Resumé

The purpose of this article is to try to discern certain general features in the origins
of the demands for referenda, and why these demands succeed or fail. A
distinction is drawn between (a) the demand for a referendum on a certain issue
and (b) the demand to amend the constitution to allow for referenda of one form
or other.

(a) Nation-wide referenda are rare almost everywhere, but demands for
referenda arise more frequently. The article argues there are often regularities in
the way these demands arise. A request for a referendum on one specific issue can
be regarded as a minority weapon. Those who have had their standpoint voted
down demand a referendum. If their demand is to succeed, two conditions are
important: parties which are split and strong commitment on the part of the
voters. To understand why an original minority weapon can get support from the
majority, two functions of the demand for referendum are central: the function as
a mediation device and as a lightning rod.

(b) Concerning interest in the introduction of the referendum as a more or less regular principle of government, we find that referendum enthusiasts often share a similar ideology or view of society. The most central catchword here is antiparty sentiment.

Introduction

Nation-wide referenda are rare occurrences in almost all countries. The nations of the world might be divided into three categories according to their use of the referendum: regular users, sporadic users, and those which have never made use of the referendum mechanism in any form. Switzerland is, strictly speaking, the only example of the first type. Certain states in the United States, among them California, are often mentioned in this connection, but state referenda are, of course, not nation-wide. The majority of the nations of the world have never held a

---

* I would like to thank Sten Sparre Nilson, Evert Vedung and Hans Fredrik Dahl for valuable comments on earlier drafts.

nation-wide referendum, although several countries have recently gone over to the ranks of the sporadic users. Among the nations of the industrialized Western world, we might mention Great Britain, which held its first referendum in 1975 in connection with the issue of continued membership in the EEC. Over one-third of the nations in the United Nations have held at least one referendum, and many of these again have held only one (Butler & Ranney 1978, 7).

One feature of much of the literature on referenda is, naturally enough, that it has been dominated by studies of and references to those political systems in which referenda are a regular feature of political life: Switzerland and various states in the United States. The basis of analysis is therefore two non-parliamentary forms of government in which the decision to hold a referendum will be made more or less automatically, either because the constitution demands a referendum in cases of constitutional amendment or because a certain number of voters have a constitutional right to demand a referendum. Both circumstances make these cases somewhat peripheral to the problem we shall raise here-. We are concerned with what can be called the irregular use of the referendum mechanism. We are interested in the demand that a referendum be held. Although referenda are held sporadically, there are often regularities in the way the demand for them arises irregularly. We ask: Why does the demand for a referendum arise and under what conditions does it succeed?

Butler and Ranney (1978, 18) discuss why referenda are held and state that 'common elements can sometimes be detected'. But they also argue the opposite: The reasons for each referendum fail to fit any clear universal pattern. 'Each seems to have a special history'. It is undoubtedly difficult to generalize on this subject, but I believe it is possible to find certain patterns.

We shall draw a distinction between the demand for a referendum on a certain issue and the desire to amend the constitution to allow for referenda of one form or another. As is often the case in social research, this hard theoretical distinction is somewhat more permeable in practice. Interest in an abstract principle like the referendum is often sparked by the demand for a referendum on a particular issue._'

The demand for a referendum on a particular issue will be limited to so-called voluntary referenda in which a political majority can decide to hold a referendum on a particular issue. Evidence will be taken mostly but not exclusively from Norwegian cases. Empirical data on the demand for a constitutional amendment to allow for referenda must, of course, necessarily come from countries which do not already have this provision

in their constitutions. The data examined in this article will come
exclusively from Norwegian cases. It may therefore be useful to give a
short summary of the referendum mechanism's status in Norway.

Limited Use of the Referendum in Norway

General Referendum

The referendum is not institutionalized in the Norwegian constitution, though many attempts to incorporate it have been made. The question has been up for consideration in the Storting twenty times, but it has never achieved the required two-thirds majority for constitutional amendments (Holby 1958; Björklund 1974).

The proposals have overwhelmingly concerned referenda. Proposals concerning obligatory referenda have been put forward in connection with constitutional changes and mostly in the form of a conservative 'guarantee'.² A referendum would come in addition to the present requirements for constitutional amendment. These proposals have, as a rule, been put forward by the political Right.

Most proposals have concerned facultative referenda and would give a minority in the Storting the right to demand a referendum. The direct popular initiative - where the voters themselves are given the lawgiving authority - has never been proposed in the Norwegian Storting.

Specific Issue Referendum

In terms of constitutional law, the failure of the constitution to mention the referendum has made consultative referenda the only kind possible. A majority of the representatives in the Storting can decide to submit an issue to the whole electorate. Referenda in Norway are thus voluntary referenda.

Nationwide referenda might almost be considered irregular phenomena in Norway. True enough, five have been held and in comparison with other West European countries this is not so very few.³ Pressure from abroad was in large part responsible for the two Norwegian referenda held in 1905 concerning the dissolution of the Union with Sweden and whether the newly independent country would be a republic or a monarchy. The demand for the first referendum - that concerning dissolution of the Union - was Swedish in its origin (Vedung 1971). And the referendum held three months later on whether Norway would be a republic or a monarchy was possible primarily because the republicans' demand for it was supported by the Danish royal candidate himself.

The referenda concerning prohibition (1919 and 1926) and that concerning Norwegian membership in the EEC, on the other hånd, were mandated primarily by internal political considerations and are therefore of special interest to this article.

Statement of the Problem and Evidence

We will first consider interest in the institution of the referendum in
general in Norway; after that we will discuss the demand for voluntary
referenda on specific issues and under what conditions it succeeds.

Political ideas do not exist in a vacuum unaffected by the political environment but must be seen in relation to strategical or tactical considerations and the political contexts which shape political opinion. As one Norwegian historian put it, The arguments are determined by the goal one pursues and not by the issue itself (Steen 1951, 181). Likewise the demand for the referendum will often arise as a means to achieve goals other than that most often expressed, namely concern for democracy.

