from Referendums around the world - the growing use direct democracy, eds. David Butler and Austin Ranney, Washington D. C.: The American Enterprise Institute 1994, 69-78. Notes and references provided by the original source have been omitted.

Note that the article was published in 1994. Since then and up to now, April 2002, two more referendums have been held in Denmark, both dealing with the EU: one on The Amsterdam Treaty (28.5. 1998, 74.8% turnout, 55.1% yes, 44.9% no) and the other one on whether Denmark should join the Euro-Zone (28. 11. 2000, 86.6% turnout, 46.8% yes, 53.2% no). So, Denmark signed the Amsterdam Treaty but did - so far - keep its national currency, the Danish Krone. Moreover, referendums on joining the EU were held in Finland (16. 10. 1994), Sweden (13. 11. 1994) and Norway (28. 11. 1994). While a majority in Finland and Sweden voted in favor of the EU, a majority of the Norwegian electorate voted against joining.

Scandinavia

by Vernan Bogdanor, Professor, Fellow of Brasenose College, Oxford, UK

In the 1980s, however, both Finland and Sweden amended their Constitutions. Finland in 1987 made provision for consultative referendums, while Sweden in 1988 provided for the use of the referendum to amend the Constitution. So far, however, no referendum has been held either in Finland or in Sweden under the new procedures.

Not only has the referendum been hardly recognized in the Scandinavian countries, but when it has been used, it has not always yielded the beneficial results claimed for by its advocates. Sometimes it has resolved political problems, but on other occasions it has exacerbated them. Far from enabling public opinion to be expressed, it has on occasions been a means for the parties to manipulate it. Thus the referendum has not always succeeded in fulfilling its legitimizing function, and its use in Scandinavia remains a last resort.

Until the 1980s, referendums in the Scandinavian countries, apart from Denmark, were held on a consultative and ad hoc basis. They were held for two types of issue. The first was to mark a change of constitutional status, a new beginning: when Norway in 1905 separated from Sweden, but decided to remain a monarchy, and when Iceland in 1918 decided to end the union with Denmark, and in 1944 decided to separate entirely from Denmark and become a republic.

The European Community referendum in Norway in 1972, perhaps the most important of all in its political consequences, lies, however, slightly outside these two categories. For the decision to submit the issue to the electorate was in part constrained by the Norwegian Constitution, since Article 93 of the Constitution requires a three-fourths majority in the Storting for the transfer of powers to an international organization, and it was very doubtful if this would have been secured in 1972.

Denmark is the only Scandinavian country in which the referendum was required for constitutional change before the 1980s. The referendum was first introduced in the 1915 Constitution, Article 93 of which provided for all constitutional amendments to be submitted to the people after being passed by the then two-chamber Parliament. This became Article 94 in the Constitution of 1920, and it was adapted, with modifications, as Article 88 in the current Constitution of 1953. The 1915 and 1920 Constitutions provided for a qualifying majority of 45 percent for constitutional amendment - that is, in addition to securing a majority, at least 45 percent of the electorate had to vote Yes for the amendment to be approved. In 1953, the requirement was reduced to 40 percent, but the constitutional referendum has not been used since that date.

One of the main provisions of the 1953 referendum was the abolition of Denmark's second chamber. In exchange for abolition, the Danish Right secured the passage of Article 42, providing for a referendum on nonconstitutional legislation. Under this provision, one-third of the members of the single-chamber Folketing can demand a referendum on any item of legislation (with the exception of bills concerning finance, government loans, salaries and pensions, naturalization, expropriation, taxation, and bills discharging treaty obligations) within three days of the bill in question having been passed. For the law to be rejected, there is also a qualified majority requirement: in addition to there being a No majority, at least 30 percent of the electorate must vote No.

In addition, referendums must be held under the provisions of Article 42, with a 30 percent qualifying requirement to reject, for two specific types of bills. The first type, provided for under Article 29, alters the age qualification for the suffrage. So far, four referendums have been held under this provision, and these have progressively lowered the voting age from twenty-three in 1953 to eighteen in 1978.

