliamentary session.

Let us now return to the assertion of Rice, mentioned in the opening lines of Section 3. Rice told us that "disagreement upon the final disposition of a question is the exception rather than the rule." He would further lead us to believe that his generalization possesses universal validity. This is certainly not true — irrespective of what we mean by "rule" and "exception". What Rice does not tell us is why it is so. Neither does he tell us that different political systems are characterized by the different extent of agreement they reveal.

Obviously a comparative political science must probe these questions. It is
relevant to ask questions about the extent of agreement in the final divisions of

different political systems and to consider the factors which can explain the
differences and the similarities.

When I said that the Dan sh divisions are characterized by a "considerable high agreement" among the parties, I, of course, based this on an implicit comparison with other political systems. And when I now enter a short and fragmentary dicussion of some of the factors which can help us to understand this high extent of agreement, I steadily make comparisons implicite. What kind of factors are operating in Danish politics but are absent from systems with low agreement in the final divisions?

First, Denmark is a country totally lacking those cleavages which elsewhere produce bitter conflict among the citizens and among parties in the legislature. No minority problems exist, ;>uch as religious, ethnic, regional or other types of subcultures, within the territorial borders.³⁷ In the post-war period no party — not even the Communists — has stated as its goal an alteration of the structure of the political system; there has been no "opposition of principle" (Kirchheimer, 1957). And only DKP and, later, SF have desired to make substantial changes in the socio-economic structure.

If we define consensus as high agreement on policy questions, those of fundamental as well as secondary importance, then the Danish voters are characterized by consensus. It is a plausible hypothesis that, all other things being equal, a political system with consensus in the electorate will tend to display consensus among the representatives in the final divisions. If relatively few political questions produce disagreement among the voters, then we would except relatively slight disagreement in the legislative process too.

Second, typical of the Danish political system, most of the governments after 1945 have been minority governments. With two exceptions — the majority coalitions of 1957—60 and 1960-64 - the governments have been of this type, either a oneparty government such as in the periods 1945—47, 1947—50, 1953—57 and 1964 ff., or a minority coalition such as from 1950 to 1953. Foreigners often wonder that minority governments of such a stable and enduring character can exist in Denmark (Andrén, 1963, p. 61). I will not pursue this question but only mention that the Danish political system is characterized by a relatively stable electorate,³⁸ a fact, which has some essential consequences for the government formations and for the legislative process. E ections have never been decisive but have always opened up several possibilities for government formation. Many combinations of parties have been mentioned in the deliberations during cabinet crises,³⁹ although the traditions for intercourse between SD and RV and between V and KF hinder the establishment of an "Allgemeinkoalitionsfähigkeit" such as the one Val R. Lorwin refers to in Belgium (Lorwin, 1966, p. 178).

What is relevant here is that a majority government for long periods has been perceived as impossible. Doubtless this perception has molded the attitudes of the parties in favor of legislative cooperation. We can see that the governments almost without exception have stressed willingness to enter into discussions with opposition parties and generally have tried to produce "broad

agreement", although this has meant greater modifications of the proposals than strictly necessary to form a simple majority. And since the oppositions at any given time have had no hope for improving their positions by forcing the government to resign or to dissolve the Folketing, they have preferred, instead, to persuade the government to make concessions during the committee stage.

What I propose is that the existence of minority governments — all other things being equal — makes it more probable that consensus will exist among the parties in the final division than would be the case with a majority government. Several other factors related to this minority status of the governments can be noted, all of them supporting the proposition.

The distribution of seats (Table 4) shows the conditions under which these

post-war governments and oppositions have had to work. The governments' dilemma has been that only if they were able to obtain support from one or, as has happened more often, more than one party, could they be assured of survival, that is, of getting their bills through the legislature.

Under such circumstances, government bills, to a great extent, reflect the anticipated reactions of the government, to use the term of Carl J. Friedrich (Friedrich, 1963, chapter 11). In other words, only those bills are forwarded which have a real chance of being accepted by the legislature. That is, we would expect higher agreement on the bills at the outset in legislatures in which the government only commands a minority than would be expected in systems with majority governments.

