The Storting election of 11 September 1989 resulted in great gains for the extreme parties of the left and right, the Socialist Left Party and the Progress Party, respectively. Jointly they obtained nearly one out of four votes cast in the election. The losers were the two big parties, Labor and the Conservatives, while support for the parties at the center remained stable. The long-term trend since the beginning of the 1970s has been increasing support for the two big parties, gradual decline for the parties at the center, while the two wing parties have remained small. At the Storting elections of 1981 and 1985, their joint support was only about 9 percent of the votes. Thus, the 1989 election represents a marked polarization the system.

Until 1985, Norwegian politics had been characterized by stability. But the election that year created an impasse resulting in unstable and weak governments. The three bourgeois coalition parties, the Conservatives, the Christian People's Party and the Center Party lost their majority, but continued after the election as a minority cabinet. In May 1986 the coalition was defeated and substituted by a Labor Party minority government. The latter government, maneuvering from case to case, most of the time with parliamentary support by the Socialist Left Party, the Center Party and the Christian People's Party, managed to remain in office for the rest of the election period.¹ The task of governing was aggravated by an economic recession which hit the country in 1986. On top of domestic over-consumption,falling prices and the falling exchange rate for the US dollar created huge deficits in Norway's foreign trade. A policy designed to restore the balance of the economy evidently brought some improvement, with visible costs, however. At the time of the 1989 election, the inflation rate had been clearly reduced, but unemployment was rising. An analysis of this interesting election requires a systematic investigation of recent political development, including the dynamic relationships between social structure and political change.² At this point, an attempt will be made to explain the election in the context of parliamentary instability and emerging frustrations

in the electorate. But first, a more detailed presentation of the results of
the election is required.

Election results

Table 1 presents the nationwide results for each party at the elections of 1985 and 1989. The distribution of strength between the two main blocs remained unchanged: the four bourgeois parties maintained their majority, but Labor and the Socialist Left Party jointly obtained 80 seats, only three short of a majority.

As a measure of the magnitude of electoral change we can summarize the positive values of differences between 1985 and 1989 for individual parties.³ In this case, the total figure is 15.2 percentage points, which is very high in a Norwegian context. By comparison, the corresponding figure for 1981-85 was 4.4 percentage points and for 1977-81 11.2.

The increasing support for lists competing with the established parties
may be seen as another indicator of instability or dissatisfaction in the

electorate. However, only one of these 'other' lists managed to be representedin new Storting, the Aune List, in the most northern province, Finnmark. The list, headed by the retiring province governor, called for a more active policy to improve economic conditions in Finnmark, which is badly hit owing to shrinking fish resources in coastal waters. Although the list obtained only 0.3 percent of the nationwide votes, it won a sufficient share (21.5 percent) in Finnmark to secure the election of Mr Aune.

Several parties competed in the 'green' market. A Green Party founded in 1988 obtained only 0.4 percent of the overall vote. Equally unsuccessful was an election alliance, called 'The province lists for environment and solidarity' formed by the two tiny communist parties. The alliance obtained 0.8 percent of the votes on a nationwide basis, exactly the same as the joint share for the two parties in 1985.

The impact of 'green' ideas may, however, have been greater among other parties. In their electoral platforms, all parties expressed concern about environmental problems. Most articulate in this regard were the Socialist Left and the Liberal Party. As Table 1 clearly indicates, the former party enjoyed an impressive victory. The Liberals on the other hand, gathered increasing support during the spring and summer of 1989, and according to opinion polls the party was likely to gain some 6 percent of the electoral vote. But on election day it polled only 3 percentage points, in spite of its devoted campaign in favor of new environment policies. Indeed it is puzzling how this issue may have affected the electoral outcome for the two environment parties (cf. Aardal 1990).

Finally, it should be observed that electoral turnout reached 83.2, only slightly lower than at the preceding elections. This result was rather unexpected the fact that the political alternatives facing the electorate far from clear.