We will examine interest in the institution of the referendum both inside and outside of the Storting. Referenda have that consequence that decisions usually made by politicians - either through political parties or the government - are transferred to the voters themselves. It is therefore natural to look at the desire to introduce a referendum clause of one form or another into the constitution in relation to anti-party sentiments and the relative power ratios between the political parties which are the parliamentary basis for the formation of a government. One suggestion is that the interest in the institution of the referendum is connected with anti-party sentiments and an opinion that the deciding politicians are not on the same side as the voters, which again can be seen in relation to the parliamentary situation.

The fact that we limit our attention to voluntary referenda in itself sets limits on the evidence we can bring forward. Of the seven Danish referenda in the postwar era, five have been obligatory (1953: New constitution and voting age; 1961, 1969, 1971, 1978: voting age) and one was facultative (1963: Agricultural acquisition law).

Referenda are obligatory in cases of constitutional amendment in Denmark. There is also a clause in the constitution which gives a one-third minority in the Folketing the right to demand a referendum on certain kinds of proposals. The Danish referendum on membership in the EEC (1972) took the form of a voluntary referendum as a result of a

tactical manoeuvre (Bjørklund & Martens 1979; Martens 1979). All of the
nation-wide referenda in Sweden have been voluntary.

The Referendum Institution in General

Anti-Party Sentiments

In the history of the referendum in Norway, there is a slight anti-parliamentary tradition. The referendum issue played a certain role in the opposition to the establishment of parliamentarism in 1884. The desire for a referendum was put forward by those who found themselves on the losing side in 1884. It was often tied to two other demands - the right of the government to dissolve parliament and the introduction of the two-chamber system - both of which were much more prominent than the referendum issue. All three of these desired reforms had their origin in a fear of a too radical and capricious Storting, and were considered conservative 'guarantees' (Seip, A.L. 1972). It was a question of holding referenda on acts already passed by the Storting so that the voters could have opportunity to veto them.⁴

A more inclusive viewpoint than anti-parliamentarism from which we can study the demand for referenda is anti-party sentiment. The relationship between anti-party sentiments and support for referenda can be studied with the aid of several sources.

Anti-party movements which become parties often have a referendum plank in their electoral platform. Trie folkevalgte', who first appeared in the 1953 Storting election campaign, unambiguously criticized party politics. They also demand more frequent use of the referendum mechanism.

Denmark has a similar party, 'Dansk samling', which is almost as peripheral as 'Frie Folkevalgte'. Their platform also combines anti-party sentiments with the demand for referenda.⁵ A more recent example of this combination - and a more important one - is Glistrup's support for frequent use of referendum (Nielsen 1979, 7).

Typically, political 'may-flies' - which stand for election in one campaign, fail and disappear - often demand referenda on a wide spectrum of issues. We could look at these demands for referenda from a political power perspective. Those who stand outside the established political party system and want to enter it can, in addition to starting a political party, demand referenda. The guiding idea behind the creation of the party and the demand for a referendum is the belief that the existing parties are out of step with popular opinion.

But in addition to this political power aspect, one can detect an ideological affinity between anti-party sentiments and the belief in referenda. The demand for a referendum is often basedon the belief that political parties confuse the popular will. The use of the party whlp can transform the will of a minority of the representatives into a majority decision. As John Stuart Mill put it, A majority of the majority, who may be and often are but a minority of the whole'. Another factor contributing to the lack of coherence between party platforms and the popular will is that a voter is seldom in total agreement with (or even aware of) all the various party viewpoints. Moreover, new issues can emerge which have not been considered in the party platform. And finally, electoral arrangements seldom assure a perfect fit between votes and representation.

Direct Democracy and Liberal Belief in Elites

Anti-party sentiments can often be linked with elitism. Referenda and idolization of elites can undoubtedly at first glance appear irreconcilable. Scanning the literature written by Norwegian referendum enthusiasts available at the University Library in Oslo, one often finds the belief in elites combined with a belief in direct democracy (Bjørklund 1974).

The perhaps unexpected connection between support for direct (democracy and the idolizing of elites must be seen in connection with the depoliticized view of society which referendum enthusiasts often have. The goal of politics is to discover 'the correct solution'. And there is a solution which is to the best for everyone. In this quest for the correct solution, the gifted person is not a slightly dangerous person who brilliantly defends the interests of his own group or class. Instead, he is the one who more clearly sees the correct solution. A depoliticized view of politics can also easily lead to a contempt of political parties. Parties create antagonisms. And the party organizations formalize and harden these antagonisms, complicating necessary compromises. The free and independent men - 'the best men' - cannot accept party'blinders'.

One representative for the combination of the two theories - direct democracy and the liberal belief in elites - is Johan Scharffenberg, who might be called Norway's most steadfast supporter of the referendum idea through the years. In his book 'Quo Vadis Norvegia?' (1945), concepts like 'quality', 'expertise' and 'capable men' are highly praised and placed in something of a political vacuum. The underlying idea was that the referendum would give the politically outstanding person more 'elbow-room' than he has in the closed corridors of party politics.

Scharffenberg is by no means the only representative among enthusiasts
who combine elements of liberal belief in elites and support for the
referendum idea.

Apolitical Trends and Interest in the Referendum

The political interest in introducing legislation to allow for the widespread use of the referendum mechanism can be seen in connection with apolitical trends. The Norwegian word for referendum, folkeavstemning', in itself has a certain apolitical ring to it. It consists of the politically unifying word 'folk' or 'people', which according to a well-known Norwegian historian 'emphasizes the unity of and disguises cleavages in the social landscape' (Seip 1958, 459).

