Recent research on minority governments suggests that, everything else being equal, minority government would prefer purely ad hoc coalitions (Strom 1990, 108). By negotiating each issue separately and on an ad hoc basis, Strom contends that the typical minority government can in each case select the least 'expensive' coalition partner available (Strom 1990, 108-109). This conclusion is based on a series of implicit assumptions, one of which considers parties to be unitary actors. The objective of this article is to break free from this assumption and examine the intra-party consequences of majority building by minority governments in the Danish and Norwegian party systems.¹ The study examines the impact of conflicts within highly institutionalized parties on their coalitional behaviour insofar as such conflicts may influence the bargaining power of party elites in the parliamentary arena. The focus is on how highly institutionalized parties cope with internal conflicts, or, alternatively, avoid them in the first place.

Before going any further, it is appropriate to clarify what is meant by

Resumé

The purpose of this article is to analyse the impact of intra-party conflicts on the coalitional behaviour of highly institutionalized parties in Denmark and Norway. The focus is on how highly institutionalized parties either prevent or cope with internal conflicts. The account proposed is based on the idea that organizational rigidity, that is the lack of heterogeneous and diffused mechanisms for internal dissent, does not allow the party to handle internal dissent in a variety of flexible ways. When inter-party negotiations induce internal conflicts, members may be forced to leave the party as their primary mechanism for the expression of dissent. A major implication of this view, which is supported by the empirical findings, is that in multi-party systems where minority situations occur, the most attractive strategy (in terms of bargaining power) for highly institutionalized parties is the formation of informal minority governments. When such parties form formal minority governments, the party elites tend to impose structural constraints on the day-to-day operation of the government and modify their alliance strategy in order to cope with the derived internal conflicts.

highly institutionalized parties. According to Panebianco's (1988) analysis, a highly institutionalized party is characterized by a high degree of control over its environment (i.e. it is relatively free from the influence of external or extra-parliamentary organization), and interdependence among its subgroups.Operationally, party can be classified as highly institutionalized according to five indicators:

1. The degree of development of the central extra-parliamentary organization.
rule here is that a highly institutionalized party possesses a
developed central bureaucracy.

2. The degree of homogeneity of organizational structures at the same hierarchical level. In highly institutionalized organizations the local associations tend to be organized in the same way throughout the national territory.

3. The method of finance. The highly institutionalized party has at its
disposal a revenue system basedon a regular flow of contributions from
numerous sources.

4. Relations with the external collateral organizations. A highly institutionalized
dominates its external organizations, such as tråde
unions.

5. The degree of correspondence between a party's statutory norms and its actual power structure. This correspondence tends to be greater in highly than in weakly institutionalized parties. Hence, people, groups or associations formally outside a highly institutionalized party cannot play leading roles within the organization (Panebianco 1988, 58-59).

In this paper, Panebianco's criteria for classification are accepted without delving into the theoretical and practical problems which complicate the various approaches of party institutionalization. 'Party institutionalization' will be defined very loosely as the ways in which political authority and instruments of control over internal uncertainty are distributed within a party. Given this definition, Panebianco's indicators of the degrees of institutionalization are used to determine whether a party is highly institutionalized a given period.

The account proposed is based on the idea that the lack of heterogeneous and diffused mechanisms for internal dissent - a characteristic of highly institutionalized parties - undermines the ability to cope with internal dissent. As a result, such parties cannot enter into conflict inducing coalition negotiations with other parties without risking their hold on their own members. Thus, highly institutionalized parties can bring about their own disintegration when they enter serious coalition negotiations with other parties.

A major implication of this view - the main hypothesis in this case - is
that in multi-party systems in which minority situations occur, the most
attractive strategy in terms of bargaining power for highly institutionalized

parties to pursue is the formation of informal minority governments. This argument is based on the assumption that party elites are motivated, above all, by the desire to remain party leaders (Luebbert 1986). In other words, for party leaders who are in office, it is more important to remain party leaders than to remain in office. If intra-party conflict threatens their position, rather than stay in government they will leave office in order to remain party leaders. Party leaders, in short, will emphasize central features of party policy so as to minimize dissent within various sections of the party and secure their position. Broadly speaking, therefore, the model is driven by policy-based, rather than office-based, preferences. In what follows, the theoretical basis for this argument is elaborated, and then the argument is illustrated through selected case studies of Danish and Norwegian party behaviour.

The Theoretical Framework

The basic perspective guiding this analysis is the presumption that most political conflict situations are to be seen as bargaining situations, that is to say, situations in which the ability of a party to gain its ends is, to a significant degree, dependent upon the choices or decisions made by other parties. The bargaining process is thus shaped by the tactics which parties use to gain their desired ends. In this case, according to Schelling, winning

[...] does not have a strictly competitive meaning; it is not winning relative to one's adversary. It means gaining relative to one's value system; and this may be done by bargaining, by mutual accommodation, and by avoidance of mutually damaging behaviour (Schelling 1960, 4-5).