Next, two factors are at work during the deliberations in the Danish legislature, both tending to strengthen the consensus in the final division. First, as we have seen in Section 3, some bills are "buried" in the committees and more bills are "buried" under minority governments than under majority governments. These bills usually have met with resistance from the oppositions. Second, as we already have noted, many bills are amended during the legislative process. As a result, parties outside government become willing to support them. If we define a controversial bill as one which is met with resistance by some of the parties, then we can state that the subset of controversial bills tends to diminish during the legislative process, partly because controversial bills "fall out" and partly because they are made non-controversial, usually during deliberations at the committee-stage.

It is difficult to establish empirical evidence for the existence of these subtle relations in Danish politics. Here we cannot follow the thread further. Much research primarily of a comparative character must be done as to the consequences for legislation stemming from the different forms of governmental strength and from the different distribution of political attitudes in the electorate. Nonetheless, with caution, we can conclude that the lack of serious cleavages in the Danish electorate and the specific minority-status of most Danish governments seem to be the principal factors explaining the pronounced consensus in the final divisions in the Folketing.

---

NOTES

1 A survey of the literature on voting behavior in the Northern countries up to 1960 has been given by Stein Rokkan and Henry Valen (Rokkan—Valen, 1960). A similar survey of the modest literature on legislative behavior does not exist but has to be extracted from the research reports in this yearbook. A symptomatic statement on legislative and administrative research in general up to 1960 has been given by Jan-Magnus Jansson (Jansson, 1960). He concluded: "Notwendig ware meines Erachtens eine methodologische Revision des empirisch-wissenschaftlichen Denkens gerade auf diesem Forschungsgebiet. Die nordische politische Wissenschaft hat den herkömmlichen Begriffsapparat der praktischen Politik und der staatsrechtlichen Dogmatik mit einem historisch gefärbten Studium der politischen Wirklichkeit vereint. Diese Synthese ist nicht schlecht und diirfte uns wahrscheinlich vor einigen Ubertreibungen der Soziologie und Psychologie geschiizt haben. Aber die Zeit ist nunmehr dazu reif, die Resultate der anderen Sozialwissenschaften, insbesondere die der Soziologie und Volkswirtschaftslehre, mehr als zuvor zu beriicksichtigen."

2 Bibliographical summaries can be found in Meller, 1960. For more critical evaluations see Eulau, 1961, and Wahlke, 1962.

3 A summary of indices following this line of presentation is found in Farris, 1958, pp. 316—320. Besides Farris, Duncan Macßae, Jr. (Macßae, 1958, pp. 298—312) has evaluated the different techniques and compared them with the technique of scaling.

4 Roll calls are only used a few times every year either in the case of bills, which have been met with strong resistance in the Folketing; or in case of doubts concerning quorum. See further Møller, 1949, p. 104.

5 For example the concept of "party vote" (Lowell, 1901, p. 323), defined as a roll call on which 90 % of one party votes "yea" and 90 % of the other votes "nay", is of no interest in a multi-party system.

6 Sep abnvp all MPVPT* 1985. rhantprs 7 8 and 12. tneethpr w^it.h Rasmussen. 1958.

7 The concept of "two-bloc-parliamentarism" was coined by Erik Rasmussen (1958). Similar, but not identical, views have been set forth by Poul Meyer (1965, pp. 95 and 121) and Nils Andrén (1963, p. 61).

8 In this article I have overlooked the so-called "private" bills, proposed by single members. These bills are limited in number, and most do not survive the committeestage.

---

9 Since 1945, 6—7 parties have gained representation in the Folketing after every election. Four of them (Socialdemokratiet, Venstre, Konservative Folkeparti and Radikale Venstre) have been represented all the time, while a certain circulation has taken place as concerns the minor parties. See further below, Section 5.