In the'home of the referendum' - Switzerland - the government is a coalition based on the various parties' representation in the national assembly. The idea is that of an all-party coalition. Belief in this form of government is based on the assumption that deep cleavages in political values have been overcome. A broad basis of agreement on common values exists between the largest parties. This is in harmony with a depoliticized view of society.

The publication date of books and articles written by referendum enthusiasts can give us some idea of what kind of political climate nourishes the desire for referenda. The literature I have been able to find in the University Library in Oslo is hardly voluminous and certainly not exhaustive. Of the nine titles I have found, four fall into the period 1905 - 1913 and four in the period 1945 - 1952 (Björklund 1974). This literary activity corresponds well with interest in the referendum in the Storting. The years just after 1905 and 1945 were a time in which interest in the referendum idea was relatively strong in the Storting and less than usually conditioned by tactical considerations. Helge Holby writes in the conclusion to his dissertation 'Government by Representation or by the People: Storting Debate Concerning the Referendum 1892- 1932 and 1932- 1956':

According to my understanding, the necessary political conditions for the introduction of the referendum have only existed two times. The first time was from around 1900 until 1908. Proposals were submitted by both Høyre and Venstre, later also by the Socialists . . . The second time was just after the last war, while we still lived in the time of the Common Political Program (Fellesprogrammet) (Holby 1958, 123).

It is tempting to draw attention to a remarkable similarity between these
two periods which might be characterized by the expression apolitical
trends and, we might add, optimism. After the dissolution of the Union

with Sweden in 1905, 'the new work-day' was to begin. The hatched was to be buried. The idea of national unity was also strong during the reconstruction after the last war, and resulted among other things in the Common Political Program which contained a clause concerning the referendum: The question of referendum and the right to dissolve the Storting shall be put forward for new consideration'. The idea was that communists and capitalists who had both fought against the Nazis during the occupation should continue to work together to rebuild the country.

Interest in the referendum in these two periods can undoubtedly also be explained in other ways, for example by arguing that a time of occupation inevitably gives rise to a desire for ultrademocratic reforms. And in the years before and after 1905 the nation had undergone changes which led to the extension of democratic rights such as universal franchise. It was a time in which democratic institutions were being built up, and the introduction of the referendum could be presented as a logical next step. And the two referenda held in 1905 had made the institution an actual alternative.

When the Government is Regarded as Either too Strong or too Weak

Anti-party sentiment can of course be the result of other things than a depoliticized view of society. We shall mention two forms of anti-party sentiment which have their basis in two diametrically opposed situations: a too-weak government and a too-strong government.

Interest in the referendum in the years between 1905 and 1940 could have taken as their starting point discontent over a too-weak government. Referenda, whether called for by the government or demanded by the voters, could possibly rectify this somewhat and restrain 'government by parliamentary manoeuvring'. In the postwar period, on the other hand, it was usual to regard the referendum as a weapon to curtail the power of a government which had become too powerful in relationship to the Storting and the people. The parliamentary majority enjoyed by the Labour party from 1945 to 1961 had created discontent which came to expression in concepts like 'the one-party state' and a feeling that power was no longer exercized by the Storting but was vested in the Labour party organisation's central committee. To introduce legislation to allow the government to call for a referendum - the solution which some had hit upon to correct the imbalance of power in the inter-war years - would now only worsen the situation.

Actually, the referendum proposals submitted to the Storting both
before and after the war have most often concerned the right of a

minority of the Storting to appeal a decision of the majority through a referendum. In the Storting debates, the referendum has most usually been regarded as a minority weapon - an opportunity for the opposition. No constitutional amendment to allow the government to call a referendum has ever been submitted. There has thus been a certain correspondence in the post-war years between the debate within the Storting and that outside it, more so than in the period 1905 - 1940.

The period from 1905 until the Second World War was largely marked by minority parliamentarism. There was one continuous period from 1913 until 1918 with a one-party government and majority parliamentarism under Gunnar Knudsen. But there were ten different governments between 1920 and 1935. And in the period just after 1905 when there was a certain amount of interest in the referendum, the parliamentary situation was shifting and unclear. Officially, the Liberal Party (Venstre) had a majority in the Storting between the elections of 1906 and 1909, but the party never managed to form a government that had the support of the whole parliamentary group. The unclear situation meant that the government never sat securely (Björnberg 1939). In accordance with this parliamentary situation, the referendum literature published just after 1905 was often concerned with the over-powerful Storting (Heistein 1909; Koht 1909).

One would get an exaggerated estimate of interest in the referendum if party platforms were our only source. Referenda are often mentioned here. In the elections of 1965, 1969 and 1973, every party with the exception of Labour made some reference - never negative - to the referendum in their platforms in each of these years.

The referendum disappeared from the Labour party programme in 1930. It would be natural to see this in connection with the Labour party's electoral breakthrough in the election of 1927 which set the stage for the short-lived Labour government led by Hornsrud. It is also interesting to note that the Liberals, the party which today has the most extensive references to the referendum in its platform and the only party which has included a referendum plank in all its election platforms since 1945, did not go in for the referendum while it was at the height of its power. Under Gunnar Knudsen's second government in 1914, when the Liberal Party had a majority in the Storting, both Labour and the Conservatives submitted referendum proposals. The Liberal Party voted down these proposals in 1917.⁶ This underlines what we have previously mentioned: the referendum has most often been regarded as a weapon of the minority in the Storting.

In the post-war period the referendum has mostly been regarded as an

instrument to strengthen the hånd of the people or a minority in the Storting vis-å-vis the government or the parliamentary majority. The unbroken parliamentary majority for Labour in the period between 1945 and 1961 contributed to the interest in referenda during this period.

