Although local elections in Norway are predominantly concerned with local issues, they happen to be of considerable interest for national politics as well. This was especially true for the elections of September 1971, in which members of local councils were elected for another four-year period. The; elections occurred hi the middle of important national events, and national issues were brought prominently into the campaign.

Our main concern on the subsequent pages will be the problem of how the reactions of the voters may be related to major political events. First, we shall describe briefly the political development during the period preceding the 1971 elections, with particular emphasis upon the emergence of the Common Market issue. Secondly, the election results will be analysed on the ba.sis of aggregate data from electoral statistics. Thirdly, we shall report briefly on a secondary analysis of data from Gallup polls during the period 1969-1971.¹ Finally, we shall discuss the findings from the empirical analysis. Of course, in this brief review article the analysis has to be superficial, and a number of interesting problems have to be followed up in later work.

1. The Political Situation

Since 1970 the dominating issue in national politics has been the question of Norway's entry into the Common Market (EEC). In June 1970 the Storting decided to open negotiations for full membership, and this decision triggered off a heated public debate. Within the Bourgeois (or non-socialist) coalition government, which had been in office since 1965, the EEC issue created serious controversies. The position of the government had been weakened at the Storting elections of 1969, which reduced the representation of the four coalition parties to 76 out of a total of 150 members, while the leading opposition party, Labour, returned with 74 members.²

Formally a unanimous government supported the application for Norwegian EEC membership, but basic disagreements between the coalition parties soon became evident. Consistent with their position in 1961-1963, when the EEC issue was first debated, the Prime Minister, Mr. Per Borten, and his Centre Party indicated a reluctant position toward the negotiations. The Conservative Party, on the other hand, was the most ardent supporter of EEC membership.

In February 1971 the growing tensions between the coalition parties resulted in a cabinet crisis. The key event, which triggered the crisis, was an indiscretion by the Prime Minister. It became known that certain leaders of the well-established anti-EEC opposition had seen a confidential diplomatic report on the negotiations. Prime Minister Borten first denied, but later had to admit that he was the source of the leakage. The other coalition parties, Liberals, Conservatives and Christians, demanded that the Prime Minister resign. When the Centre Party refused to abandon Mr. Borten, a cabinet crisis was unavoidable. Attempts to form a new Bourgeois coalition proved to be unsuccessful, and the Labour leader, Mr. Trygve Bratteli, was eventually asked to form a minority government.

The cabinet change crystallized the debate on the EEC issue. The top Labour leaders held long-standing commitments toward the idea of EEC membership. They now took over responsibility for the membership negotiations and conducted them with greater vigour than the former coalition government had been able to do. Furthermore, the

* I am grateful to Dr. William Lafferty and to Mr. Ole Berg, who read the manuscript
and made valuable suggestions, but I assume full responsibilty for the contents.

Centre Party announced a new policy; they advocated a break in the negotiations for
Norwegian membership and instead pushed for a special trade agreement with the EEC.

The EEC issue will be the main focus in the subsequent analysis.

2. The Elections of September 1971

Despite the fact that a heated public debate was going on at the time of the election, voter participation turned out to be surprisingly low with only 73 per cent of the eligible voters turning out. At recent elections the turnout level has been much higher: at the local elections of 1967 it was 77 per cent and in 1963 it reached 81 per cent.

It has been convincingly demonstrated by students of electoral behaviour that high clarity of political alternatives and high campaign intensity tend to mobilize the electorate to high participation." The low turnout level at the 1971 elections seems to contradict this theory. We shall argue that the low turnout in 1971 was due to the special circumstances at this particular election; the public debate did not refer directly to the election, the parties were internally split over the EEC issue, and this created a great many cross-pressures, which led a great number of voters to abstain from participation. We shall return to this problem.

Table I presents the distribution of the vote in 1971 compared with the preceding
local elections (1967) and Storting election (1969).

A comparison, election by election, indicates trends in recent political developments, but one should be careful in interpreting the differences since, on the local level, list alternatives may differ from one election to another. Particularly, this is important to have in mind when we compare local elections with Storting elections. In the latter elections all national parties appear with separate lists all over the country with only minor exceptions.⁴ Local and non-partisan lists occur only at local elections, as do various joint local lists between several parties. Consequently, at local elections the proportion of votes cast for these kinds of lists reduces the strength of national parties compared with their positions at Storting elections.⁵ A direct comparison between two local elections may also be slightly misleading, because in quite a few communes the character and number of lists may differ from one local election to another.