Political Alternatives

For a number of elections, beginning in 1965, Norwegian electors had been facing two competing government alternatives: either a bourgeois coalition, or a minority government of labor, based upon a joint socialist majority (i.e. Labor and the Socialist Left Party). These alternatives, which fitted the main division along the left-right axis, managed to provide the country with viable governments. At the 1985 election, these alternatives still prevailed, but failed to produce a workable majority. The two socialist parties obtained 77 seats in the Storting, the three coalition parties 78, and the Progress Party 2 seats. Thus the latter party arrived in a balancing position, which destabilized the system. In May 1986, the Progress Party joined the socialist parties in defeating the coalition government. For the

rest of the election period, the country was run by a Labor minority
government, although there was a bourgeois majority in the Storting.

How should this paradoxical situation be explained? It might look like a reasonable solution for the coalition parties to broaden their parliamentary basis by including the Progress Party, but such a solution was never considered. Because of its populist character, the Progress Party differs in style and behavior from all other parties. Ideologically, the Progress Party is liberalist and reflects the 'New Right' in Norwegian politics. It may be argued that the three coalition parties are unable to cooperate with the Progress Party owing to big political distances in relation to the latter party. A recent analysis (Valen 1990) suggests that mutual distances between bourgeois parties not particularly big along the left-right axis, but the Progress Party and the two parties at the center are widely apart along other traditional cleavages. More specifically, the Progress Party and the Christian People's Party constitute opposite poles along the moral-religious division, and the Progress Party and the Center (Agrarian) Party differ sharply along the urban-rural dimension which is highly salient in current politics.

The unstable parliamentary situation, together with the economic recession, nurtured dissatisfaction and protest reactions in the electorate (Aardal & Valen 1989, 276-285). Such sentiments were manifest in the local and province elections of 1987, which the Progress Party won by a landslide (Bjørklund 1988). During the subsequent two years, numerous opinion polls confirmed that the Progress Party had become the third biggest party in the country, although its strength fluctuated considerably over time. In June 1989 the polls indicated a support of more than 20 percentage points.

In the summer of 1989, the three former coalition parties, together with the Liberals, declared that they were ready to take over government power after the election. But the increased support for the Progress Party implied that the two traditional government alternatives had lost their validity. According to opinion polls, a socialist majority was unlikely, and the probability of a majority in favor of the coalition parties was even lower. Thus the outlook was for a continuing deadlock situation in the competition between the two major party groupings, while the Progress Party was likely to return with a much stronger representation than in the period 1985-89.

In this situation the individual voter did not have a choice between fixed political alternatives. A vote for a given party did not imply support for some specific government alternative. Which kind of government did the electors prefer? An answer may be found in Gallup polls conducted monthly from January to August 1989.⁴ The respondents were asked two questions: (1) 'Which of the following two alternative governments would you prefer after the Storting election this autumn: a Labor Party government or a non-socialist government without the support of the Progress Party?'; and

(2) 'Which one would you have chosen from the following two alternatives: a Labor Party government or a non-socialist government supported by the Progress Party?' The first of these questions produced a slight predominance in favor of a Labor Party government, approximately 43 percent compared with 40 percent for a non-socialist government without the support of the Progress Party. Some 17 percent refused to take a stand. Surprisingly enough, the overall distributions of these answers remained quite stable from January through August. When the choice was between a Labor government and a bourgeois government with the support of the Progress Party, nearly 50 percent indicated a preference for a Labor government, while only 33 percent of the respondents preferred the bourgeois alternative.In instance too, opinion remained stable over time.

My main concern is to study reactions within the different parties. For this purpose all the monthly polls have been lumped together and the results are presented in Figure 1. Naturally, the desire for a change of government was least marked in the socialist parties, and here the condition of Progress Party support did not matter at all. It is equally natural to find that the voters of the Progress Party were most interested in a bourgeois government relying on the support of their own party. On the other hand, in the other four bourgeois parties, there was a solid majority in favor of bourgeois government, providingthat government was quite independent of the Progress Party. Support was much weaker if the government should have to base itself upon the support of the Progress Party. The differences between the two conditions for forming the government were least among Conservative Party voters, whose desire for a change in government was strong under all circumstances. In all the parties at the center, however, we find a definite majority saying 'no' to a bourgeois government, if the condition is that it should be based on the support of the Progress Party. The differences were most marked in the

Center Party and in the Liberal Party. This discrepancy is consistent with the
former observation of large distances between the Progress Party and the
parties at the center.