In 1961 an organization calling itself the 'National Association for the Referendum' ('Landsforbundet for folkeavstemning') was formed. It never achieved the kind of popular support its founders had hoped for. The aim of this organization was to amend the constitution to allow for the referendum. And in June 1964 a proposed constitutional amendment was submitted to the Storting. The proposed amendment was more or less identical with proposals that had been introduced three times before and voted down by the Labour Party in 1956, 1959 and 1964. In brief, the proposal would have given a one-third minority of the Storting the right to demand a referendum on certain issues. Again it was the non-socialist parties which supported the proposal. In 1968 it was voted down by the Labour Party, which was then in the opposition.⁷

Since the referendum has most often been regarded as a minority weapon in the Stoning, it is in a way logical that every proposed constitutional amendment has been voted down. The fact that an amendment requires a two-thirds majority does not make the situation easier. The simple majority which the referendum achieved in 1968 was not sufficient.

Individual Referenda

The Demand for a Referendum in Norway on a Certain Issue

Looking at the demand for a referendum on a certain issue, we might argue the same as we did for referenda in general: one reason there have been so few nation-wide referenda is that the demand has often been raised by the minority. For those whose standpoint has been been defeated through the ordinary political channels, the demand for a referendum can be seen as the issue's last chance. The demand for a referendum on a certain issue can also be seen as a barometer of voter dissatisfaction. The demand often arises on issues in which voter interest is strong and uncompromising. It has been raised with varying degrees of intensity in many of the traditional controversies of Norwegian politics.

At the local level the referendum, combined with the initiative, is being used regularly with two issues: local prohibition (licensing of the sale of liquor) and the language controversy (the use of the 'nynorsk' or 'bokmål' language in primary schools).

On these issues there have also been demands for referenda on the national level. Norway has held two nation-wide referenda on the prohibition issue. But the demand, raised by the 'riksmål'-speakers union, for a national referendum on the language issue has never succeeded. The referendum has likewise been proposed in connection with certain other post-war political controversies such as NATO membership (1949, 1965-69), the price control and rationing laws (1952), election laws (1952) and foreign aid. This list is not exhaustive. But what is certain is that the demand has in only one case been raised with enough vigour and insistence to succeed: the issue of Norwegian membership in the EEC.

In an analysis of why Norway has held five nation-wide referenda, two referenda are of special interest: the prohibition referendum of 1919 and the EEC referendum of 1972. The two referenda of 1905 were, as mentioned earlier, conditioned primarily by external political considerations, and the 1926 referendum on prohibition followed logically from the one conducted in 1919.

The Three Functions of the Demand for a Referendum: Minority Weapon, Mediation Device and Lightning Rod

On the basis of an examination of the background for the prohibition referendum (1919), the EEC referendum (1972), and the abortive demand for a referendum concerning continued Norwegian membership in NATO in the late 1960'5, it is tempting to say that the demand for a referendum has three central functions.

At first glance it may appear paradoxical that a political majority in a voluntary referendum consents to uncouple itself from the decisionmaking process and turn the question over to the voters. Consider the case in which those whose standpoint would be victorious in the absence of a referendum oppose the referendum idea. Here we can observe the demand for a referendum in its function as a weapon of the minority. Those whose standpoint would be voted down if it went through the normal channels of representative democracy can embrace the demand for a referendum on the issue. In the absence of a referendum the battle is lost, one has nothing to lose and everything to gain by demanding a referendum. The referendum is the issue's last chance. This demand for a referendum as a minority weapon may come from a political party which is united on an issue or from a faction within a party. In the first case it is directed against the parliamentary majority and in the other case it is directed against the majority within one's own party.

In regard to standpoints on a particular issue, we might distinguish between opposition parties and establishment parties. The demand for referenda is raised as a rule by opposition parties and the opposition within the establishment parties.

To clarify this let us distinguish between politicians who find themselves in the majority position and those who hold the minority position on a particular issue. Consider this distinction in relation to their standpoint on the demand for a referendum - as proponent or opponent. This can be illustrated in the following way:

One would expect most of the antagonists on both sides of the issue to cluster themselves in the two hatched areas. The problem then is to explain how someone might fall into area number two. The two other functions of the demand for a referendum - mediation device and a lightning rod - help to explain these cases.

The referendum may be used as a tool for mediation in a conflict. When a party or a government is divided on an important issue, it can be in danger of breaking up. The smaller the majority and the more important the issue, the greater the threat of lasting cleavages. In such a situation a party may embrace the referendum as a mediating device. The minority which is voted down can be reassured that the decision is in a way only temporary. The voters will have the last word. The war is not lost just because the majority of the party has voted the other way. And the majority has reason to hope that a destructive cleavage has been avoided because the minority has not been provoked.

The third function of the demand for a referendum is that it may act as a lightning rod for dissent. Certain politicians may come to regard a particular issue as an irritating source of dissatisfaction which opponents confront them with time and again and use as a rallying point to mobilize opposition. The call for a referendum can then act as a lightning rod and

remove the issue from the party political agenda. The idea can be that one is almost certain to win the referendum, and the legitimacy with which such a democratic procedure invests the decision will settle the issue once and for all. If the will of the people is expressed unambiguously and without charges of manipulation or pressure, it will be difficult to oppose it.

The idea of the referendum as a lightning rod often presents itself before an election campaign where the need to uncouple certain issues from the campaign can be powerful. The election result may depend heavily on what issues are salient. The demand for a referendum can be a well-suited instrument for removing a certain issue from the campaign by arguing that it does not belong there. Everyone will have a chance to express his or her view later, through the referendum.

Norwegian EEC-Referendum (1972)

The best example of the referendum as a mediating device occurs when the demand for a referendum has the support of both sides from the very beginning and is not regarded as a demand of the minority. It was just an example such as this which occurred in August/September 1961, when Bent Røiseland became the first leading politican, and the Liberals the first party, to demand a referendum on the EEC question. The Liberal Party was split on the question about Norwegian membership in the EEC, but not so deeply as in 1972 (Björklund 1977).

