Resumé

Much current research into the working of liberal democracy tends to emphasise and concentrate on the corporate channel of influence. The linkage to centres of power provided by political parties has not received sufficient attention. There is a need for more research into the operation of parties as linkages between grass root activists and national elite politics. Studying party democracy is to study internal distribution of power, and the article turns to the debate on community power studies for guidelines on how to approach empirical research on intra-party democracy.

Introduction

How is power distributed in a political party? Is it a tyranny of the mass membership against the enlightened opinions of the few? Is it an oligarchy suppressing the democratic will of the membership? Are the back-room boys the secret powerholders of an allegedly democratic party system? The questions are abundant — in the literature as well as in the press. The number of times these questions are raised, however, is clearly out of proportion to the attempts made at offering an answer; possibly this is just another derivative of the reverence shown within the political science profession when approaching the study of power directly.

Although the purpose of this paper ultimately is constructive, the problems underlying studies of party power will permeate much of the following. The plan here will be to explore some basic themes and draw some distinctions which might be useful to the study of internal party power. But if research is characterised by one "open" phase of seeking the proper guidelines for an empirical study, and another rather "closed" phase of binding oneself to the mast in a systematic search on the basis of some kind of "conscious intolerance", this is a phase-one article. It is an exploratory journey into parts of the literature in order to work out the broad characteristics of the terrain

— a knowledge which hopefully will make it easier to point out a promising
route for a later and more thorough expedition.

---

* Prepared for the ECPR workshop on party organisations, Freiburg, March 1983.

The point of departure for this paper is a recognition of the basic similarities of "power studies" — whether on the international arena, at the state level, in the community, in an interest organisation, or in a political party. The assumption is that in the open-minded search for empirical entries, making the party debate a bandwagon of the general power debate will possibly bring forward some fresh perspectives to the debate on intra-party power.

What I propose to do, then, is first briefly to recapitulate some characteristics of the tradition of party studies, and second to sketch a few basic parameters of the general power-debate. This will provide the setting for the presentation of some approaches to intra-party power. In the following section I will survey some recent trends in party research as well as actual party change — with an emphasis on Norwegian parties and Norwegian party research. Finally, I will tie some of the threads together by suggesting possible entries for an empirical study of intra-party power.

Before turning to our subject matter, however, some comments on the relevance of an intra-party focus to the issue of national-level political processes are needed. What is the relationship between intra-party decisionmaking and decision-making at the national level? At the national level, one line of argument goes, intra-party democracy is not crucial. Whether the political party is internally democratic is of no more importance than democracy in the Football League when it comes to evaluating democracy at the national level. According to Schumpeter, there are other characteristics of the political system which really are decisive in this respect: "The democratic method is that institutional arrangement for arriving at political decisions in which individuals acquire power to decide by means of a competitive struggle for the people's vote" (Schumpeter 1942, 269). The point here is that political choice at the polls is more important than ability to further one's cause within a political party. In this breach with the Rousseau-inspired tradition within democratic political theory, it is the method which counts, not primarily the question of a harmony of content between majority opinions and official policies. But of course a Downsian type of argument enhances the thesis that the competitive struggle — the very method of selecting the political elite — will also shake the elite opinions in line with the majority opinions of the electorate. That is in the very interest of the elites themselves (Downs 1957).

Is on the contrary internal party democracy counter-productive to national democracy? Taking national democracy to mean majority electoral support for the decisions made within national political institutions (which of course leaves a great deal unsaid), this may require a swift-moving, elastic national party leadership able and willing to make quick manoeuvres to respond to changing electoral opinions and competing party platforms. In this Schumpeter-Downs "battlefield model" of party politics, the troops must obey the

orders of their generals; internal decision-making costs are comparative
disadvantages in the game.

Let us try a rough sorting out of the issues involved, which for the purpose of this article will suffice. We can distinguish between two levels — the party and the nation — and two ways of reaching decisions — the democratic and the non-democratic. Putting these together in Table 1, we get four different kinds of political systems with different voter-party member relationships:

Clearly, we cannot assume a priori either that internal party democracy will bring national democracy (the "party member democracy" category), or that non-democratic internal party processes necessarily lead to a non-democratic national scene (the "electoral democracy" category). The relationship between the two levels cannot be dealt with thoroughly here, but obviously the nature of internal party processes has consequences for how national politics operates. And at the end of the day, I even think the issue has considerable importance for the question of national democracy, taking into account the ability of parties to shape electoral opinions, their agenda-setting abilities and the overall inertia provided by the parties in the national political processes.