The second type of bill, provided for under Article 20, transfers powers to international authorities unless there is a five-sixths majority of the Folketing, when the referendum need not be held. So far, two referendums have been held under this provision - in 1972 to ratify Denmark's entry into the European Community and in 1992 to ratify the Maastricht Treaty. The 1986 referendum that ratified the Single European Act, however, was not held under Article 20. It was a consultative and advisory referendum to which the provisions of Article 42 did not apply. A referendum of this kind is provided for under Article 19 of the Constitution, and the 1986 referendum is so far the only advisory referendum to have been held since the 1953 Constitution came into force.

lt may seem a paradox that the Single European Act, which could not have gained a majority in the Folketing, received a majority in the country, while Maastricht, which enjoyed the support of parties with 80 percent of the seats in the Folketing, was rejected by the voters in 1992. But the paradox is not difficult to explain. The Danish electorate has always been hostile to political union but fearful of being economically disadvantaged if it is excluded from EC markets. In 1986, Danish Prime Minister Poul Schluter assured voters that "the political union is stone-dead" and this assurance was accepted.

By 1992, however, it seemed that Denmark was being asked to transfer further competences to a rapidly evolving federal Europe. lt was not until they were given assurances at the European Council summit at Edinburgh in December 1992 that the Danes were prepared to endorse Maastricht in a second referendum in 1993. This second referendum was not held under Article 20 but under specific legislation passed by the Folketing in March 1993. Unlike the 1986 referendum, the 1993 referendum was to be binding. Rejection was to require, as under the Article 42 procedure, 30 percent of the electorate to vote against it as well as a majority of those voting.

One of the features of the Edinburgh Declaration was that Denmark emphasized that it could not, under the terms of its Constitution, participate in the third stage of monetary union without a further referendum. Thus the referendums on the European Community have been of value to Denmark in enabling it to define to its partners the limits of its commitments to the European Community. They were also the first indication that a vision of the future that had been widely endorsed in the chancelleries of Europe was supported much less enthusiastically by the voters.

Thus the significance of Article 42 is not to be measured solely by this single occasion when it was brought into use. Rather, it serves to emphasize and encourage the search for consensus in Danish politics. Danish governments will be under an incentive to seek the support of opposition parties so as to disarm the threat of the referendum. Indeed, the referendum was held in 1963 precisely because the normal Danish search for consensus had broken down, and the Social Democrats were determined to push through their legislation against the opposition of the parties of the Right. Thus, Article 42 prevents a Danish government, even if it enjoys a secure majority in the Folketing, from passing legislation that is unattractive both to a minority in Parliament and to the electorate. The provision gives a minority in the Folketing a strong position to challenge legislation it believes is unpopular in the country. But were the opposition in the Folketing to seek to use Article 42 in an aggressive manner, it would be endangering the convention that parties outside the government ought to be consulted on legislation before it is brought to Parliament.

The referendum in Denmark, then, has performed a valuable function. But the same cannot be said of its use in the other Scandinavian countries, where it has been employed in an ad hoc and consultative way at the behest of governments. Precisely for this reason, its two most striking political effects in Scandinavia, by contrast with many other democracies, have been to split parties and to exacerbate rather than resolve the issues it has been called on to settle.

For the Scandinavian countries are consensual democracies, in which it is usually possible to resolve issues within the parliamentary arena. Putting an issue to referendum is often a sign that the normal consensual mechanisms have broken down. The referendum may be used not for reasons of democratic principle but in the desperate hope that the people can resolve a question when the politicians have failed. It is not surprising if this hope is rarely fulfilled. And since use of referendum is unconstrained by the Constitution (except in Denmark) it can be used by the political class in its own interest. The results, however, are often unexpected. In Scandinavia, referendums have revealed cleavages that the party system has hidden, and this has been a potent force in reaking up seemingly united parties. Only in Denmark has the referendum succeeded in resolving differences within parties; only in Denmark has it performed a healing function.

In Sweden, indeed, none of the four referendums that have been held have succeeded in their supposed function of problem solving. The 1922 and 1980 referendums on prohibition and nuclear energy both failed to remove their respective issues from the political agenda; the issues in the 1955 and 1957 referendums were settled despite, not because of, the referendums. The 1955 referendum resulted in an 83 percent vote to continue driving on the left side of the road. This outcome was ignored by the politicians who, by 1967, had achieved a consensus favoring driving on the right side of the road, which was duly implemented. The 1957 referendum on supplementary benefits was so ambitious that it took a general election in 1958 and several close votes in the Folketing [Riksdag] before the issue was finally resolved.