The anti-membership parties Socialist Peoples' Party (SF), Norway's Communist Party (NKP) and the Agrarians (SP) saw the demand for the referendum mostly as a minority weapon. Even here, however, there were those who were sceptical toward the idea of a referendum and believed it would be easier to mobilize the required 38 or more opponents in the Storting than it would be to win a majority in a referendum.

It was up to Labour to decide whether a referendum would be held. If they had opposed it, they would most likely have achieved a majority with the Conservatives' support. The majority in the Labour party was, from the very beginning, opposed to a referendum.⁸ However on April 10, 1962 the Labour party Storting group voted unanimously for the referendum. By that time it was clear that the moderate centrist parties were in favour of a referendum, and the Labour party would have found itself in the Conservatives' corner had they opposed it. But a more important reason why the Labour party supported the referendum idea was that they viewed it as a potential mediation device. The party was split on the EEC question, and in contrast to the Liberal Party in

1961/62, this split was along an old line of conflict: The Right/Left dimension. The problem for the party leaders was exacerbated by the fact that in May 1961 a new party - SF - had been founded by members of the party's Left wing. Labour party voters who opposed membership in the EEC therefore had an alternative in a party which was unambiguously opposed to membership and which had done surprisingly well in the last election. The decision to support the demand for a referendum can be seen as an attempt to hold the party together.

Prohibition Referendum (1919)

The prohibition referendum of 1919 can also be analyzed in terms of these two functions: the referendum as a minority weapon and as a mediation device. The Liberals were the leading party of the second decade of this century, and it was during Gunnar Knudsen's term - a Liberal government - that the first prohibition referendum was held. The Liberal Party's election platform of 1918 called for the prohibition of all liquors containing more than 12% alcohol by volume. The one condition was that this standpoint had to be supported by a majority of the voters in a referendum. The addition of this condition might be explained by the fact that the Liberals were split on the prohibition issue. The demand for a referendum was a concession to the minority which opposed prohibition and functioned as a mediation device.

In the beginning, the demand for a referendum had been a tactic of the minority who supported prohibition. When the first demand for a referendum was raised in the prohibition movement's newsletter Afholdsbladet' in 1909, prohibition supporters were in the minority in the Storting. But the lobbying and campaign efforts of the prohibition movement soon bore fruit. The so-called prohibition party seldom nominated candidates on its own, but demanded instead that the various other candidates take a standpoint on the prohibition party's demands.

Since prohibition activists presented the question in terms of a test of the candidate's moral character, the question was tactically sensitive, and several probably felt themselves pressured into publically supporting the demand. The prohibition movement thus had a majority in the Storting in 1919 - 69 of 129 representatives (Sverdrup 1972, 30). The desire for a referendum had meanwhile waned in the prohibition movement. The idea was rejected by the 'Avholdstinget'. The referendum was no longer seen as the last chance of imposing prohibition. The prohibitionists' extraordinary national convention of 1919 nevertheless accepted the idea of a referendum, although it was with obvious reluctance (Nilson 1972, 100).

It would have been difficult for the prohibition movement to oppose the demand for a referendum which the Liberal Party now supported. The referendum idea was by this time an almost unshakeable article of belief in the prohibition movement, and once the demand for a referendum is raised, it has a tendency to capture the imagination.

The Demand for Referendum on NATO-membership

Aside from the question of Norwegian membership in the EEC, there is only one other issue in which the demand for a referendum has achieved some support among leading politicans: the question of continued membership in NATO. This referendum demand illustrates the lightning rod function and also shows how a split within a party can be more easily handled when strong engagement on the part of the voters is lacking.

After twenty years of membership the possibility of withdrawing from the military alliance was allowed. This was one of the reasons that the debate on Norwegian membership in NATO arose around the end of 19605. The history of the referendum demand in the NATO question shows many similarities with the demand for a referendum on the EEC question. Again, the Liberals were the first party to raise the demand. The Liberal Party's 1965 national convention unanimously passed a platform which supported continued NATO membership in 1969, but added: Tf the situation demands it, an advisory referendum on continued membership in NATO should be held before the Storting makes the final decision'. The condition for this unanimous proclamation of support for continued membership in NATO was the concession given to NATO opponents in the form of support for the referendum idea. Here again the referendum was used as a conciliatory mediation device.

The demand for a referendum on the NATO question did not long remain an internal demand from the Liberals. More interesting than the fact that SF and NKP supported the demand, is the way in which the Labour party handled the question. At the Labour party national convention in May 1965, NATO opponents demanded a platform plank inserted calling for an advisory referendum on the NATO question in 1969.⁹. However, the demand met much resistance.

In the Labour party, it was not only opponents of NATO membership who raised the idea of a referendum on the question. On January 28, 1966 Einar Gerhardsen held a speech for the Student's Society entitled The Labour Movement and the Future'. According to the report on the speech in 'Arbeiderbladet', he mentioned NATO and argued for continued membership, but at the same time said he 'thought there were

many good reasons for holding an advisory referendum.¹⁰ In addition to the possibility that the referendum demand could build bridges between proponents and opponents of membership, Gerhardsen no doubt had other possible consequences of a referendum in mind. The justification which he mentioned for a referendum in his speech was:

If the Norwegian people are allowed to express their opinion and the result is a clear
decision, there will perhaps be better reason to hope that we can put an end to what the
opponents call a fruitless debate.11

This is the idea of the referendum as a lightning rod. Opinion polls unmistakeably indicated that a solid majority of Norwegians were in favour of continued membership. An unambiguous referendum result might put the question to rest; the will of the majority would have to be respected.