Alternatives and Turnout

In a well-known article, Angus Campbell (Campbell et al. 1966) has argued that there is a direct relationship between electoral turnout and clarity of political alternatives facing the electorate. His thesis is that the clearer the alternatives, the more intense is the election, and the higher is the turnout level. Data from Norwegian elections have tended to support this hypothesis & Martinussen 1972). Thus, in the 1987 local and province elections, which occurred in a situation of parliamentary instability and lack of a coherent opposition, the turnout was a record low (Bjørklund 1988). Owing to unclear alternatives, a similar result might be expected in 1989. It did not happen. As Table 1 indicates, the turnout level in 1989 was almost as high as in 1985, when government alternatives were clear. In this respect, the 1989 election is a deviating case.

It should be observed, however, that in spite of confusing alternatives the intensity of the election was high. In the election survey which was carried out immediately after the election, the respondents were asked how they perceived the situation and the intensity of the campaign.⁵ Without entering into a detailed analysis, the major tendencies are outlined. The proportion indicating that it was easy to make a decision was 35 percent of the sample in 1989, compared to 50 percent in both 1981 and 1985. Similarly, the proportions indicating that they cared a lot about the outcome of the election dropped from 46 percent in 1985 to 39 percent in 1989. Presumably, these figures reflect the lower clarity of alternatives in 1989. But the picture is entirely different when we consider the frequency of political discussions, which is likely to express campaign intensity. In 1989 the proportion indicating that they had been discussing politics daily was 40 percent. In 1981 and 1985 the corresponding figure was 28 percent. The data suggest that the intensity of the 1989 election was higher than normal, which is likely to have contributed to the overall participation level of the electorate. The data suggest another significant conclusion: high electoral intensity is not necessarily a result of clear-cut political alternatives. But then we are left with a new question: what made the 1989 Storting election so exciting to the electorate? A brief description of the campaign may elucidate this question.

The Election Campaign

In the election study, respondents were asked: 'Let us look back at the

election this autumn. Would you please mention one or two issues which were particularly important to your voting decision?'. The results presented in Table 2, include only voters who were able to mention one or two issues, altogether some 80 percent of the sample.

In the present paper, I am not concerned with the question of how issue position may have affected the electoral choice of individual voters. The main purpose is to describe the salient issues at the election. Nearly half of the respondents who mentioned some issue were concerned with social welfare problems, such as health care, insurance and care of the aged, slightly less than at the 1985 election. Environmental problems were mentioned by more than one-third of the respondents, which is a strong increase compared with 1985. As might be expected, unemployment problemsattained significance, while defense and foreign policy issues declined in saliency. Immigration and criminality were new issues in 1989, but actually the concern expressed by the respondents was lower than one would expect on the background of public debate. Surprisingly enough, nobody mentioned the question of who should form the government after this election. This topic was mentioned by 5 percent in the 1985 study.

Presumably the difference reflects the absence of clear-cut government
alternatives in 1989.

A characteristic aspect of the party system is the distinctiveness of individual parties in their position towards political issues (Aardal & Valen 1989). Table 3 demonstrates that a similar distinctiveness is also evident in the electorate's perception of issues. Questions concerning social welfare are most frequently mentioned by voters of the Labor Party, with the Progress Party in second place. Voters of the Liberal and Socialist Left Parties are most concerned with issues of energy and the environment. All party groups are concerned with unemployment, but Laborites and Conservatives more than others. Foreign policy has been mentioned most frequently by supporters of the Center Party, with the Conservatives in second place. The particular issue referred to is Norway's relationship with the European Communities. The Conservative Party is clearly in favor of joining the EC while the Center Party takes the opposite position. As usual, taxation is largely a concern for the parties of the right. Finally, Table 3 indicates that decentralization is a favorite issue of Center Party voters, and so are moral and religious issues for the Christian People's Party, while the immigration issue largely interests voters of the Progress Party.

It is not surprising that social welfare issues are seen as the most important theme in the 1989 campaign (Table 2). Early in the campaign the leader of the Progress Party launched an attack on the abuses of the welfare system, particularly social insurances. All other parties, but most notably the parties of the left, expressed their disagreement. The debate illuminated inconsistencies between the basic ideas of the welfare state and economic liberalistic ideology. In addition to welfare state ideology, the debate focused upon the shortcomings of welfare arrangements, particularly in the areas of health and care for the aged. Presumably, it was the debate on

welfare issues which created the unusually high intensity in the campaign.
This debate apparently overshadowed the second most important theme
of the campaign, energy and the environment.