Genealogies of Party Research

In his discussion of political parties, Giovanni Sartori draws our attention to parties as instruments "for representing the people by expressing their demands" (Sartori 1976, 27). The representative function of parties points to the party as a democratic unit, while the expressive function emphasises their major task as organisations. This dual purpose of parties is reflected in the literature by two main lines of inquiry — that concentrating on intra-party democracy and that which Sartori labels "the organisational approach" (1976, 70). Of course one may at first glance question this distinction. In a formal sense it is certainly not satisfactory in that studies concerned with party democracy clearly also must be concerned with organisational structures;

Sartori fully acknowledges this (1976, 70). Nevertheless, as a first screening
of the literature, the distinction is useful.

1. As for the question of democracy, party research is still marked by the original fears of friend and foe alike at the time of the entry of mass political parties into political life around the turn of the century. On the one hand there was a Burkean fear that the party organisation would place restrictions on the free deliberations of the elected representatives. It was bad enough that the voters might try to impose their will on the parliamentarians, but it was assumed to be even more of a problem when sections of the electorate were to sustain their pressure by organisational means. The contenders of this view were not only doubtful about the prospects of party democracy, but also about the effects of extra-parliamentarian politics. Moisei Ostrogorski's classic Democracy and the Organisation of Political Parties from 1902 reflects this focus in its defence of individualism against machine politics (Ostrogorski 1902, Lipset 1968, Epstein 1967).

On the other hand, there was a fear that the party as an instrument of expression might not prove sufficiently efficient, that the party might evolve into an apparatus for domination rather than for grass root democracy. Or the discovery that the political party was possibly not the missing link in a Rousseau-inspired image of democracy. This of course is the argument of Michels in Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracy (1911). Here the ability of the party apparatus to channel upwards is questioned and the tendencies strengthening the power of the party leadership emphasised (Michels 1962). But whether the concern is for too much or too little democracy within the political party, there is a continuous tradition within the party literature occupied with the balance between the upward and the downward flow of authoritative decisions.

2. While the party democracy project is parallel to the rise of the mass parties under universal suffrage, the organisational approach is of more recent origin. The sine qua non here is to view the party as an organisation with a particular task which can be analysed with the tools of general organisational analysis. The focus is on the organisational processes as such and the causes and consequences of these. Maurice Duverger's study from 1951, Political Parties, subtitled "Their organization and activity in the modern state", is of course the classic study within this field. Duverger's dictum is that "presentday parties are distinguished far less by their programme or the class of their members than by the nature of their organisation" (Duverger 1964, xv). Since the publication of Duverger's book, studying the organisational ramifications of parties no doubt has been the mainstream tradition within the party literature. And this organisational approach has of course been under the influence of the changing winds within organisational theory. There have been pleas for the importance of "incentive systems" (Clark & Wilson, Wilson

1962) in determining how parties work. Emphasis has been put on the maximisation of self-interest within parties seen as "office-seeking organisations", with the ambition of the candidates in a pivotal role (Schlesinger 1965). The more general interest in functionalism and systems theory is echoed in works from Eldersveld to Sartori.

This organisational approach to political parties probably puts the field at a more realistic level in terms of what can be achieved in empirical studies, particularly in contrast to the Michels tradition of oligarchy studies. It also has the merit of exploring categories potentially useful for comparative research. Still, it seems to have "organised" the main interest of the field away from the study of power distribution. As already indicated, this is certainly not inherent in the organisational approach as such. Power is clearly embedded in the organisational structures and processes involved, and this is of course an old theme within organisational research. Nevertheless the shift in focus — away from the power studies towards the more manageable questions of how parties work as organisations — has restricted the field deplorably , and left the arena to the work of the political journalists.