On March 6, 1966, the National Committee of the Labour Party youth organization AUF released a statement which took no standpoint on NATO membership but unanimously supported the idea of a referendum on the question.¹² This statement might be given as a textbook example of the referendum as a mediating device. One avoids taking a standpoint on the issue and instead unites on the demand for a referendum. Likewise, 'Radikale Venstre' in Denmark decided at their national convention in June 1971 not to take a standpoint on one of the most important issues in post-war Danish history - the question of Danish membership in the EEC. The party was split, so the convention united instead on the demand for a referendum and the wish for a unbiased information campaign.₁₃₎

Thus many of same conditions that led to a referendum on EEC membership were present in the NATO debate. But at least one condition was lacking - strong engagement in the question on the part of the voters, an engagement which manifested itself in the EEC campaign in widespread organizational activity outside the regular political parties. One of the main slogans of the EEC opponents' demonstration in 1962 was: 'Ask the people'.

The Mediation Device: Theory and Reality

The Liberal Party - perhaps Norway's most heterogeneous political
party - has several times tried to use the demand for a referendum on
certain issues to keep various groups inside the fold. But the intended

result has not always been achieved. There was true enough no split in
the Liberal Party as a result of the prohibition referendum of 1919.

The Swedish liberal party 'Frisinnade landsföreningen' on the other hand was split partly as a result of the Swedish prohibition referendum of 1922. The party was divided in its standpoint on prohibition and had been among the initiators of the referendum idea (Lundkvist 1974, 347; Thulstrup 1968; 116). Fifty years later - in 1972 - the Norwegian Liberal itself was split as a result of the EEC referendum which politicans from the Liberal Party and the Liberal press had been the first to demand. The newspaper 'Dagbladet', with EEC-opponent Helge Seip as editor, was the first newspaper in Norway to demand a referendum in an editorial published August 1, 1961 and entitled 'Give the People the Last Word'. The then chairman of the Liberal Party gave slightly hesitant support to this editorial demand in an interview in the same paper. Eleven years later, Helge Seip, by then national chairman of the Liberals and a proponent of membership in the EEC, walked out of the Liberal Party's national convention in Røros in November 1972 together with a number of other Liberal leaders to form 'Det nye Folkeparti'.

One cause for holding a referendum is that conflict dimensions do not fit the party-structure: Political parties are divided. The demand for referendum can function as a bridge builder between conflicting wings, but often temporarily: The referendum campaign can widen instead of reduce party cleavages.

The Three Functions in Relation to Referenda in Great Britain, Denmark and Sweden in the Postwar Period

Denmark, Great Britain and Ireland also arranged referenda concerning EEC membership. The Irish referendum was obligatory. In Denmark and Great Britain, on the other hand, the politicians themselves decided to put the decision in the hands of the voters in the form of a voluntary referendum, as was the case in Norway.

In the case of Great Britain, it was Labour which launched the idea of a referendum on British membership in the EEC. The background for this decision was first of all that the party was deeply divided. The demand was raised as a minority -weapon by opponents of membership, not only from Labour, but from 'a handful of anti-Marketeers in all parties' (Goodhart 1976, 26). And the referendum demand did not function as a typical mediation device in the Labour party in this case. There was

strong disagreement on the desirability of a referendum, and the first
party resolution in favour of a referendum prompted Roy Jenkins to
resign his post as vice-chairman of the party (Grimond & Neve 1975, 19).

The Danish decision to hold a referendum on EEC membership was taken in May 1971. This decision must be seen in connection with the imminent election. The Social Democrats and 'Radikale Venstre' were both divided and the political distance to alternative parties which were clearly opposed to membership SF and 'Retsforbundet' was not great. Voter defection was therefore a present danger (Bjørklund & Mentens 1979; Martens 1979). The demand for a referendum functioned both as a lightning rod for dissent, uncoupling the issue from an election campaign, and as a mediation device. After the election in September 1971, the number of EEC opponents in the Danish Folketing exceeded the critical number needed to bring about an obligatory referendum (Martens 1979). In other words, without the decision in May, it would have been a referendum concerning Danish membership in the Common Market.

An important background factor for the three Swedish referenda In the postwar period is political splittage, not so much within parties as between parties in coalition government. The three postwar Swedish referenda concerned: Right-hand driving (1955), the supplementary pensions question (1957) and nuclear energy (1980). The first two referenda came during the era of coalition government formed by the Social Democrats and the Farmers' Party (Landsbygdspartiet Bondeförbundet), a coalition which was established in 1951 and which ended in 1957 as a result of disagreement over the supplementary pensions question. The referendum result had provided no clear resolution of the issue.

The background for the referendum on right-hand vs. left-hand driving is, as far as I can see, not thoroughly studied in the specialized literature on the subject. In a Norwegian newspaper interview, Professor Jörgen Westerståhl stated that 'the referendum never would have taken place had there not been divided opinion in the government'.l4 Westerståhl's source for this contention is the prominent Social Democrat and former Swedish prime minister Rickard Sandier.¹⁵ Worth mentioning is also that ten years before this referendum, in 1945, the idea of holding a referendum on right-hand driving was launched in a parliamentary committee (Wallin 1966, 334). At this point of time the different stands on right- versus left-hand driving were most probably not yet clarified. Perhaps an important background factor for the demand was the fact that the solution adopted would have direct consequences for daily life of every voter. It seemed therefore natural that they themselves should decide. The final impetus for this referendum appears to have been the

following: Internal disagreement led to conflict on the government's
position, and so as not to complicate intra-governmental cooperation the
final decision was left to the voters.