As usual, the election campaign lasted for only about three weeks. Apparently, the impact of the campaign was greater than normal. Table 1 indicates that overall changes at the election were substantial. An investigation individual shifts among voters also confirms that volatility was unusually high. A panel study in which some 800 voters were interviewed, both in 1985 and 1989, indicates that 39 percent of the respondents have shifted position between the two elections, either by shifting back and forth between different parties, or by shifting between voting and non-voting. Similarly, the proportion of voters indicating that they arrived at their vote decision some time during the campaign was 42 percent in 1989.⁶ The corresponding figure was 21 percent in both 1981 and 1985.

The impact of the campaign can also be directly measured by following opinion polls during the 3-4 weeks before the election. Two parties lost the campaign, the Progress Party and the Liberals. The former party decreased from around 21 percent of the vote in the summer to 13 percent on election day. During the same period, the Liberals dropped from nearly 7 to 3.2 percent. The Socialist Left Party increased its support most substantially during the campaign, but Labor and the Conservatives improved their support.

Regional Variations

So far we have been concerned with nationwide tendencies in the election. Since there is a long tradition of regional contrasts in Norwegian politics (Rokkan & Valen 1964) a pertinent question is: to what extent are the general tendencies reflected throughout the territory? Can territorial variations to our understanding of the election result? This question may be approached by the application of aggregate data for the 448 local communes. Table 4, which is based on electoral statistics, is limited to the four parties which were strongly affected by the election winds in 1989. The country has been divided into five regions consistent with previous research (Rokkan & Valen 1964).⁷ Table 4 confirms traditional patterns. Both socialist parties have their weakest support in the South and West. Otherwise support for the Socialist Left does not vary much from one region to another. Labor on the other hand, enjoys its strongest support in the Interior East and in the North. Both parties of the right have traditionally had their strongholds in the Oslofjord Area, but in later years they have greatly improved their position in the South and West. Throughout the country the two latter parties enjoy far more support in urban than in rural areas (Aardal & Valen 1989, 223-228).

But for our present purpose, changes from 1985 to 1989 are the most interesting. Naturally, gains and losses have to be considered in relation to the previous strength of the respective parties. In general, Table 4 confirms that the main national tendencies (see Table 1) are evident throughout the country. But some regional deviations are observed. The Conservative Party has lost slightly more in the Interior East than in other regions. The decline in Labor support is considerably greater in the North than in other parts of the country. At the same time, the gains of the Socialist Left are relatively small in the Northern region. The Progress Party has the most interesting profile. Until 1985 the party's support was largely limited to the Oslofjord Area and the Southwest. In 1989 it managed to establish itself throughout the country, but the gains were relatively biggest in the North and in the Interior East.

The most remarkable changes occurred in the northern region. The Aune list in the province of Finnmark partly accounts for the turmoil. But even apart from this list, the changes in vote distributions are sizeable. The main result is that the joint socialist strength has been reduced by more than 7 percentage points compared to 1985, while the Progress Party has gained a foothold in this peripheral region. Big electoral changes in peripheral areas were evident already in 1985, most notably in the North and in coastal areas of Trøndelag and the West, and electoral volatility coincided with unemployment and other indicators of social and economic distress (Aardal & Valen 1989). In 1985, Labor, the leading opposition party, profited most from the electoral changes. In 1989 economic recession was most evident in the northern provinces. Now the Labor party, which had been in governmental office for more than three years, was challenged from the right.

In an attempt to explain territorial variations, the communes have been
classified by the amount of unemployment in 1989 and by the increase

in the unemployment rate from 1987 to 1989. When all communes are considered, electoral change does not seem to be directly related to unemployment.But the Northern region is considered separately, the picture is different. The size of unemployment in 1989 does not seem to be a good predictor of electoral change. But as indicated in Table 5, electoral results in local communes coincide with the increase in unemployment. For the governing Labor Party the tendency is clear: the higher the rate of unemployment, the bigger the drop in the Labor vote. For other parties tendencies are less clear. It should be observed, however, that the joint gains for the two extreme parties are biggest in communes with middle or high increases in unemployment. Unfortunately, figures for the Aune list, which is limited to one single constituency, cannot easily be fitted into the table. But as far as Finnmark is concerned, the Aune list has been particularlyattractive communes with rising unemployment. Table 5 does suggest that unemployment has affected the election result, but a more thorough analysis is required, taking into account social structure as well as other indicators of economic distress in local communes.