Power

This academic caution is obviously related to the elusiveness of the concept of power: Can one provide "scientific proof for the production of intended effects? No wonder Robert Dahl in his article on "power" in the International Encyclopedia of the Social Sciences emphasised the difficulties of definition and measurement by pointing to the similarities between the analysis of power and the analysis of causality (Dahl 1968). These difficulties indicate a dilemma which certainly is just as much to the point in party research, namely between rigour and relevance:

"Attempts to meet high standards of logical rigor or empirical verification have produced some intriguing experiments and a good deal of effort to clarify concepts and logical relationships but not rounded and well-verified explanations of complex political systems, in the real world. Conversely, attempts to arrive at a better understanding of the more concrete phenomena of political life and institutions often sacrifice a good deal in rigor of logic and verification in order to provide more useful and reliable guides to the real world" (Dahl 1968, 410).

The question then is whether anything but increasing scepticism and paralysing dilemmas will turn up by surveying the field of "power studies" for guidance in research on intra-party power. On the other hand one can raise the — to my mind — more proper question, which is whether we can avoid the encounter. Certainly the parallels can be ignored or just tacitly acknowledged, but that would not further the study of internal party power. Consider for a moment:

some introductory lines in Samuel Eldersveld's study of parties in the Detroit
metropolitan area, where he stresses the party as an "social organism" as
well as a "miniature political system":

"It has an authority structure, although the manner in which authority is graded and legitimated may differ considerably from other social groups. The party has distinctive patterns of power distribution. It has a representative process, an electoral system, and subprocess for recruiting leaders, defining goals, and resolving internal system conflicts. Above all, the party is a decision-making system, although, again, how "authoritative" the decision-making process is remains a subject of inquiry" (Eldersveld 1964, 1).

Clearly, Eldersveld is building up towards his particular brand of functionalist approach here. But that aside, a power study is a power study whatever the ramifications. Of course one can quarrel a lot about the differences in approach imposed by differences in the level of inquiry, but I still think it is fair to emphasise the family resemblance.

In what sense can this genealogical point of view direct our search for a potentially helpful focus on the study of internal party power? By drawing some rough lines from the general power-study debate, we can compare these with what has actually been done within party power research. First to the main theatre of power studies:

During the 1960's there emerged two antagonistic positions within the field, known as the "pluralists" and the "elitists". The pluralists maintained that western democracies were governed by way of a free competition among a plurality of interests: parties, organisations, pressure groups. Through their competition for support as well as for influence, the political systems came close to the classical notion of democracy — a "polyarchy" after the major protagonist of the camp (Dahl 1971). The elitists on the other hand did not deny the group competition within the system, but maintained that the various elites were nothing but different sections of one unified power elite, a homogeneous stratum divided only by less significant issues (Mills 1956).

From the point of view adopted here, however, the most interesting aspect of these camps was how their maps were worked out, how one went about measuring "the actual distribution of power", and not so much their later labelling. Let us for the sake of simplicity point to four major methodological "schools": The positional, the reputational, the decision-making approach and the focus on non-decision-making.¹ The positional approach applied to the study of party power would imply a rather formal analysis of the kind so abundant in the prebehavioural epoch. The questions here are what does the "constitution" say about the distribution of power. Which bodies are assigned the power to adopt authoritative decisions and who occupies these positions? Where do they come from and what are their background?

The reputational approach means asking people with particular knowledge,

observers or participants, to name the most influential persons within the
bodies or processes in question. In principle this is the worktool of the journalist,
although rarely used in a systematic fashion in the press.

To focus on the decision-making process means neither the exclusive focus on formal structures nor its elite personnel, but to connect both to particular issues working their way towards an official decision. How and by whom is influence exercised in a "fair" selection of major decisions? And across the different policy areas, is power cumulative or dispersed?

Lastly, there is the non-decision-making approach. How and by whom are important issues let in or denied access to the arena of decision-making at all? In this approach the decision-making process is nothing but one stage in a process which starts with the question of which "issues" are allowed to become "an issue". Or in other words, the politics of agenda-setting. And with these four general models of "power measurement" in mind, we shall turn to the more recent literature on internal party power.

Empirical Approaches to Intra-Party Power

As argued earlier, the organisational approach in party studies has made the original and most classic question in party research rather secondary. "Distribution of power" has been approached in a rather circumscribed way — as a structural derivative of organisational types (see, e.g., Wright 1971).. But of course, the literature is far from devoid of studies having a reasonably direct relevance for the study of party power. To mention just one, a very important book in the classic tradition from the mid-fifties was R.T. McKenzie's British Political Parties, which contained the following bold opening sentence: "The scope of this book is indicated by its sub-title: its purpose is to examine the distribution of power within the two major British political parties" (McKenzie 1955, 1963). My task here, however, is not to survey the complete list of such studies, or to evaluate their respective merits and suggestive powers for that matter, but to attempt a systematic categorisation of their basic approach to party power. And the basis for that categorisation is provided by the literature on power studies in general as suggested above.