Viewing conflict behaviour as a bargaining process provides us with an image of the bargaining problems which party elites may face. In multiparty a bargaining problem refers to situations in which there is a need for party elites to reach some settlement in parliament, but, at the same time, they wish to settle on terms favourable to themselves (Sjöblom 1968, 272). The bargaining problem can be divided into two components: (i) the interaction problem within a party, and, (ii) the bargaining power of a party (Maor 1990).

An interaction problem within a party may vary from mere disagreements on the one hand to extreme internal conflicts on the other. For purposes of this article, in which the focus is on the latter extreme, intra-party conflict is taken to signify intra-elite (e.g. within the parliamentary group) as well as elite-follower (e.g. elite contra party members) conflict following an alliance which results in a deterioration of the party's stability and/or cohesion.² Operational criteria, both of which are necessary to designate an internal dispute as an intra-party conflict, include: (i) party elites'

perception of the intra-party strife as an attempt to change their coalitional
behaviour,3 and (ii) patterns of organizational decline, i.e. exit and voice
by party members (Hirschman 1970).4

Bargaining power, in turn, refers to the power of a party to bind itself (Schelling 1960). Self-binding is closely related to the credibility of threat, which depends on how visible the inability of the threatening party to rationalize its way into or out of its commitment may be to the threatened party (Schelling 1960,40). From an intra-party point of view, the credibility of a party's threat is significantly affected by the ability of party elites to cope with internal conflicts. In the parliamentary arena, both voice and exit are costly in terms of a party's bargaining power. The cost of devoting even a modicum of time and resources to resolve internal conflict undermines elites' bargaining power. Elite perceptions of the credibility of threats, moreover, commonly provide an accurate picture of intra- and inter-party dynamics. For operational purposes, therefore, bargaining power may be defined as the power of a party to commit itself to parliamentary over its 'decisive preference' (Luebbert 1986) as perceived by the party and competing elites.⁵

In minority situations, party elites which continuously face conflictual situations will use tactical moves and countermoves in the parliamentary arena in an attempt to resolve the bargaining problem. They will have to commit the party to cooperative relationships in such a way as to ensure party cohesion. This brings out very clearly — as Figure 1 shows - the significance of intra-party conflicts and their consequences with respect to coalitional behaviour.

Figure 1 indicates that the fundamental relationships being considered here are those pertaining to a party's leadership, linkages between the leadership and the party base, and the effect of these considerations on the party's coalitional behaviour. Internal relationships are not only a matter of democracy, but also a matter of manageability. Under these circumstances, party institutionalization, which affects the elites' ability to cope with

internal conflicts, becomes an intervening variable in the relationship
between intra-party conflicts and bargaining power.

Few scholars have considered the nature of the structural context in which party elites operate as a variable in coalitional behaviour. Sven Groennings, a pioneer in this field, has argued that organizational rigidity enhances the effectiveness of the party as a coalition actor:

[...] the more centralised the party structure, the easier it is for the party to remain in the coalition. The a priori hypothesis that a party weakened by factional dispute will find it difficult to formulate a coalition policy leads quickly to the hypothesis that the greater the organised dissensus within a party, the lesser is the tendency to coalesce, even if the dissensus has nothing to do with coalition policy. It should be noted, furthermore, that it is easier for a party with loose central control to coalesce with another party of the same character than one with tight discipline, because a highly centralised party can present a threat to a loosely structured party (Groennings 1968, 454).

A similar conclusion has been reached by Angelo Panebianco, i.e. that
alliances amongst parties inevitably destabilize the less institutionalized
organizations (Panebianco 1988, 219).

A logical problem is created here: why are highly institutionalized parties considered to be effective coalitional actors if they lack the internal fluidity necessary to adjust to dissent among their members? At the outset one might expect that highly institutionalized parties would possess more defences with respect to internal challenges, as their instruments of control over such uncertainties are concentrated in the hands of the party elites. Such parties, however, in fact tend to lack heterogeneous mechanisms for the diffusion of internal dissent such as dissensions in parliament and factions within the party. The absence of these mechanisms does not allow them to handle internal dissatisfaction in a variety of manageable ways. Such parties are, therefore, at a disadvantage when they enter parliamentary

Hence, the highly institutionalized party cannot enter into conflictinducing negotiations with other parties without risking its hold on its own elite and party members. When inter-party negotiations induce internal conflicts, members may be forced to express dissatisfaction outside the party due to the lack of internal network for the diffusion of dissent. Rigid organization, in other words, can lead to party disintegration when such a party enters serious coalition negotiations.

Specifically, a highly institutionalized party limits the margin for dissenting by internal actors. When intra-elite conflicts occur within a highly institutionalized party, as a result of its alliance strategy, resignation may well be the primary means for elite members to express dissent. Similarly, when elite-follower conflicts evolve within a highly institutionalized as a result of its alliance strategy, activists and militants may either leave the party or express their dissatisfaction through extremist actions outside the party (riots and violent demonstrations, for example).