The supplementary pensions question was a recurrent issue in Swedish politics during the last half of the 19505. The demand for a referendum on this issue was put forward by the Liberal Party (Folkpartiet) shortly after the referendum on right-hand driving was held. The Conservatives (Högern) came out in support of the referendum shortly thereafter (Molin 1965, 60). Both these parties were opposed to the compulsory supplementary pensions for all employees, a plan which the Social Democrats had proposed. The demand for this referendum might thus be characterized as an opposition demand or a weapon of the minority. But the decisive factor that led to the referendum was the fact that the two parties in the coalition government - the Social Democrats and the Farmers' Party (Centerpartiet Bondeforbundet) had divergent views on the question of supplementary pensions. This disagreement was manifest in the formulation of the referendum question. The government was placed in the unusual situation of supporting two of the three alternatives to be voted on. The Social Democrats supported compulsory supplementary pension for all employees to be financed by the employers and administered by government authorities. The Farmers' Party (Centerpartiet) defended a government-administered supplementary pension which would be voluntary. The opposition - the Conservatives and Liberals - proposed a third alternative. Since none of the alternatives received a majority, the referendum did not resolve the conflict. Instead, the coalition government was dissolved and a Social Democratic minority government was formed.

The March 1980 referendum on nuclear power in Sweden can be
described in three phases:

a) The idea was launched in the Riksdag in 1975 by representatives from
VPK (Communist Left Party) as a minority weapon (Vedung 1978).

b) Secondly, Fålldin's coalition government declaration of 1976 brought up the idea, albeit in very unclear terms. If 'opinions' on the next important energy issue to be decided in 1978 were 'strongly divided', 'the possibilities' for an advisory referendum would be 'put in order' (Vedung 1978). The question of nuclear power was the most divisive issue the Fålldin government faced, and one of the motives behind this mention of the possibility of a referendum was the desire to hold together not a divided party but a divided government coalition. In other words, the referendum was intended to serve as mediation device. Evert Vedung has also shown how the Farmers' party used the

mention of a referendum in the coalition government declaration as a threat - the stick attached to the carrot - to make the other two parties in the coalition, the Conservatives and the Liberals, agree to the demand not to begin construction of the eleventh reactor (Vedung 1979, 114). The demand for a referendum arose again and became a central issue in connection with the fall of the Fålldin government in October 1978. It was the nuclear energy issues which finally brought down the coalition. Fålldin had embraced the referendum idea as a last chance to save the coalition, but failed. According to newspaper accounts the two other parties in the government coalition - the Liberals and the Conservatives (Moderaterna) - demanded that the government stand united behind one alternative as a condition for their support for the referendum idea. Perhaps it was the memory of 1957, when the two governing parties each supported its own alternative, which haunted them.

c) The third phase and the decisive go-ahead for the Swedish referendum on nuclear energy to be held in 1980 were provided by the Social Democrats' support for the idea. The Social Democrats' conversion to the referendum idea took place in the beginning of April 1979 in connection with the nuclear accident at Harrisburg, Pennsylvania. There were certain reasons for rethinking the nuclear energy issue at this time, but mostly political motives lay behind the demand for a referendum. The Social Democratic support for the referendum idea can be viewed as an issue_-uncoupling mechanism or, in other words, as a lightning rod. They wanted to remove the divisive nuclear power issue from the upcoming (September 1979) election campaign and thereby avoid a repetition of the 1976 election in which the nuclear issue was prominent and the Social Democrats suffered a severe setback.

Conclusion

Norway is one of the many countries in which the referendum is used, in
the words of Butler and Ranney, as an 'ad hoc solution to a particular

. . . problem'. (Butler & Ranney 1978, 221). Despite the irregularity implied by the expression 'an ad hoc solution', we can indicate certain general features which probably have validity not only under Norwegian conditions but also outside its borders.

We have drawn a distinction between the demand for a referendum
on a specific issue and interest in the introduction of the referendum
mechanism as a more or less regular principle of government.

Concerning the patterns we find surrounding interest in the referendum
as a general principle of government, the result can be summarized
as follows:

The question of amending the constitution to formalize the right of referendum has often been discussed by the Storting. A review of the various proposals shows that the referendum mechanism is often intended to strengthen the position of the opposition and function as a minoriy weapon. Gunnar Wallin's study of the debate on the referendum institution in general in the Swedish parliament comes to similar conclusions as far as the postwar period is concerned (Wallin 1966).

We also find that referendum enthusiasts outside of the Storting often share a similar ideology or view of society. The most central catchwords are anti-party sentiments, which have been associated with apolitical trends, the liberal belief in elites and two opposing poles in parliamentary practice: minority parliamentarism with frequent changes of government and majority parliamentarism with one party in a seemingly unassailable position.

The demand for a referendum on a specific issue arises as a rule from a minority - those who have had their standpoint voted down on its way through the normal political channels. If the demand is to succeed and the minority become a majority, two conditions are important: parties must be split internally, and the issue must be a salient one making the cleavage visible to the voters.

We have emphasized three functions of the demand for referendum. It can be used as a weapon by the political minority, as a conciliatory mediation device by divided parties and a lightning rod for dissent which can remove the issue from the normal political game. The referendum as a mediation device can be used as a political bridge-builder between two conflicting wings in a party. The need for this mediation device can be accentuated in an election year when party divisions can become easily visible and when the need to show party unity and strength to the voters is particularly acute. Raising the prospect of a referendum in an election campaign cannot only unite the divided party but can also function as a lightning rod and uncouple the issue from the campaign agenda. The issue becomes irrelevant in that forum since the voters will have a chance to express their opinions later: there will be another contest.

---

NOTES

1. A survey conducted by the Norwegian Central Bureau of Statistics just before the 1972 Common Market referendum showed that 42% of the opponents of membership vs. only 13 % of the proponents desired a more frequent use of referenda. The Common Market referendum has, however, not led - as one might have expected - to any new or strengthened interest in the use of the instrument.

2. But there have also been proposals to facilitate constitutional changes through the use of referenda. A member of the Norwegian Communist Party and of the Electoral Arrangements Commission of 1948, Jörgen Vogt, proposed that motions on constitutional amendments "which after second consideration do not receive two-thirds of the votes in the Storting, but over one-half, could be sent by the Storting to a binding referendum'. Quotation from: Proposals by the Electoral Arrangements Commission of 1948 (p. 82) (Instilling fra Valgordningskommisjonen av 1948).