Positional Studies

In an extreme form this is the approach adhered to by the press spokesmen of the parties themselves: "The decision was taken according to the statute book",² which of course would make most studies superfluous. Very few, though, and certainly not the party activists themselves, believe that parties are like open books. And even if the proper bodies or persons take the decisions within their prerogative, the question remains as to how they reached

their conclusion: who were the mighty and who were not behind the closed doors of the committees in charge. Still, it is fairly evident that the formal structures of parties no more than the constitutional arrangements of states will give us the final story on influence and power in the system. On the other hand, it would certainly not be wise to disregard these factors altogether. And taking this less formalistic point of departure, there are clearly approaches of interest which need not at all be blind to informal structures.

Biographical. Biographies of political leaders (as well as autobiographies) will most often take the positional point of departure: These were his or her positions — in what way and to what extent did he influence his party? Was he a figurehead alone? Or did he rule the party with an iron fist? I shall not go into examples of this approach — for one thing because there are so few academic biographies of Norwegian politicians.3

Organisational practices. Positional research is preoccupied with the power embedded in organisational structures. What gives the leader his power to decide is his role as defined by party statutes. Not only the leader, however, has a place in the statutes — there is the national conference as well as the more exclusive national committees and the party secretariat. They are all involved in the formal no less than the real process of policy formation. One way to approach the question of power is therefore to ask how the organisational structures and rules operate in practice. What is the relationship between form and content, between intention and effects? How close is the fit? Ulf Torgersen has undertaken a study along these lines, focusing on the position of the national conference in Norwegian parties up till 1940 (Torgersen 1969 and 1967). We shall return to this, however, as it is also an example of another kind of approach.

R.T. McKenzie's study of British parties centres on the Leader of the party, noting at the outset regarding the Conservative Leader that he is "formally accorded what seems at first to be almost overwhelming power and authority" (McKenzie 1963, 17). Studying the role of the Leader, his road to power and the relationship between the Leader and other party bodies, would, in McKenzie's view, reveal more about the true distribution of power within the party.

To the degree that positions give power, it is clearly of great importance who fills these positions. And McKenzie's emphasis on the Leader's ascendance to power amply demonstrates this. Approached more broadly, this is of course the rationale behind the interest in the process of elite recruitment. The argument here is that party power may reveal more of its true face in the structural characteristics of the party elite. The degree of trade union background within the social democratic leadership may illustrate the point. This is part of the

background for an earlier study of mine on the recruitment of the Norwegian
Labour Party elite before 1940 (Heidar 1974).

Reputational Studies

To my knowledge, there does not exist any systematic study of how people
evaluate the power of major party figures. As I will come back to, this is
perhaps an unjustified lacuna in the research on intra-party power.

Decision-Making Studies

Together with the studies that I (in rather broad terms) characterised as "positional", the decision-making approach is that most frequently found in the literature. And no wonder: Decision is obviously at the heart of power studies. What this approach in particular brings into focus, however, are the issues. The issue, the cause, is clearly a fairly basic component of the political process. And power is a question of fighting successfully for causes, for making them part of the official party platform.

Case-studies. Often these are not bound by the borders of a single party. They follow the issue across parties, interest groups, etc. toward the final political outcome in Parliament or in the government. Nevertheless, there are many case-studies of internal party decision-making — in Norway though, these are mostly carried out by historians. Many are indeed exciting presentations of the party policy-processes in operation. Examples can be the studies of the Labour Party's policies toward Franco's Spain in the immediate post II period and the decision to join the NATO-alliance in 1948-49 (Benum 1969, Eriksen 1972). The problem of course, as it also was encountered by Robert Dahl in his New Haven study (Dahl 1961) — is whether the case-studies alone or even several cases taken together really are indicative of the "true" distribution of power within the party.