In describing this interplay of 'challenging' intra-party environment and 'adopting' elites, the intent is not to depreciate the nature of the conflict itself. On the contrary, precisely because political parties can be seen to be divided into competing sub-groups (Duverger 1954), the nature of intraparty retains a fundamental role in the elites' ability to cope with it. Top party elites differ in the extent to which they control parliamentary group members and party followers. The fact that parliamentary group members are interested primarily in 'selective incentives' (Olson 1965) emphasizes their dependency on the top party leadership (Panebianco 1988, 27). This, in turn, enhances the elite's ability to control their behaviour. Less influence can be exerted on party followers, however, as they are primarily interested in 'collective incentives' (Panebianco 1988, 26). Consequently, the elites' ability to cope with the dissenting behaviour of parliamentarians is likely to be greater than their ability to cope with the hostility of party followers.

The above discussion gives rise to a number of hypotheses, one of which
is as follows:

1. If intra-party conflicts occur within a highly institutionalized party as a
result of its coalitional behaviour, party elites tend to possess a relatively
low level of bargaining power (i.e. as they are most likely to modify

their coalitional behaviour in order to maintain party stability). In order to comprehend the complexity of bargaining power, however, cases in which internal conflicts do not occur should also be taken into account. Two additional factors are of relevance in this connection. First, intra-party conflicts are most likely to evolve as a result of a formal, rather than an informal alliance, as the former is more binding and visible than the latter. Secondly, a governmental position gives an advantage to the party occupying it in terms of control over policy formation and implementation. a second hypothesis can be formulated as follows:

2. As long as a highly institutionalized party which occupies a governmental position forms informal minority governments, it tends to avoid intraparty and, thus, possesses a relatively high level of bargaining power.

Formal minority governments refer to minority cabinets whose external support is negotiated prior to the cabinet formation through explicit, comprehensive and more than a short-term contract (Strom 1990, 94). As intra-party conflicts seem most likely to evolve following a formal legislative alliance, it is reasonable to extend the above definition by treating a coalition minority government (i.e. when the formal aspect of the alliance exists at the governmental level) as a formal one. Informal minority governmentscan defined as minority governments which lack the abovementionedcharacteristics. generalize, the most attractive strategy in terms of bargaining power for highly institutionalized parties is the formationof

mationofinformal minority governments. However, this is restricted to
polities in which parties are more important as organizations than as labels
(Wilson 1973, 95).

Methodological Considerations

At this stage of the analysis a method of inquiry is needed which is capable of achieving insights into the processes considered here, a method which reaches beyond aesthetic appreciation and approaches acceptable scientific canons of intersubjective validity. The fundamental methodological premise this enquiry is that the best way to study intra-party processes and coalition bargaining is to talk with party elites systematically and to listen carefully (Putnam 1973; Aberbach et al. 1981). Hence, in May 1989 and October 1990, 38 members of Danish and 39 members of Norwegian party elites were interviewed in open interviews lasting somewhat more than half an hour each. The Danish and Norwegian party systems were chosen, among other reasons, because of the existence of major parties - the Danish Social Democratic Party (SD) and the Norwegian Labour Party (DNA) - which have rejected any form of formal parliamentary cooperation in the post-war period. Such a phenomenon has never been examined from an intra-party point of view.

The systematic analysis of these interviews provides the basis for the discussion presented in this article. An analysis too detailed to be presented here demonstrates that members of parliament, ministers, and prime ministers remarkably well the views of the elites within each party.⁶ The strategy of the interviews themselves was to raise specific questions concerning the position and leadership style of the party leader, the intraparty of the party's coalitional behaviour and the relative bargaining power of the party elites. Any one interviewee's perception of intra-party conflicts and attitudes towards coalition partners was confirmed by at least two elite members to be considered valid. All the interviews were tape-recorded with notable advantages for the subsequent analysis and with very minimal effects on the frankness of the respondents.

The preceding discussion of highly institutionalized parties, intra-party politics and coalitional behaviour has been pursued at a rather high level of abstraction. In the remainder of this article the theoretical argument will be illustrated by analysing six case studies which involve strategies of majority-building by highly institutionalized parties competing in minority situations.⁷

Highly Institutionalized Parties which Face Internal Conflicts

The Social Democratic Party in Denmark, 1978-79

The fundamental reason for the formation of the Social Democratic-Liberal Party coalition government in Denmark in 1978 was the desire to form a stable cooperation which could adhere to a long-term economic policy that would bring about economic recovery (Damgaard 1989). The coalition agreement provided for the implementation of an economic stabilization plan designed to tackle adverse economic conditions. The accord reached between the parties involved an understanding that three economic policy proposals strongly supported by the trade unions would not be pursued. These proposals were (1) a tax reform, (2) housing reform, and (3) the introduction of wage-earner co-ownership of industry (i.e. 'economic democratization').

In the months following the coalition agreement, strife between the party and the trade unions was recorded, and conflict emerged within the Social Democratic parliamentary group. At the SD Congress in December 1978, an open dispute took place between Anker Jørgensen, the party leader and Prime Minister, and the trade unions' chairman, Thomas Nielsen. The latter criticized the way the government formation process had taken place and called it a 'coup' (Jørgensen 1990). It was the conflict within the SD parliamentary group, however, which was manifested by an unprecedented expression of dissatisfaction. The leader of the SD parliamentary group, Jens Risgaard Knudsen, resigned during the first meeting of the SD parliamentary subsequent to the conclusion of the SD-Liberal accord.