The main results of the election may be summarized on the basis of Table I:

1. The Centre Party and the left-socialist parties, which run joint lists in a number of communes, increased their proportion of the vote, while Labour, the Liberals and the Conservatives declined. These tendencies are apparent, with varying strength, whether we compare 1971 with the local elections of 1967 or with the Storting election of 1969. For the Labour Party the defeat was particularly severe when the Storting election forms the basis of comparison. The Christian People's Party is in a middle position; the party enjoyed a net gain compared with the preceding local elections, but its proportion of the vote decreased slightly between 1969-1971. This latter decrease would disappear, however, if we take into account the local and nonpartisan votes in 1971, of which the party may certainly claim its share.⁶ Thus, the Christian People's Party improved its position between 1967-1969, but remained fairly stable between the two later elections.

In general, the local elections of 1971 resulted in victory for the anti-EEC parties, i.e., the Centre Party, and the left-socialist parties, and defeat for the pro-EEC parties, Labour and the Conservatives. The Christians and the Liberals, who have been less identified with the EEC issue, by and large maintained their position of 1969.

2. The proportion of the votes for local and non-partisan lists which had been consistently declining at a number of elections, suddenly increased in 1971.⁷ Traditionally these lists have been concerned with territorial controversies within individual communes and with specific economic or social interests like teetotalism or occupational group interests. At the 1971 election a new type of non-partisan lists appeared in addition to the older ones; lists advocating anti-pollution, decentralization of democratic decision-making, and women's liberation (i.e., pure lists of women). Apparently the slight increase in the non-partisan vote was due to these new lists.

3. The number of votes for joint Bourgeois lists has been declining rapidly in recent years, and this tendency continued at the 1971 election.⁸ It is more remarkable, however, that joint lists between socialist parties appeared - for the first time. The local alliances between Communists and left-socialists reflect the fact that due to the EEC issue the distance between these parties has decreased.⁸

Table I presents overall figures for the total electorate. Our next question is, how do the election results differ from one commune to another according to region and social structure? In this analysis, which is based on data from the official election statistics, the 1971 elections will be compared with only the preceding local elections.

By and large, the main tendencies on the national level are reflected in each of the regions. However, some regional variations are apparent. Thus the greatest gains for the Christian People's Party and the greatest losses for the Liberals occurred in the South. The Centre Party improved its position more in the North than in other regions. The Labour vote also varied from region to region. In the interior East the party slightly improved on its 1967 position. In all other regions it lost ground, although more in the West where its share of the vote declined by 3 per cent.

However, the regional variations do not reflect systematic tendencies and, consequently,
they are difficult to interpret. The patterns are far more regular when party
distributions are related to urbanization of communes.

Some main tendencies are evident in Table II:

1. The decline in turnout is greater in the more urbanized communes. In cities and towns, vote participation in 1971 was about 5 per cent lower than at the preceding elections. In the most extreme rural communes, i.e., sparsely populated fishing communes, the turnout in fact increased between 1967 and 1971.

2. The losses of the Labour and Conservative parties are substantially higher in urban
than in rural communes. The relationship is however, not a linear one; both parties
suffered heavy losses in fishing communes.

3. The Centre Party has improved its position most in fishing communes. In all other
types of communes its proportion of the vote increased by about 2 per cent.

4. Communists and left-wing socialists have improved their position in urban areas, but lost in sparsely populated communes. Part of the explanation for the latter tendency is that both parties in 1971 abstained from running lists in a number of rural communes in which they participated in 1967.

5. The Liberal Party lost more in rural than in urban communes. This trend was the
opposite for the Christian People's Party, however, which gained most in densely
populated communes.

By and large, the tendencies observed in Table II prevail hi all regions, but some
notable variations do occur. The losses of the Labour Party in urban areas are particularly
large in the South-West region. Both for densely and for sparsely populated areas the

Christian People's Party has improved its position more in the South-West than in other
regions. The Centre Party has made its greatest gains in the Middle and Northern regions,
irrespective of type of community.

In this brief commentary it is not possible to go into a detailed statistical analysis.
Furthermore, only aggregate dai:a are available, which do not permit us to study changes
on an individual level.