Systematic summaries of case-studies. To smoothen the problem just mentioned, there has been suggested a more systematic approach to the study of decision-making. Ottar Hellevik has proposed an empirical operationalisation of Michels' oligarchy-thesis by conducting a series of case-studies on issues where the party experienced a major division of interests (Hellevik, n.d.). The point was to map the decision-making process — the positions of the main actors as these developed over time, the determinants of this development and the organisational criteria for the final "party decision" — in every case, and then to "summarise" them in a measurement of degrees of oligarchy.

Focus on the party conference. The national conference has in many parties

the final authority in matters of party policy. And at the least the conference is the main intersection between the party activists — or the grass root leaders if one likes — and the party elite, parliamentarian as well as organisational. A recent book taking the pivotal position of the conference in the policymaking process as its point of departure is Lewis Minkin's massive study on the conference of the British Labour Party (Minkin 1978). Another is the study already mentioned on the pre-war parties in Norway by Ulf Torgersen (1969).

The nomination process. A fourth way to search for the distribution of power within parties can be to survey the recurring decision-making on the party's electoral personnel (rather than to focus on its policies). These are decisions on what kind of personnel the party wishes to be associated with among the electorate. Here is if course an opportunity to conduct systematic research over time on the kind of "interest profile" emerging in the party. There are several studies of this brand covering Norwegian parties by Henry Valen and Tom Christensen (Valen 1966, Christensen 1976).

Non-Decision-Making Studies

Like the reputational studies, this is a rather slim category within the party power literature. Non-occurring events, it has been argued, do not readily invite empirical research. The point is, however, that these events do occur; they are only preceding the most visible part of the process (Bachrach & Baratz 1970). And like Bachrach and Baratz on community power, Lewis Minkin has addressed the problem within party politics (Minkin 1978). The topic obviously involve difficulties and hard work, but the politics of agendasetting certainly is both an interesting and rewarding area. No equivalent study exists, however, on Norwegian parties.⁴

The Present Context of Party Power Studies

What are the proper functions of a political party? How will it best serve democracy? And what is the relationship between its ideal and real performance? The questions are as we know many, and most of them have continued to be pertinent since the days of Ostrogorski. In this section I shall concentrate on the question of trends in party development. Is there a convergence towards a particular type of party, a "type" which has implications for the kind of power distribution particularly to be watched out for? Which again may have consequences for the kind of empirical approach to be adopted?

Is there a model of the future party? A search in the literature will reveal
some curious zigzags. At the turn of the century, Ostrogorski thought that

the United States as the first mass-suffrage democracy (which produced the mass party) would also be the first to develop the prototype of a "modern" party. The "machine politics" he observed in American parties was to him a scaring insight into the future. Duverger, writing half a century later, pointed, however, in a rather different direction. In his view, the future did not reside with what he considered the rather loosely knit cadre organisation of the U.S. parties, which were to him closer to the old-fashioned conservative and liberal parties in Europe. The future lay with the mass-membership parties — the "parties of integration" to use Neumann's term (Neumann 1956) — which were typical of the West European socialist movements. Too much ink has, however, been spilt on the question of whether North America or Western Europe can give us the best signposts for the future. Epstein, in contesting Duverger's European mass-party bias, argued that the different conditions provided by these systems in fact had to produce differences in party types, and that placing them within a time-table was not particularly useful (Epstein 1967, 129). Nevertheless, he could not quite resist the temptation himself, since later in his book he denoted the "business-oriented middleclass parties as the pioneers of the new style of campaigning via the mass media" (Epstein 1967, 257).

These are some of the prescriptions then. Do we now know anything more about the trends in party development? What has actually happened to "the political party" during the 35 years which have passed since Duverger described the "contagion from the left", the rise of the "community type" mass party? It would probably be overdramatising the state of the patient to take Walter Dean Bumham's diagnosis of American politics in the 1970'5, "Beyond Parties", as a general point of reference (Burnham 1975). Nevertheless I think some notion of "decline" if not "dissolution" is appropriate. What is Burnham's argument? Basically that the American electorate "now" (mid-seventies) is experiencing a "most sweeping transformation", undergoing a "critical realignment". And this realignment, "instead of being channeled through political parties as in the past, is cutting across older partisan linkages between rulers and ruled. The consequence is an astonishingly rapid dissolution of the political party as an effective intervenor between the voter and the objects of his vote at the polls" (1975, 308). This concerns U.S. politics, and Burnham bases his argument on electoral data, opinion polls,, etc. to reach his bleak verdict on the state of the parties.⁵

But what about European parties in general and European mass parties in particular? Here data on party identification and other measurements of the parties standing in the electorate are rather scarce (Kaase 1980). But these data would in any case not have told us very much about the shape of the party organisation. Nevertheless, there are plenty of scattered reports that the once so proud parties of integration have been changing — if not in the

direction of advertising agencies, then at least bridging the gap to the liberal/
conservative associational type. Are there any hard facts on Norwegian parties
to bring out these points?