Another topic which has probably contributed to electoral changes in recent years is immigration. At the local and province elections of 1987 the immigration issue attracted a lot of attention, and apparently contributed to the gains of the Progress Party (Bjørklund 1988). At the 1989 election the issue seems to have been less salient (see Tables 2 and 3). Nonetheless, it is worthwhile relating electoral change to number of immigrants in local communities. According to Table 6, the gains of the Progress Party, which has signalled the most restrictive position towards immigration (see Table 3), are significantly higher in communes with high rather than low immigration The tendencies are less clear for other parties, except that the Conservatives have lost most in communes with high immigration rates. In further analyses, immigration has to be explored in a multivariate perspective. Nonetheless, the main tendencies reflected in Table 6 are likely to prevail also when control is established for social and economic structure.

Concluding Remarks

This initial analysis reveals some important tendencies in the 1988 Storting
election:

Electoral volatility was unusually high.

In spite of unclear government alternatives, the intensity of the campaign
was high.

A polarization occurred in favor of the extreme parties of the left and
right.

For the populist Progress Party, the election resulted in a breakthrough
in peripheral parts of the country, most notably in the Northern region.

Further analyses will focus upon the question: why did the 1989 election deviate so much from previous Storting elections? At this point some tentativeexplanations be suggested. The polarization in favor of the ProgressParty the Socialist Left Party constitutes an important topic. Why did it occur? For a possible explanation we may look first at a long-standing tendency among the major parties to move in the direction of the political center, which has apparently caused frustration in the political wings. There is a near consensus among the established parties concerning some important policy areas, such as social welfare policy and foreign policy. Actually, the political distances among Norwegian parties are rather small (Valen 1990). The politics of coalitions has made it necessary for the two big parties to cooperate closely with the parties at the center. This was clearly demonstratedby bourgeois coalition of 1965-71, which was headed by the leader of the Center Party, Mr Borten. Important social reforms, like the general pension scheme, were introduced by this government. Public expenses as well as taxes increased substantially during these years. The emergence of a right-wing populist party in 1973 may be seen as a reaction against this policy. Characteristically enough, the original party label was 'Anders Lange's Party for a strong reduction in taxes and public expenses'. At the death of its founder in 1977, the party label was changed to the Progress Party. Frustrationsin

trationsinthe right-wing were also evident during the coalition of the 1980s, although this government was headed by the Conservative leader, Mr Willoch. Government policies had to express compromises between the platformsof Conservative party and the two parties at the center.

A reverse tendency occurred during the Labor minority government from 1966 to 1969, headed by Mrs Brundtland. Since the government had to seek parliamentary support at the center, policies had to be, to some extent, adapted to the demands of the Center Party and the Christian People's Party. At the election, the Socialist Left Party profited from leftwing caused by these policies. The 1989 election result may be interpreted as a call for stronger confrontations along the left-right axis.

At the same time, however, frustrations and protest have become a dominant aspect of Norwegian politics during the latter half of the 1980s. 'Contempt for politidams' has become a household expression in public debate. This tendency may be explained partly as a result of the economic recession after the oil boom, but possibly the decline of parliamentary stability is a more pertinent explanation. Parties and politicians seemed to be unable to formulate constructive policies in response to the problems facing the country. In this situation the extreme parties of the left and right permitted themselves to be rather outspoken since they had not been directly involved in decision-making on government level.

However, a fuller understanding of the election requires a more thorough analysis of voter reactions towards specific issues in the campaign. The reactions against established parties in peripheral regions suggest that the economic recession created protest attitudes which benefited the Progress Party. Reactions against immigration policies seem to have worked in the same direction. It is a puzzle, however, that support for the Progress Party declined so sharply during the campaign. Several forces may have been at work. But the most likely explanation is found in the intense debate on social welfare policies. The Progress Party made an assault on the basic principles of the welfare state, to which all other parties subscribe. Particularly, parties of the left argued strongly in defense of the welfare system. It is likely that this debate mobilized the electorate in favor of the Socialist Left Party. The Labor Party, however, did not profit much from this confrontation, probably owing to the fact that the party had run the government in a period of increasing unemployment.