In an attempt to pacify the intra-elite conflicts, Jørgensen imposed constraints on the day-to-day operation of the coalition government. A coordination committee (i.e. an inner cabinet) was formed and a principle of 'mutual veto' in the governmental decision-making process was established. The coordination committee, which comprised of four Social Democrats and three Liberal ministers, discussed issues before ministers were allowed to publicize their opinions. The principal of 'mutual veto' was operationalized creating a system of 'contact ministers'. SD ministers had to contact ministers from the Liberal Party and obtain approval for proposals which they wished to promote, and vice versa. Because of the continuation of internal dissatisfaction within the SD, however, Jørgensen was ultimately forced to break the SD-Liberal government coalition in order to maintain party stability.

The failure of the highly institutionalized SD to neutralize internal
opposition was translated into an inferior position in the bargaining arena.
This was confirmed by Knud Engaard (Liberal), the Minister of the Interior:

We had double the work and double the influence. We could stop everything that we did not like. That is a problem with a coalition government between two parties of very different principles. If you cannot reach a compromise, then such a government has to stay away from legislation in such areas.⁸

The fact that the Liberal Party blocked the SD's main proposals for 'economic democratization', tax and housing reform, all of which were strongly supported by the trade unions, exemplifies the inferior position which the Social Democrats held in the bargaining plane.

The Centre Party in Norway, 1985-86

The starting-point for the analysis of conflicts within the Centre Party in Norway is the 1985 general election, after which the Conservative-led government coalition lost its majority in parliament. At the outset, a government coalition comprising the Conservative Party (CP), Centre Party (SP) and the Christian People's Party (KrF) was formed in 1983 based on a decisive preference for 'fighting inflation'. Given the adverse economic trends apparent during the early 1980s such as falling competitiveness the non-oil industrial sectors, rising inflation and increasing unemployment, it is hardly surprising that such a preference did not alienate the non-socialist elites and party members. Even so, after the government lost its parliamentary majority, conflicts were recorded within the SP.

The source of intra-elite conflict within the Centre Party was basically an organizational one. Members of the SP parliamentary group refused to serve as a 'rubber stamp' for governmental decisions. Anne Enger Lahnstein, the deputy leader of the SP parliamentary group during 1983-86, the guideline of the group's strategy as follows:

That is the way we behave in my party. I cannot remember another way of behaviour of the group. [As members of parliament] we are elected by the people to take care of special issues. So our people in the government negotiate with the other parties and must find a compromise, and then we have to do something with it inside the parliament. Of course we not only provide votes for the government. We also have to think on our own and we have responsibility for what we are doing. That is my basic attitude to this. For them [SP cabinet members], this was a problem.⁹

The intra-elite conflict prompted determined action from the SP cabinet members. The government had to accept a pattern of two-stage compromise of governmental decisions. First, a compromise was reached at the government then the compromise was renegotiated and amended within the SP and KrF parliamentary groups. As the Prime Minister, Kare Willoch, noted,

In reality, members of government would consult with members of parliament when the
government was discussing controversial issues and sometimes the government postponed
decisions in order that its members might consult with their groups in parliament. What

should have been respected to a greater degree was that members of government should have authority to negotiate compromises which members of parliament should accept. Signs by parliament should be given in advance so that there might be negotiations leading to a result. But what gradually developed was a system through which compromises had to be adjusted, and always in the same direction - that is, always in the direction of greater expenditure.¹⁰

The SP parliamentary group, in short, established itself not merely as a
relevant actor, but as a necessary actor in the governmental decisionmaking

Attempts to neutralize the dissatisfaction within the SP parliamentary group were at first initiated by the Prime Minister. Willoch implemented a twofold strategy: (1) an informal inner cabinet was formed which included the Prime Minister, the Finance Minister and the leaders of the two coalition partners; and, (2) leaders of the parliamentary groups were invited to various informal meetings of the government, such as lunches. As for the personnel aspect of this strategy, Willoch noted that a clear reasoning lay behind the allocation of portfolios. The reasoning was as follows:

I wanted their leading personalities in the government. It was my demand that their party leaders should be in government because I did not want to strengthen the other centres which would be in parliament. That was my absolute condition for having three parties in government."

Indirectly the SP parliamentary group was to some extent neutralized, although it was not totally pacified by the allocation of governmental portfolios to its prominent leaders in governmental positions and the simultaneous incorporation of its non-prominent parliamentary leaders into the informal structure of the government. As a result of the continuation of dissatisfaction within the parliamentary group, the SP cabinet members had to modify their coalition behaviour in order to maintain party stability. This change occurred in April 1986 following the defeat of the government in parliament. Instead of cooperating with the Conservative Party in opposition, Centre Party formed an informal minority government with the Labour Party which lasted up to 1989. The internal opposition of the SP parliamentary group was consequently pacified and, at the same time, the party leader retained his position.