The study of political parties was not included in the Norwegian "study of power".⁶ This was by Norwegian standards a large social science project, officially commissioned to unveil the real distribution of power in the Norwegian society. Nevertheless, the findings of this project have obvious implications for party research. One main conclusion of the power study was that Parliament had declined in importance compared to the civil service, that the corporate channel of influence had been strengthened, while the numerical had declined (to stick to terms familiar within Norwegian political science, Rokkan 1966). And as the civil service is besieged by the interest groups, Parliament is the construction site of the parties. This of course would imply that parties — along with Parliament — have declined in importance and power. If so, would not declining party power together with the continued decline of a "gemeinschaft" type society, deprive the mass parties of the very bases of their vitality? There are few figures available to pursue this argument. Looking at the data on party membership, the picture is mixed — some parties have rising figures, others are in decline. Still, one can hardly fail to notice the declining grass root activity within Norwegian parties, with the possible exception of the Conservative party. There are also signals of a decreasing importance attached to the national party conferences.

Does any of this, then, provide more clues for a study of intra-party power? To the degree that the political parties are losing out in the national power game, and to the degree this is acknowledged in the political milieu — can one expect less of a power-struggle within the parties? The history of small and rather powerless parties does not at all concur with a thesis of appeasement. Nevertheless, there may be a change of groups contending for power under the new circumstances. If interest groups take better care of their policies on their own, or, at the least, see a higher "return" on their efforts somewhere else (in the civil service), they might possibly reduce their activities in political parties. And if so, major changes in intra-party distribution of power may occur.

Some Lessons for a Study of Intra-Party Power

Although hardly a convergence toward a "common type", there are some signs that the former parties of mass integration are changing to become less membership oriented. But this is clearly a slow and long-term change, where the traditions of the past will continue to make their impact on the policymaking processes. At the same time, a party like the Norwegian Conservatives with its extraordinary success the last 10 years,⁷ makes the "cadre" type less

applicable, although the party has clearly not become a party of integration. This I think suggests that reading party trends along a uniform scale of development, as a realisation of a "modernisation syndrome", is not particularly useful.⁸

As stated at the outset, the purpose of this paper has been to survey sections of the party literature in order to do some spadework for an empirical study of party power. I shall end, therefore, with some provisional remarks on how this literature perhaps usefully can be employed. Let me again, however, repeat that the question at hand is the distribution of intra-party power. Ultimately that means he who is able to influence the organisational workout of official party policy platforms. What later happens to party policies — in parliament or government for that matter — is outside the restricted scope of interest here. But this of course may well imply that the "party power" pictured will be different from the power-holders of the "final" round. (But then, "where" do we find the "final" outcome of a political process working its way through the different stages of decision-making — to parliament, to government, to civil service, to ... Where is "actually" an "authoritative" decision situated?)

Firstly let me note the close resemblance between the "elitists" of the general power debate and the proponents of the "iron law of oligarchy". The value of both the general and the party specific elitists, it seems to me, is to broaden the scope of research to also cover the less visible corners of the power process: A refusal to take the public debates alone as objects of inquiry, to look for issues not opened for debate, an emphasis on the organisational handling of important debates, a search for dimensions of elite homogeneity on issues with doubtful member support.

What then are the main lessons of this search through the literature, lessons supporting an approach of relevance as well as parsimony? First of all, that the difficulties are alarming for any kind of power study. And in this morass one might possibly start by enlisting the help of the comparative method: The "more than'V'less than" terminology is clearly less problematic to handle than the absolute denotations. Debates on "democracy" vs. "oligarchy" in parties suggest — at least to me — that comparing the structure of power in party A with that of party B may be a better way to start the discussion. I think it would be harder to reach a verdict on the question of the Leader's power within The Norwegian Labour Party than on the question of her power compared to the counterpart within the conservative party.