10 The newspapers and Folketingsårbogen serve as firsthand sources on several occasions.

11 Nor is it possible to establish which members of the Folketing were present during the divisions.

12 Primarily varying editorial practice in the Folketingsårbogen.

13 No consideration is given to the representatives from Greenland, to Slesvigsk Parti (1 member!), to members who stood outside the parties, or to Liberalt Centrum (1964/65) and Socialistisk Folkeparti during the 1958—60 period. In Table 1, certain short sessions of the Folketing during the period are combined with other sessions to create a clear survey of the cohesion of the party groups during the various government periods. The same simplification is used throughout this paper (i.e. Fig. 1, Fig. 2, and Table 3). A list of party initials used :.n Table 1 and in this paper: SD = Socialdemokratiet (Social Democratic Party) RV = Radikale Venstre (Radical Liberals) KF = Konservative Folkeparti (Conservatives) V = Venstre (Liberals) DR = Danmarks Retsforbund (Singletaxers) SF = Socialistisk Folkeparti (Socialist People's Party) DKP = Danmarks kommunistiske Parti (Communists) U = De Uafhængige (Independents )

14 E. g. Ralph Burton, 1936, has computed an average cohesion = 100 for the French Communist Party, cf. also Peter Campbell, 1953. For the Finnish Communist Party, Pekka Nyholm, 1959 has computed an average cohesion = 94 for the period 1948—51, the average for the other parties ranging from 77 to 87. A general treatment of the problem of party discipline in Communist parties can be found in Duverger, 1959, especially pp. 173-174 ard 197-202.

15 Folketingsårbogen 1957/58, p. 481—482 and 1959/60, pp. 439—451. In the first case (Lov om yderligere forlængelse af lov om bemyndigelse til opretholdelse af en dansk styrke i udlandet), as far as can be seen DR's parliamentary group, including one of the ministers, abstained, while two ministers voted for the bill. In the other case (Lov om forsvarets organisation m.v.) the group, including two ministers, voted for the bill, while the third minister abstained.

16 Holger Eriksen (SD), Thorkil Kristensen (V) and Aage Fogh (RV) have thus on several occasions deviated from the party line.

17 See examples in Meyer, 1965, chapter 10, where a general discussion of these loyalty conflicts may also be found, compare same place pp. 51—56.

18 As an example of this type of bills see Lov om projektering m.v. af en nordsydgående motorvej i Jylland "ned forskellige tilsluttende vejforbindelser (Folketingsårbogen 1964''65, pp. 308 ff.) ar.d Lov om oprettelse af et universitet i Odense (Folketingsårbogen 1963'64, pp. 374 f f.).

19 The word is used here in a different way from that which is the parliamentary usage; in that usage unanimity is said to exist when no party votes against a bill. Here "unanimity" indicates that all the parties vote for the bill. In the official usage, the number of unanimous divisions would be much larger.

20 "Broad agreement" means, in accordance with parliamentary usage, that the four "old" parties together with eventually one or more of the smaller parties voted for a bill, without there being unanimity in its above stated meaning. "No agreement" is N (I abs +11 abs >X 10° defined as: - %, where N (number of passed N bills) = 2600.

21 The idea of pairing the parties stems from Rice, 1928, chapter XVI, "The Identification of Blocs in Small Political Bodies".

22 Summarizing this description in mathematical terms, we get the following: Let Xab represent the number of divisions where parties a and b voted the same way, Y ab the number of divisions where one of the parties abstained, while the other voted either for or against, and Z ab the number of divisions, where one party voted for and the other against. If the total number of divisions during a given period amounts to N, we have the relation: X_ab ~^~ Yab ~^~ Zab =N_' The distance betwen the two parties is then defined as: <°XX_ab+l/2XY_ab + IXZ_ab)XIOO. Dl_ab -- N and it will be valid that °-Dl_ab- 10° as DI ab =O for X&b=N, and DI flb = 100 for Z =N. In passing, it should be mentioned that the index allows measurement of the distance not only between a government party and an opposition party, but between arbitrarily chosen parties (see Tables 3 and and 3 b).

---

23 This view is in accordance with that of Poul Meyer (1965, chapter 12), who has treated the same problems reviewed in this section.

24 An account of this formation can be found in_12—19.

25 When Lov om forbrugerpriser på rugbrød og mælk was passed in 1953, SD abstained after an amendment proposed by the party had been rejected. See Rigsdagsårbogen 1952/53, pp. 223—228. When Lov om ændring i lov om offentlig forsorg (vedrørende den offentlige sko t,e o e spisning) was passed in 1951 RV moved an amendment supported by SD and DKP. At the division in third reading the amended bill was passed by SD, RV and DKP against DR, while the parties in government, V and KF, abstained. About this unique situation, see Rigsdagsårbogen 1950/51, pp. 322—327.