The Centre Party in Norway, 1989-90

The source of the intra-elite conflict within the Centre Party subsequent to the formation of the 1989 formal alliance with the Conservative Party and the Christian People's Party was again an organizational one. Members of the SP parliamentary group refused to serve as a 'rubber stamp' for governmental decisions. Consequently, SP cabinet members were confrontedwith trade-off between government solidarity and parliamentary

group hostility during 1989/1990. According to John Dale, the party secretary:

This is a classical conflict because you have negotiations in the government and you have negotiations in the parliamentary group. [Yet] the question you posed is surprising. You asked what is the parliamentary group allowed to do. This is not a question. The question is what are the cabinet ministers allowed to do because, in our system, the parliamentary group has a 'say'. They are the people that must take responsibility here in parliament by their voting. They [cabinet members] cannot come to the parliamentary group and say 'Listen, we demand from you'. It's more the other way around.¹²

Because of the intensity of the intra-elite conflict, SP cabinet members were unable to embrace policy proposals fully without offending the parliamentary A classic example is the inter-party negotiations over the 1990 budget in which the Centre Party acted inconsistently. The party broke up the negotiations once the Progress Party (FrP) had been allowed to take part in the discussions, but in the end signed the agreement all the same. A further cause for dissatisfaction of the SP parliamentary group was the direct negotiations which were held between the Prime Minister, Jan P. Syse, and the leader of the Progress Party, Carl Hagen.¹³

How did the SP cabinet ministers resolve the conflict? Johan J. Jakobsen, the SP leader and Minister of Local Government and Labour, implemented a two fold strategy. First, a decision-making 'troika' responsible for shaping party strategy was established. This troika included Johan Jakobsen, Anne Enger Lahnstein, the leader of the parliamentary group, and John Dale, the party secretary. Second, the SP broke up the government coalition on the issue of the 'European Economic Area'.

To understand how the modification of the SP alliance strategy was an inevitable consequence of the establishment of the 'troika' requires a more detailed examination. The discussion around the evolution of the SP strategy regarding the 'European Economic Area' began with the fall of Gro Harlem Brundtland's minority Labour Party government in 1989. A newspaper interview with John Dale provided a starting-point. In this interview, Dale expressed the view that Labour's policy toward the European was one of the main reasons why the Centre Party wanted a change in government (Nationen, 6.10.89, 2). SP had tried to start a debate on Norway's and EFTA's adaptation to the single market within the EC during the 1989 electoral campaign, but Gro Harlem Brundtland had not wanted to divulge the government's strategy to the public. Immediately following the election, however, Gro Harlem Brundtland drafted a series of memos on adaptation which substantiated the SP's arguments that the final aim of the Labour Party was to make the question of EC membership a mere formality (Rommetvedt 1990, 21).

The Centre Party was at this point, as earlier, strongly opposed to both
membership of the EC and a full customs union between EC and EFTA

for all products. Even so, a commitment to a 'broader trade agreement' with the EC was a part of the 1989 three-party bourgeois coalition agreementwhich Centre Party accepted (Keesings 1989, 36985). A year after the coalition formation, however, a governmental crisis surfaced over the issue of the formation of a free internal market in Western Europe, 'A European Economic Area', between EFTA and the EC. While SP insisted that Norway should refuse to abandon its concessionary laws which impede foreign ownership of property, financial institutions and industrial enterprisesin the Conservative Party adopted a more conciliatory position.

To understand how the Centre Party's position on this issue was formulated, is necessary to appreciate the role of Lahnstein and Dale in the 'troika'. From its beginning, SP strategy over this issue stemmed from the opinions held by both Dale and Lahnstein. According to Ole Gabriel Ueland, SP member of Storting,

Our tough leaders are represented in the parliament, in the government and in the administration. You have a 'troika' there of Jakobsen, Lahnstein and Dale, they are all in the centre of the discussions [...]. But the person that without any doubt had used most time and has had most interest for this question during the last year was John Dale.¹⁴

As party secretary, Dale could not initiate a governmental crisis without the firm backing of party members, nor could he and Lahnstein act without the approval of the parliamentary group. Yet their political skill enabled them to establish an informal support base within both the party organization the group in parliament. Within the party organization, for example, Dale met SP leaders at the local level and presumably won their support. He did not, however, seek formal approval from the various party bodies.

Discussions in the parliamentary group, in turn, took place approximately two to three weeks before the actual crisis occurred. Once again, the two leaders did not seek the formal approval of the parliamentary group. The 'troika', which was dominated by Lahnstein and Dale, in other words, essentially acted independently with respect to the issue in question.

The failure of Centre Party elites to deal more directly with internal conflicts was translated into a relatively low level of bargaining power during this period. Relatively minor concessions made by the Conservatives were a major source of irritation to other party elites. During negotiations over the 1990 budget, for example, the value added tax for milk was decreased and the amount of money for environmental protection was increased.¹⁵ The relatively low level of the SP bargaining power in this instance was confirmed by John Dale, the Party Secretary:

The Conservatives do not respect our attitudes. They think that this is a small party and
[therefore] they could be the bosses, that we, as a small party, we should only obey them.

It was wrong. Their mistake was that they thought they could more or less force us to obey them instead of respecting us and respecting our programme, our own right to decide our policy. [They] try to use all kinds of force and threats, year by year. [For example, they said] 'if you do not obey us we shall crush you'. This is their big mistake.¹⁶

Not surprisingly, in the crisis over the 'European Economic Area' the SP
parliamentary group militantly opposed its elite, ultimately forcing it to
modify its coalitional behaviour.