Looking at the positional approach, it is close to a truism that this provides an unavoidable point of departure. There is more to it than that, however. Institutional values, statutes, organisational rules and practices are "hardened power", the left-overs of yesterday's power struggles. Comparisons of "the

natural way of doing things" may therefore broaden the scope of an otherwise
predominantly pluralist-bent school of decision-making studies.

It seems to me that reputational studies have suffered too much from undeserved disrepute. No doubt there are serious problems embedded in this approach: On what bases is a shortlist of alleged power-holders obtained? How are the respondents selected? On what bases do these actually rate the power of those on the list — do we in fact obtain an "index of rumors"? Still, knowing as we do now that no ultimate answers are readily available through other means, I believe a systematic use of this journalistic worktool may throw at least some additional light on the structure of power. It would, I think, be of more than passing interest to see the Labour national conference delegates' rating of their chairman versus the national trade union leader, and to compare this to the standing of major interest group figures in other parties.

As for the study of decision-making in the minute, historical way, this is of course an extremely time-consuming approach. And there is the question of the validity of generalising single-issue, or even select-issues studies. The intensive, detailed scrutiny has of course major advantages in terms of reliability and relevance, but in a multi-party study I fear this approach would be far too demanding. An easier way out, retaining some of the advantages, would be to ask participants at different levels of the party about their experience of decision-making within the party — the policy-process in general as well as particular issues. Again we will face the problem on what grounds the respondents base their judgements. But again the counter would be that opinions of the participants themselves clearly are far from irrelevant.

A mapping of what actually happened at the party conferences would certainly also be of great interest. The question is what kind of return we might expect on an investment like that. I fear it may be low considering the work necessary for this approach. In terms of Norwegian parties at least, there appears to be a declining importance attached to the party conferences — possibly related to the opening of the conference to the press. Still, it would be premature to write off the conference as an arena of real influence at the outset, even though the preparations before them have been of far greater interest lately than the events themselves (as the political journalists of course have already noted.)⁹ There is, however, abundant evidence to the fact that important and unexpected decisions are made at the conference itself.¹⁰ A short-cut here would be to ask the different level activists about the importance they attach to the national conference, compared to other arenas for making decisions in the party (the committees, the secretariat, the central committee, etc.).

A discussion of the non-decision-making approach would seem to involve
the same kind of problems and solutions as that of decision-making. Lewis
Minkin amply demonstrates the amount of hard work necessary to describe

the process of conference preparation. And even if the journalists on a fulltime basis can bring to light lots of interesting material, in a limited research project there would be a question of capacity. And even if this is an area in need of more work and attention than hitherto attributed to it, a simpler device would be necessary. One possible way could be to carry out a systematic comparison of the kind of issues reaching the final conference agenda of the various parties, and to compare the different ways of handling them organisationally on their v/ay towards a "final" party decision.

---

NOTES

1. Connecting the three last approaches to some names and books: Hunter 1953; Dahl 1961; Brachrach & Baratz 1970.

2. Duverger (1964, xvi) mentions a study within this formal tradition which undertakes an analysis of the "constitutions of twelve of the principal parties of Central Europe", namely Hartman 1931.

3. A few exist, and some do marginally touch the question of party power. See for example Koht's biography of the 19th century Liberal leader Johan Sverdrup, vol.lll, Koht 1925, ch.XIII.

4. Although not exploring this theme in particular, Torgersen 1967 touches several aspects of the problem.

5. Burnham's argument falls very much within the so-called "crises of democracy" debate following the turbulent years of the early 1970'5. See, for example, Crozier et al. 1975.

6. The Norwegian "Study of Power" has produces numerous reports. A summary is presented in the official Norges Offisie/le Utredninger, no. 3, 1982.

7. Even though the party suffered a relatively small setback at the 1983 local elections.

8. The most ambitious and broadly based empirical study of recent years, Janda 1980, appears to me to have chosen an analytical focus too abstract and general to be of much help in testing this assertion. See also Torgersen 1967.

9. See for example A. Hompland's article in Dagbladet, 8/3-83 following the national conference of the agrarian Centre Party.

10. Taking but a few examples from, the Norwegian Labour Party (and it is perhaps not by chance that conference decisions from this party most readily come to mind): Pro abortion (1969), election of party chairman (1975), supporting the 6-hour workday (1981).

REFERENCES

Bachrach, P. & Baratz, M. 1970. Power and Poverty. Oxford: Oxford University Press.