26 1943—47 a little party, DS, was represented; it tried in vain to regain the lost position in the 1964—election. In 1960 two new parties were represented: SF, formed in 1959 by the purged Communist leader Aksel Larsen, and U, formed in 1953 by the former prime minister and chairman of V, Knud Kristensen. U had participated in vain in the elections in 1953 and 1957. In 1965 a new party, Liberalt Centrum, was formed as a splinter party to V by two representatives (see below). It has not yet passed through the electoral purgatory.

27 In this section I have drawn heavily on a series of concepts introduced by Robert A. Dahl. Definitions and a general discussion of the patterns of opposition can be found in Dahl, 1966, chapters 11—12.

28 The position of the party can best be compared to the Swedish Center Party, cf. Stjernquist, 1966.

29 This position has been taken by Poul Meyer (1965, p. 175).

30 In 1966 the parliamentary procedure has undergone a slight alteration giving the small parties admittance to the committees, which are treating bills proposed by the party concerned (Folketingsårbog, 1965—66, pp. 525 ff).

31 To quote an example of supporting statements by representatives of the small parties, the leader of U, Iver Poulsen, stated: ". .. what we want, are to participate in the solution of the community problems, and not be kept outside the wall set up by the four old parties around themselves and the work of the Folketing" (Folketingstidende 1964/65, col. 236).

32 Lassen, 1961, pp. 24 ff and Thorsen, 1965, p. 116. The total lack of electoral surveys explains the vagueness of the statements.

33 How important this conception has ben in the post-war period are fully stated in Tage Kaarsted's account of the cabinet crisis in 1957 (Kaarsted, 1964). In this connection, it has to be mentioned that the concept of "two-bloc-parliamentarism" as developed by Rasmussen (1958) and especially as used by Meyer (1965, p. 95) is based upon the existence of these relatively stable perceptions in Danish politics of what sort of party-constellations were most likely to form a government jointly.

34 A provisional survey of the voting patterns in the divisions in the session 1965/66 does in fact indicate a new pattern of opposition. V did not vote against the government at any occasion, and only in one division the party abstained. In the same session KF voted against the government bills in 7 divisions, and abstained in 5 divisions.

35 Book length discussions by Danish politicians of the problems of Danish Parliamentary Government are utmost seldom. From the later years it is only possible to mention one contribution, namely by the former chairman of the Folketinq, Gustav Pedersen (1962). The memoirs of one of the leaders of RV, Bertel Dahlgaard, (Dahlgaard, 1964) gives a good picture of the main arguments of RV.

---

36 The quotations are taken from a speech of Poul Møller, Folketingstidende 1962/63, col. 146. For other examples so Meyer, 1965, chapter 12.

37 In the Folketing three regional minorities have been represented in the post-war period. The two representatives of the Faroe Islands have been members of the parliamentary groups of SD and V, respectively, that means integrated in the general party system. The two representatives of Greenland most of the period have taken a neutral stand on the government bills. In 1960 one entered the SD—RV government and became the pivotal vote in the Folketing. Without him, the occurrence of ties had been possible; with him, the government commanded a minimal majority. This situation, which ended in 1964, nevertheless did not mean a continuing abandonment of the principle of neutrality. Finally, the German minority had one representative from 1953 to 1964. The Germar representative, too, tried to avoid a direct intervention in the balance between the other parties. On one occasion, he nevertheless gave pivotal support to the government, arguing that he would not risk that Slesvigsk Parti caused a situation in Danish politics reminiscent of the situation in the Weimarrepublic (Folketingstidende 1961/62, col. 783—786). In other words, the latter two minorities have generally uphold an attitude of modified neutrality. They have deliberately tried to avoid to be the vote(s), that was pivotal against the government.

38 See Jan Stehouwer, "Long Term Ecological Analysis of Electoral Statistic in Denmark" in this volume.

39 A survey of the intricate deliberations can be found in Kaarsted, 1964, especially chapter 1. A comment on thi« aspect of Danish government formation is given by Herbert Tingsten (1966, chapter 6, especially p. 182).

40 The government 1960—64, composed of SD, RV and one of the representatives of Greenland is sui generis, although formally a majority coalition. See note 37.

References:

Andrén, Nils, Nordisk statskunskap, Stockholm: Bokförlaget Liber, 1963.