Highly Institutionalized Parties which Avoid Internal Conflicts

The Social Democratic Party in Denmark, 1975-78

From 1975 to 1978 the Social Democratic Party in Denmark formed informal minority governments with the Conservative Party (CPP), the Christian People's Party (KrF), the Centre Democrats (CD) and the Radical (RV), as it was unable to build a majority with the parties on its left. A classic example of these legislative alliances was the agreement between SD and its partners over the passage of the 1976 budget, which introduced heavier penalties for breaches of current collective wage agreements a complete wage and price freeze. Although Anker Jørgensen subsequently went back on the first part of this agreement following strong opposition from the central federation of labour (LO), the proposals for wage and price freezes were passed in December of 1975 with the support of the four parties mentioned above (Keesing 1978, 29332-1).

Differences of opinion with regard to SD coalition behaviour were evident during this period. While Jørgensen preferred to cooperate with the Liberal Party, the elite members found it much easier to cooperate with the Conservatives. As Svend Jakobsen recalled:

The Prime Minister, Anker Jørgensen, was more in favour of co-operation with the Liberals than the Conservatives. He saw the Conservatives as the party to the right [..] Some of us were discussing the situation already in the middle of the 70's because in concrete discussions it was much easier to negotiate with the Conservatives than with the Liberals.¹⁷

The importance of the above difference between Anker Jørgensen and some elite members should not be overemphasized. After all, the SD alliance strategy mainly involved informal alliances with the CPP as well as with other centre-right parties. Participation of the Liberal Party (V) in some of the SD legislative alliances during 1977, moreover, was only on an informal, ad hoc basis which did not antagonize centre-right elements within the party elites.

Given the ad hoc nature of informal minority governments, Social Democraticelites
select the least 'expensive' cooperation partner available.

Two examples concerning the inclusion of the Radical Liberals (RV) in some of the 1977 alliances and its exclusion from others spring to mind. On the one hand, cooperation between Social Democrats, Liberals, Radical Liberals and Conservatives, in April, resulted in the adoption of an emergencybill the government's 6 percent norm for annual pay increases for a two-year period (Keesing 1978, 29331). Similarly, on 6 September 1977, the government, supported by these parties, secured parliamentary approval for a package of increases in indirect taxes on petrol, tobacco and spirits (Keesing 1978, 29331). On the other hand, the Radical Liberals were excluded from an agreement among the SD, V, KrF, CPP, and the CD over a four-year extension of the defence budget in March 1977 (Facts on File 1977,428). As these were many feasible partners, SD elites could oppose the demands of the RV and the extreme left-wing parties without risking legislative defeat over this issue. In sum, as long as SD elites formed informal alliances, the party maintained a relatively high level of bargaining power.

The Norwegian Labour Party, 1976-81 and 1986-89

The coalitional behaviour of the Norwegian Labour Party (DNA) during Odvar Nordli's and Gro Harlem Brundtland's terms of office in 1976-81 and 1986-89, respectively, provides a further set of illustrative examples.⁸ During both periods the DNA formed informal minority governments. This strategy reduced potential negative 'influences' stemming from the party's alliance strategy. The internal opposition, especially during Nordli's period, could therefore not rally itself over the issue of alliance strategy since the party committed itself to the least binding inter-party cooperation. This direction is confirmed by a Labour Party MP, Björn Tore Godal.

There has been strong tendency in the Labour Party to avoid any organized co-operation with anybody in a formal agreement. Not since the 1930s has our party gone into a formal agreement with any other party. It's really the basic belief that we are strong enough in our own right.¹⁹

This parliamentary strategy was duly implemented during 1976-81 and 1986-89. For the former period, potential alliance partners were mainly the Centre Party, which lost nine seats (42.8 percent of its share) in the 1977 election, the Socialist Left Party (SV), which lost 14 seats (87.5 percent of its share), and the Liberal Party (V), which maintained its two seats in the Storting. For the latter period, potential alliance partners were mainly the SP, which increased its share of the seats from 11 to 12 in the 1981 and 1985 elections, the Christian People's Party (KrF), which improved its share of the seats from 15 to 16, and the SV, which improved its share of the seats from 15 to 16, during the same elections.

External support from the SP was clearly expected during the redistribution wealth in favour of the rural population which began in 1976. During 1986-89, however, support from the SP, which was usually accompanied by support from the KrF, was perhaps less to be expected, but none the less evident on several occasions. In June 1986, for example, the government's austerity measures were approved by the Storting after an agreement with the SP and KrF (Keesing 1986, 34503). Similarly, the 1987 budget was approved in revised form on December 1986 with the support of SP and KrF (Keesing 1987, 35210). In December 1987 the 1988 budget was again approved with the support of SP and KrF. Additionally, in April 1988 the Storting passed legislation, effectively freezing wages and dividends, which had been proposed by the government and was supported by SP. Finally, in June, 1988, the parliament approved a new law which served to regulate the Oslo stock exchange, which had been proposed by the government and was supported by the SP (Keesing 1988, 36241).