---

NOTES

1. Observe that Storting elections occur regularly every fourth year. The Constitution does not provide for a dissolution of the Storting between elections.

2. A nationwide voter survey, which was conducted immediately after the 1989 election, is currently being analyzed. This study is part of a long-standing program of electoral research at the Institute for Social Research in Oslo. The program was initiated at the Storting election of 1957 by the late Professor Stein Rokkan and Henry Valen. Responsible researchers at recent elections have been Henry Valen and Bernt Aardal. On this program, see H. Valen & S. Rokkan, 1967. 'The Norwegian program of electoral research', Scandinavian Political Studies 2, 294-305; S. Rokkan & S. Kuhnle, 1977. 'Political research in Norway 1960-1975', Scandinavian Political Studies 12, 127-156; H. Valen, 1981. 'Electoral Research in Norway', in: Research in Norway. Universitetsforlaget: Norwegian Research Council.

---

3. This measure was applied by Mogens Pedersen (1983) in his well-known article on electoral volatility in European party systems.

4. The polis which were conducted by the Gallup Institute on behalf of the Norwegian Broadcasting Corporation (NRK) were basedon nationwide voter samples of some 1000 persons each time.

5. Three questions are particularly relevant: (1) 'Some people say it was easy to decide for which party one should vote at this election, while others think it was very difficult. How do you see it, did you find it very easy to decide, quite easy, quite difficult or very difficult?' (2) 'Looking back at the election this fall, would you say that you personally cared a lot about which party or parties lost or won, did you care somewhat, or didn't it matter for you personally?' (3) 'We should like to hear how much you participated in political discussions or conversations before the election. Would you say daily, about twice a week, more seldom, or never?'

6. The respondents were asked: 'When did you decide to vote for the party of your choice, was it long before the campaign started, was it some time during the campaign, or was it on election day or immediately before?' A total of 58 percent indicated that they decided long before the campaign, 18 percent during the campaign, and 24 percent around election day.

7. Observe, however, that in previous analyses, classifications of regions have followed province borderlines, although this is not always strictly correct. In the present analysis, the coastal areas of the provinces of Buskerud and Telemark are classified as part of the Oslofjord-area, while the rest is seen as part of the Interior East.

REFERENCES

Bjørklund, T. 1988. 'The 1987 Norwegian Local Elections: a Protest Election with a Swing
to the Right', Scandinavian Political Studies 11 (3), 211-234.

Campbell, A., Converse, P. E., Miller, W. E. & Stokes, D. E. 1966. Elections and the Political
Order. New York: Wiley.

Pedersen, M. 1983. Changing Patterns of Electoral Volatility in European Party Systems', in
Daalder, H. & Mair, P. eds. Western European Party Systems. Beverly Hills: Sage.

Rokkan, S. & Valen, H. 1964. 'Regional Contrasts in Norwegian Polities', in Allardt, E. &
Littunen, Y. eds., Cleavages, Ideologies and Party Systems. Helsinki: Westermarck Society.

Valen, H. & Rokkan, S. 1967. 'The Norwegian Program of Electoral Research', Scandinavian
Political Studies 2, 294-305.

Rokkan, S. & Kuhnle, S. 1977. 'Political Research in Norway', Scandinavian Political Studies
12, 127-156.

Valen, H. & Martinussen, W. 1972. Velgere og politiske frontlinjer (Voters and political
frontlines). Oslo: Gyldendal.

Valen, H. 1981. 'Electoral Research in Norway', Research in Norway. Oslo: Universitetsforlaget.

Valen, H. 1990. 'Coalitions and Political Distances' in Sånkiaho et al. People and their Politics
(Festschrift for Pertti Pesonen). Helsinki: The Finnish Political Science Association.

Aardal, B. & Valen, H. 1989. Velgere, partier og politisk avstand (Voters, parties and political
distances). Oslo: Central Bureau of Statistics.

Aardal, 8., 1990. 'Green Polities: A Norwegian Experience', Scandinavian Political Studies
13 (2), 147-164.