Burton, Ralph, "The Frencn Chamber of Deputies. Party Attitudes". American
Political Science Review, 30, (1936), pp. 549-555.

Campbell, Peter, "Discipline and Loyalty in the French Parliament during the
Pinay Government." Political Studies, 1, (1953), pp. 247-257.

Carr, E. H., The Twenty Years Crisis 1919—39. An Introduction to the Study of
International Relations. New York and Evanston: Harper & Row, 1964.

Dahl, Robert A., ed Oppositions in Western Democracies. New Haven
and London: Yale University Press, 1966.

Dahlgaard, Bertel, Kamp og Samarbejde. Nærbilleder af politik og politikene
gennem 40 år. København, Fremad, 1964.

Duverger, Maurice, Political Parties. Their Organization and activity in the

modern state. 2. English ed., London, New York: Methuen, 1959.

Farris, C. D., "A Method of determining ideological groupings in the Gongress",
The Journal of Politics, 20, (1958), pp. 308-338.

Folketingsårbogen 1945/46 - 1964/65 (Before 1953/54 the title was rigsdagsårbogen).

Folketingstidende 1945/46 - 1964/65 (Before 1953/54 the title was rigsdagstidende).

Friederich, Carl J., Man and his government. An empirical theory of politics,
New York, Me GrawHill, 1963.

Jansson, Jan-Magnus, "Dsr Staatsapparat als Gegenstand der politischen
Wissenschaft in den Nordischen Ländern". In Otto Stammer, ed., Politische
Forschung. Köln und Opladen: Westdeutscher Verlag, 1960.

Kaarsted, Tage, Regeringskrisen 1957. En Studie i regeringsdannelsens proces.
Aarhus: Universitetsforlaget, 1964.

Kirchheimer, Otto, "The Waning of Opposition in Parliamentary Regimes".
Social Research, 1957, pp. 128-156.

Lassen, Aksel, Folkevilje eller vælgerlune. København: Fremad, 1961.
Lorwin, Val R., "Belgium: Religion, class, and language in national politics".

In Political Opposition in Western Democracies, op. cit. pp. 147—187.

Lowell, A. L., The Influence of Party upon Legislation in England and America.
Annual Report of the American Historical Association, 1901, vol. 1.

Macrae, Jr, Duncan, Dimensions of Congressional Voting. Berkeley and Los
Angeles: University of California Press, 1958.

Meller, Norman, "Legislative Behavior Research", Western Political Quarterly,
13, (1960), pp. 131-153.

Meyer, Pou'. Politiske Partier. København: Nyt Nordisk Forlag, 1965.
Møller, Jens, "Tingenes Forretningsordner". In Den danske Rigsdag, vol. 4,

København: J. H. Sch'Jltz Fonag 1949, pp. 7-117.

Nyholm, P. G., "Havaintoja Eduskuntaryhmien Koheesiosta Vaalikaudella 1948
-51". Politiikka, 1959, pp. 42-52.

Pedersen, Gustav, Folkestyret ved korsvejen, København: Forlaget Idag, 1962.
Rasmussen, Erik, "Parlamentarismens typologi i svensk teori og dansk politik.

Nogle hovedpunkter". In Statens Almindelige Videnskabsfond, Årsberetning
1956/57. København, 1958, pp. 18-27.

Rice, Stuart A., Quantitative Methods in Politics. New York: Alfred A. Knopf,
1928.

Rokkan, Stein and Henry Valen, "Parties, elections and political Behavior in
the Northern Countries". In Politische Forschung, op. cit.

Stjernquist, Nils, "Sweden: Stability or Deadlock". In Political Oppositions in
Western Democracies, op. cit. pp. 116—146.

Thorsen, Svend, Kuling over Christiansborg, København: J. H. Schultz Forlag,
1955.

Tingsten, Herbert, Från ideer till idyll. Den lyckliga demokratien. Stockholm:
P. A. Norstedt, 1966.

Turner, Julius, Party and Constituency: Pressures on Congress. Baltimore:
The John Hopkins Press, 1951.

Wahlke, John C., "Behavioral Analyses of representative Bodies". In Austin
Ranney, ed., Essays on the Behavioral Study of Politics. Urbana: University of
Illinois Press, 1962, pp. 73-90.