What is most interesting here is the relative bargaining power of the Labour Party during the periods under study. For the former period, the Prime Minister, Odvar Nordli, could summarize the DNA's bargaining power as follows:

You cannot co-operate with different parties without giving something, but I could give those gifts from the programme of the Labour Party. If you find it at the right time, it's all right. So, I must say, in the five years I was prime minister of a minority government, I had no great problems in parliament.²⁰

The dominant policy influence of the DNA was evident in the wave of economic democratization reforms passed in the 19705. These reforms included improvement of working conditions, enhancing industrial demonracy the decentralization of banking, all of which exemplify the relatively level of bargaining power possessed by the DNA during this period (Esping-Andersen 1985, 224).

Minor concessions to alliance partners - SP and KrF - were also evident during 1986-89. SP in particular was, from the perspective of Labour Party elites, considered to be a most cooperative party. As Einar Förde, leader of the DNA parliamentary group, put it:

By far, the easiest party to negotiate with is the Centre Party for a very pure and simple reason; that is, because it is an extreme case [of] a pragmatic party oriented towards their own interests. They will always make a deal on the judgment of what they get. There was very little ideology in it and it was very clearly a case of take-and-give relations.²¹

The Labour Party, in short, could rely on the external support of the Centre Party which was forthcoming after minor modifications of DNA proposals. To sum up, the Labour Party possessed a relatively high level of bargaining power during the period under examination.

Conclusions

The model proposed implies two predictions about the impact of intraparty on the bargaining power of highly institutionalized parties: (1) when intra-party conflicts occur within a highly institutionalized party as a result of its coalitional behaviour, the party elites tend to possess a relatively low level of bargaining power; and, (2) as long as a highly institutionalized party which occupies a governmental position forms an informal minority government, it tends to avoid intra-party conflicts and, thus, possesses a relatively high level of bargaining power. The derived consequence - in line with Strom's (1990) theory on majority-building by minority governments - is clear: In multi-party systems in which minority situations occur, the most attractive strategy (i.e. in terms of the relative bargaining power) for highly institutionalized parties is the formation of informal minority governments.

The evidence examined here proves consistent with these predictions. All the cases which involved intra-party conflicts within highly institutionalized parties occurring as a result of their coalitional behaviour, underline the inability of their elites to cope with internal dissatisfaction. As a result, party elites insisted on preferences which were sufficiently focused so as to generate the widest possible support within the party and hence imposed structural constraints on the day-to-day operation of the government. Yet, the lack of mechanisms for the expression of internal dissent meant that eventually party elites had to change their parliamentary strategy in order to maintain party stability.

This conclusion raises an inevitable question: why do some highly institutionalized form formal alliances? A substantial reason for implementing such a strategy would appear, first, to lie in the desire of highly institutionalized anti-system or protest parties to obtain legitimacy through a formal alliance with a governmental party - for example, the Italian Community Party, 1976-79. Secondly, a highly institutionalized party which occupies a governmental position might be willing to commit itself to formal cooperation in order to 'share the burden' of economic crisis - for example, the Danish Social Democratic Party, 1978-79.

Thirdly, when a highly institutionalized party provides informal external support for a government on a roughly permanent basis, a demand for 'office' can be effectively raised by its elites and party members. The basis for a demand to participate in government is both ideological and organizational. Ideologically, if common understanding over main policies could be achieved outside government, it naturally raised the prospect of close cooperation in government. From an organizational point of view, a political party might also undermine its base (i.e. membership) by providing long-term external support without distributing the expected 'selective

incentives' which formal participation in Government may have to offer (Olson 1965). A classic example for such a case is the demand raised by the Centre Party and Christian People's Party elites for their inclusion in the Conservative-led 1983 minority government in Norway.²²

The research reported here is but an illustration of what can be achieved if an organizational approach to analysing party behaviour is adopted. It reveals that elites possess a higher-than-expected ability to neutralize intraparty It also indicates some fascinating variation in the evolution of intra-party conflicts attributable to the existence or the lack of channels of activity for internal opposition. Above all, it offers an example for the kind of systematic cross-nation research that may shed further light on processes of intra-party politics and their impact on coalitional behaviour.

---

NOTES

1. For recent attempts to break free from an office-seeking frame of reference in the context of intra-party politics, see Laver & Shepsle 1990a, 1990b.

2. The degree of stability is related to horizontal exchanges (elite-elite exchanges) and, in particular, to the character of compromises (whether stable or precarious) at the organizations' upper echelons. The degree of cohesion, on the other hand, is based upon the extent to which vertical exchanges (elite-follower exchanges) are concentrated the hands of the few, or are dispersed amongst numerous leaders (Panebianco 1988, 39).

3. No matter what the extent of the organizational decline, exit and voice could fail to alert the elites if they are not perceived by them as an attempt to change their strategy. One possible situation in which this seemingly quite unlikely event would come to pass is when a decline in party membership hits most parties in a party system simultaneously. In this situation each party might well garner in some of the disgruntled members of the other parties while losing some of its previous members to its competitors. In these circumstances the exit option (i.e. the decline in party membership) ineffective in alerting the party elite to its failings. Clearly, therefore, elites must perceive internal conflict as an attempt to change their strategy, and identify patterns of organizational decline related to their party.