Benum, E. 1969. Maktsentra og opposition. Spaniasaken i Norge 1946 og 1947. Oslo: Universitetsforlaget.

Burnham, W.D. 1975. "American Politics in the 19705: Beyond Parties?", pp. 308-357 in
W.N. Chambers and W.D. Burnham, The American Party Systems, 2nd ed. New York:
Oxford University Press.

Christensen, T. 1976. "Kvinner som politisk trekkplaster", Tidskrift for Samfunnsforskning,
vol. 17, pp. 135-157.

Clark, P.B. & Wilson, J.Q. 1961. "Incentive Systems: A Theory of Organizations' in
Administrative Science Quarterly, vol. 6.

Crozier, M.J., Huntington, S.P. & Watanuki, J. 1975. The Crisis of Democracy. New York
Nev/ York University Press.

Dahl, R. 1961. Who Governs. New Haven: Yale University Press.

Dahl, R. 1968. "Power", in International Encyclopedia of the Social Sciences, vol. 12. pp.
405-415. The Macmillan Company & The Free Press.

Dahl, R. 1971. Polyarchy. New Haven: Yale University Press.

Downs, A. 1957. An Economic Theory of Democracy. New York: Harper.

Duverger, M. 1964. Political Parties. Their Organization and Activity in the Modern State.
London: University Paperback Edition.

Eldersveld, S.J. 1964. Political Parties: A Behavioral Analysis. Chicago: Rand McNally.

Eriksen, K.E. 1972. DNA og Nato. Striden om norsk medlemskap innen regeringspartiet
1948-49. Oslo: Gyldendal.

Hartman. 1931. Die Po/itische Partei. Briinn.

Heidar, K. 1974. Institusjonalisering i Det norske Arbejderparti 1887-1940. stencil. Oslo.

Hellevik, O. n.d. Avgjørelsesprosesser i politiske partier. Et opplegg for empirisk analyse av
oligarkiske tendenser. Oslo.

Hunter, F. 1953. Community Power Structure. Chapel Hill: North Carolina University Press.

Kaase, M. 1980. "The crisis of authority: Myth and reality" in Rose, R. (ed) Challenge to
Governance. Studies in overloaded politics. London Sage Publications.

Koht, M. 1925. Johan Sverdrup. vol.lll. Oslo: Aschehaug.

Lipset, S.M. 1968. "Ostrogorskii, Moisei la" in International Encyclopedia of the Social
Sciences, vol. 11, The Macmillan Company & The Free Press.

McKenzie, R.T. 1963. British Political Parties. The distribution of power within the Conservative
and Labour parties. 2 ed. London: Heinemann.

Michels, R. 1962. Political Parties. A sociological study of the oligarchical tendencies of modern
democracy. New York: Collier Books.

Mills, C.W. 1956. The Power Elite. Oxford: Oxford University Press.

Minkin, L. 1978. The Labour Party Conference. A study in the politics ofintraparty democracy.
London: Allen Lane.

Neumann, S. 1956. "Toward a comparative study of political parties", in S. Neumann
Modern Political Parties. Approaches to Comparative Politics. Chicago: University of
Chicago Press.

Norges Offentlige Utredninger. nr 3. 1982. Maktutredningen. Sluttrapport. Oslo.

Sartori, G. 1976. Parties and Party Systems: A framework for analysis, vol. I. Cambridge:
Cambridge University Press.

Schlesinger, J.A. 1965. "Political Party Organization" in Handbook of Organizations.
Chicago: Rand McNally.

Schumpeter, J.A. 1942. Capitalism, Socialism and Democracy. London: Unwin University
Books.

Torgersen, U. 1967. Programutformning og partifasade: Noen trekk av norsk landsmøtepraksis.
Oslo.

Torgersen, U. 1969. Landsmøtet i norsk partistruktur 1884-1940. Oslo: Universitetsforlaget.

Valen, H. 1966. "The recruitment of parliamentary nominees in Norway" in Scandinavian
Political Studies, vol. 1.

Wilson, J.Q. 1962. The Amateur Democrat. Chicago: The University of Chicago Press.

Wright, W.E. 1971. "Comparative party models: Rational-Efficient and Party Democracy" in
Wright, W.E. (ed.) A Comparative Study of Party Organization. Columbus: Merrill.