3. In fact, there are only four countries in Western Europe which have held more nationwide referenda since 1900: Switzerland (approx. 250), Denmark (14), France (10), and Ireland (7) (Butler & Ranney 1978).

4. One of those who most clearly opposed Johan Sverdrup, the father of parliamentarism in Norway, was T. H. Aschehoug. In 1892, ten years after he had left the Storting, he submitted, through a Conservative member of the Storting, a proposal for a constitutional amendment concerning referenda. This proposal, which was the first proposal concerning the referendum ever put before the Storting, was debated in the constitutional committee in 1897 and was unanimously rejected. The referendum issue was too new and had to be weighed carefully, they argued. In brief, the amendment would have strengthened the monarchy: the king would be able to appeal a Storting resolution to the people. Documents from the Storting: Forslag 70, dok. 124, 1892- Komitéinnstilling i Inns. St. No. 150, 1897.

5. We cite from the party program of'Dansk Samling' for the period 1964 - 65: 'Under popular sovereignty (democracy) the people themselves will decide directly in as many areas as possible. Citizens will have the opportunity, through collecting signatures, to call for referenda on important questions'. Anti-party tendencies come to expression through formulations such as 'Party reign deprives the population of freedom'.

6. The minority in the Constitution Committee which voted for the referendum consisted of Finsland (Labour), Norløff (Cons.) and Thallaug (Cons.) There was dissention when the Labour party representative also proposed popular initiative, but the three stood together on a proposal for referenda on common laws 'which would give both the Storting and a group of voters the right to demand a referendum'. Document from the Storting: Indst. 5 LXXIII - 1917.

7. Documents from the Storting: Inn. S. nr. 150, 1967 - 68, Stortingstidende 7.6.1968.

8. I have read Labour party protocols from the central party committee and the parliamentary group in the period 1961-1972.

9. Protocol from the Labour Party National Conference 1965 (p. 54).

10. Arbeiderbladet (January 31, 1966). According to the report of the speech distributed by the Norwegian press service (NTB), Gerhardsen is reported to have said, "A referendum in this connection (the issue of NATO membership - T. B.) would be proper. . . ' (Adresseavisen, January 31, 1966). I believe one might best rely on rbeiderbladet in this case. Gerhardsen hardly took a definitive standpoint, but mentioned the referendum as possible alternative.

11.-4 rbeiderbladet (Newspaper) 31.1.1966.

12. Arbeiderbladet (Newspaper) 7.3.1966.

13. Berlingske Tidende (Newspaper) 7.7.1971.

14. A rbeiderbladet (Newspaper) 24.2.1962.

15. Information from Jörgen Westerståhl (November 9, 1979).

REFERENCES

Bjørklund, T. 1974. Folkeavstemning i norsk politikk - taktikk ogprinsipper. Unpublished
master thesis. Institute of Political Science, University of Oslo.

Bjørklund, T. 1977. ' Kvifor vart detfolkerøysting om EF?' Syn og Segn, 8. 498 - 506.

Bjørklund, T. and H. Martens 1979. Nok en gang: 25. september og 2. oktober bakgrunn
for datoer og folkeavstemning'. International Politikk nr. 1. 85 - 100.

Bjørnberg, A. 1939. Parlamentarismens utveckling i Norge etter 1905. Uppsala: Skrifter
utg. av Statsvetenskapliga föreningen i Uppsala 10.

Butler, D. and A. Ranney 1978. Referendums. A Comparative Study of Practice and Theory,
Washington D. C. American Enterprise Institute for Public Policy Research.

Goodhart, P. 1976. Full-Hearted Consent London: Davis-Poynter.

Holby, H. 1958. Representasjonsstyre eller folkestyre. Stortingsdebatten om folkeavstemning
1892 - 1932. Med tilleg 1932 - 1956. Unpublished master thesis. Institute of Political
Science, University of Oslo.

Koht, H. 1909. "Folkestyre' Samtiden.

Lundkvist, S. 1972. Politik, nykterhet och reformer. En studie i folkrörelsernas politiska
verksamhet 1900 - 1920. Uppsala: Almquist & Wiksell.

Mårtens, H. 1979. Danmarks ja, Norges nej, Copenhagen: Munksgaard.

Molin, B. 1969. Tjänstepensionsfrågan. En studie i svensk partipolitik, Gothenburg: Akademiförlaget.

Nielsen, H. J. 1979. Politiske holdninger og fremskridtsstemme. København: Forlaget Politiske
Studier.

Nilson, S. S. 1972. Politisk avstånd ved norske folkeavstemninger. Oslo: Gyldendal.

Seip, A. L. 1975. Vitenskap og virkelighet. Sosiale, økonomiske og politiske teorier hos T. H.
Aschehoug 1845 til 1882. Oslo: Gyldendal.

Seip, J. A. 1958. 'Teorien om det opinionsstyrte enevelde'. Historisk Tidsskrift, p.
397-463.

Siune, K. 1975..'Structure and Content of Election Campaign on Danish Radio and Television',
Scandinavian Political Studies, 10, p. 139- 155.

Steen, S. 1951. Det frie Norge, 1814. Oslo: Cappelen.

Sverdrup, J. 1972. Et statsmonopol blir til Vinmonopolet frem til 1932. Oslo: Fabritius.

Vedung, E. 1971. Unionsdebatten 1905: En jämförelse mellan argumenteringen i Sverige
och Norge, Stockholm: Almqvist & Wiksell.

Vedung, E. 1978. 'Kärnkraft, folkomröstning, regjeringshaveri'. Reaktorn 4, p. 18 - 23.
Vedung, E. 1979. Kärnkraften och regleringen Fälldins fall. Uppsala: Rabén & Sjögren.