4. We have at our disposal at least seven indicators of organizational decline. Intra-elite conflicts, for example, may be manifested by resignation of elite members (i.e. the 'exit' option) and/or by dissensions in parliament, petitions, and appeals to party elites with the intention of forcing a change in party strategy (i.e. the 'voice' option). Elitefollower by comparison, may be manifested by a decline in party elites with the intention of forcing a change in party strategy (i.e. the 'voice' option).

5. According to Bacharach & Lawler (1981), three assumptions are in order with respect to bargaining power. First, power is the essence of bargaining, i.e. bargaining power pervades all aspects of bargaining and is the key to an integrative analysis of context, process and outcome. Second, bargaining is a process of tactical action, i.e. the focus is on the use and effectiveness of specific tactics. Third, bargaining power is subjective power, i.e. perceptions are a crucial mean in the relationship among parties.

6. The analysis is contained in unpublished material written by the author.

7. Considerations concerning classification of the parties under examination according to degrees of institutionalization are also contained in unpublished material written by the author.

8. Knud Engaard, interview with the author, Copenhagen, May 1989.

9. Anne Enger Lahnstein, interview with the author, Oslo, October 1990.

10. Kare Willoch, interview with the author, Oslo, October 1990.

---

11. Kåre Willoch, interview with the author, Oslo, October 1990.

12. John Dale, interview with the author, Oslo, October 1990.

13. Carl Hagen, interview with the author, Oslo, October 1990.

14. Ole Gabriel Ueland, interview with the author, Oslo, October 1990.

15. Per Ditlev Simonsen, interview with the author, Oslo, October 1990.

16. John Dale, interview with the author, Copenhagen, May 1989.

17. Svend Jakobsen, interview with the author, Copenhagen, May 1989.

18. The faet that Harlem Brundtland enjoyed a pre-eminent position in the party during her term of office, whereas Nordli's position was to a significant degree undermined by a personal strife with the party chairman, is not of relevance for the analysis presented here.

19. Bjorn Tore Godal, interview with the author, Oslo, October 1990.

20. Odvar Nordli, interview with the author, Oslo, October 1990.

21. Einar Forde, interview with the author, Oslo, October 1990.

22. Kåre Willoch, interview with the author, Oslo, October 1990.

REFERENCES

Aberbach, J. D., Putnam, R. D. & Rockman, B. A. 1981. Bureaucrats and Politicians in
Western Democracies. Cambridge, Mass: Harvard University Press.

Bacharach, S. B. & Lawler, E. J. 1981. Bargaining: Power, Tactics and Outcomes. San
Francisco: Jossey-Bass.

Damgaard, E. 1989. 'Crisis Politics in Denmark 1974-1987', in Damgaard, E., Gerlich, P. &
Richardson J. J., eds., The Politics of Economic Crisis: Lessons from Western Europe.
Aldershot: Avebury.

Duverger, M. 1954. Political Parties. London: Methuen.

Esping-Andersen, G. 1985. Politics against Markets. Princeton, NJ: Princeton University
Press.

Facts on File Yearbook

Groennings, S. 1968. 'Notes Toward Theories of Coalition Behavior in Multiparty Systems: Formation and Maintenance', in Groennings, S., Kelley, E. W. & Leisersen, M., eds., The Study of Coalition Behavior. Theoretical Perspectives and Cases from Four Continents. New York: Holt, Rinehart and Winston.

Hirschman, A. O. 1970.£jh7, Voice and Loyalty: Responses to Decline in Firms, Organizations,
and States. Cambridge, Mass.: Harvard University Press.

Jørgensen, A. 1990. Braendingen: Fra mine dagboger 1978-1982. Copenhagen: Fremad.

Laver, M. & Shepsle, K. A. 1990a. 'Government Coalitions and Intra-Party Polities', British
Journal of Political Science 4, 489-508.

Laver, M. & Shepsle, K. A. 1990b. 'Coalitions and Cabinet Government', American Political
Science Review 84, 873-890.

Luebbert, G. M. 1986. Comparative Democracy Policymaking and Governing Coalitions in
Europe and Israel. New York: Columbia University Press.

Maor, M. 1990. 'The Dynamics of Minority Rule: A Bargaining-Based Theoretical Framework'.
presented at the ECPR Joint Session of Workshops, Bochum, Germany.
Olson, M. 1965. The Logic of Collective Action. Cambridge, Mass: Harvard University Press.

Panebianco, A. 1988. Political Parties: Organization and Power. Cambridge: Cambridge
University Press.

Putnam, R. D. 1973. The Belief of Politicians: Ideology, Conflict and Democracy in Briain
and Italy. New Haven: Yale University Press.

Rommetvedt, H. 1990. 'Parties and Governments in Norway'. Paper presented at the ECPR
Joint Sessions of Workshops, Bochum, Germany.

Schelling, T. C 1960. The Strategy of Conflict. Cambridge, Mass.: Harvard University Press.

Strom, K. 1990. Minority Government and Majority Rule. Cambridge: Cambridge University
Press.

Wilson, J. Q. 1973. Political Organizations. New York: Basic Books.