The Swedish referendums of 1957 on supplementary benefits and of 1980 on nuclear energy, along with the Danish referendum of 1953 on changing the voting age, are the only postwar democratic referendums in Western Europe in which three alternatives were offered to the voters. In 1953, a majority favored lowering the voting age to twenty-three in Denmark, the middle alternative of the three. In 1957 and 1980, however, none of the three alternative supplementary pension plans or proposals for nuclear energy achieved an overall majority. This left considerable scope for dispute concerning the significance of the result.

The 1957 supplementary benefits referendum was the first held in Sweden, where the divisions lay between rather than within the parties. It was held because the Riksdag was deadlocked, since no proposal enjoyed the support of an overall majority. The Center, although in coalition with the Social Democrats, did not support its coalition partner's proposal, and it sought to differentiate itself from them. The Center party insisted upon including in the referendum its own separate proposal, so the electorate was confronted with three alternatives.

The 1980 referendum on nuclear power was, once again, held on an issue that divided both parties and coalition partners, cutting as it did across the normal Left/Right politics of Sweden. The three-party, non-Socialist government comprised Conservatives (broadly sympathetic to nuclear power), Liberals (moderately sympathetic), and the Center party (fundamentally hostile). A referendum seemed the only way to contain a split that threatened to break up the coalition. The opposition Social Democrats, the largest single party, became converted to the referendum, since it too was split, and it sought to keep the nuclear power issue out of the 1979 general election. Moreover, after the accident at Three Mile Island in 1979, the referendum was on the Riksdag by the pressures of public opinion.

Nevertheless, the parties were determined to retain control of the in their own hands. The referendum contained three alternatives since the Social Democrats and Liberals agreed with the Conservatives in supporting the use, for the time being, of the twelve nuclear reactors already in operation, although neither party wished to be seen publicly to align itself with the Conservatives.

The Social Democrats were in opposition: the Liberals, although in coalition with the Conservatives, found themselves under pressure on the nuclear issue from their youth wing and so thought it politic to lean to the Left in the referendum. Therefore, in effect there were two pronuclear energy options, the Conservative and the Social Democrat-Liberal, against one antinuclear option, supported by the Center and the Communist parties. Thus the referendum seemed slanted against the antinuclear option. The ambiguous outcome of the referendum allowed the pronuclear power parties, who had a majority in the Riksdag, to settle the issue. But some believed that the antinuclear option would have been victorious in a straight fight with just one pronuclear option. The 1980 referendum, therefore, has been regarded by one commentator as:

"... essentially meaningless as far as the actual issue was concerned, serving primarily as a delaying tactic.... The only lasting effect of the 1980 referendum in Sweden is liable to be the further devaluation of referenda as an institution in modern Swedish politics."

Normally, however, Scandinavia has no need to use the referendum because the Scandinavian democracies, although in form majoritarian on the British model, operate in a consensual manner. Not only are the ideological differences between the parties small as compared. with Britain or France (although they have been widening in recent years), but it is customary for government to consult opposition parties about forthcoming legislation rather than to proceed where legislation would be divisive. This is the case when there is a majority in the legislature that would wish to do so. Thus the referendum, essentially a majoritarian instrument, seems incompatible with a basic value in Scandinavian political culture: that of compromise. Only when compromise breaks down-as in Sweden over supplementary pensions and nuclear energy or in Denmark over the Single European Act - is there resort to the referendum. The referendum is often a sign of the failure of the Scandinavian political system, not of the health of democracy, as might be the case elsewhere in Europe.

There is a second reason on why the referendum is used so rarely in Scandinavia. It is incompatible with the culture of social democracy, which has been so dominant since the war. For many Social Democrats have seen the referendum as essentially a conservative weapon against progress.

This attitude was beautifully expressed by Tage Erlander, Sweden's Social Democratic prime minister, in 1948:

"Referendums go together with a different form of government than the parliamentary system. Under a coalition government of the Swiss type, no objections can be aimed at the referendum system. On the other hand, it is obvious that referendums are a strongly conservative force. It becomes much harder to pursue an effective reform policy if reactionaries are offered the opportunity to appeal to people's natural Conservatism and natural resistance to change. The enthusiasm of Conservative parties for the referendum system is thus certainly related more or less consciously to the fact that it provides an instrument for blocking a radical progressive policy."