Thus, on the basis of aggregate data (Table II) it is reasonable to assume that changes in turnout are related to changes in vote distribution, but only information for individual voters can describe the character of such relationships. However, a secondary analysis of Gallup poll data during the period before the election may give some ideas of the processes of change in the ebctorate.

3. Changes in Party Preferences 1969-1971

Every month the Norwegian Gallup Institute interviews some 1500 persons drawn on the basis of nationwide, stratifi.ed probability sampling. Two standard questions from these polls are pertinent to our analysis: 1) 'How did you vote at the last Storting election?' and 2) 'Which party would you vote for in case there should be an election tomorrow?' By cross-tabulating the responses on these two questions it has been possible to study the shifts in party preferences between the 1969 Storting election and the time of interviewing.

Gallup samples for several months have been pooled for three periods according to
major political events:

1. October 1969-June 1970. The coalition government in office. Decision on EEC
negotiations at the end of June 1970.

2. July 1970—February 1971. The EEC debate had started. Increasing controversies
within the coalition. Cabinet crisis at the end of February 1971.

3. March-August 1971. The Labour Party in power, minority government. The debate
on EEC became more crystallized.

During the first two periods. Gallup polls indicated by and large a consistent tendency from month to month. In the third period, however, the trend which started in March 1971 was broken after only two months, and a new development set in. This period has therefore been divided into two subgroups: (a) March-April 1971 and (b) May-August 1971.

First we may consider the degree of stability in the electorate. Table 111 presents the
proportion of respondents who indicate consistency in party preference between the
1969 election and the time of the interview.10

Table 111 suggests a remarkable stability in the electorate. Due to the recall method which has been applied, we may assume that the data are somewhat biased in the direction of higher stability.¹¹ Nonetheless, by comparing different periods, an interesting tendency emerges. During the first period after the Storting election, very few respondents report that they have shifted party. After June 1970, when negotiations

for EEC membership started, the number of shifts increased significantly, and it reached its maximum in the third period after the fall of the coalition government. Thus the frequency of individual changes in party preferences was clearly related to major political events on the government level.

The tendency in Table 111 becomes even more revealing when we consider the patterns
of change between different parties from one period to another.12 These changes
may be briefly described:

1. During the period October 1969 to June 1970, only minor changes occurred in the aggregate distributions of party preferences. The most important change was a loss for the Liberals of approximately 1.5 per cent, while the Labour Party gained a similar proportion.

2. During the subsequent period when the coalition government was in a process of disintegration. Labour improved its position by another two percentage points, while all the coalition parties, except the Centre Party, lost support. The Socialist People's Party, which is an outspoken opponent of the EEC membership, improved its position, but by less than 1 per cent. These figures suggest that during the initial period of membership negotiations, the EEC issue had a rather limited direct impact upon the distribution of party preferences. The Centre and Socialist People's Parties, both anti-EEC, gained less than Labour which was generally perceived to be pro-EEC. Indirectly, however, the EEC issue loomed large in the competition for votes. The leading opposition party, Labour, profited from the fact that the coalition was coming apart at the seams, but the single issue which caused most of the troubles for the coalition was the question of Norway's entry into the EEC.

3. The first two months after Labour had taken over governmental power, the party continued improving its position in the opinion polls. It reached its peak in March 1971 with 50.4 per cent of the total vote, the highest figure ever obtained by a single party since Gallup polls started in Norway in the beginning of the 1950'5.

On the Bourgeois side, substantial changes occurred. The Centre Party improved its position by 2 per cent, the Liberals and the Conservatives lost in popularity, while the Christian People's Party slightly increased its vote. In a similar way, the Socialist People's Party enjoyed a slight increase.

4. In May 1971, the tide turned for the Labour Party; from then on the party suffered a decline in support from over 50 per cent in M arch-April to 42.5 per cent in September. The Socialist People's Party enjoyed the greatest gains. On the Bourgeois side, Liberals and Conservatives regained some of their lost support.

The Gallup poll of May 1971 not only signalled a new tendency in party distributions, it also indicated a sudden rise in the proportion of non-voters. This proportion remained relatively large throughout the summer of 1971 after having been consistently low since 1969.

Both major tendencies which appeared in May 1971 pointed towards the election
results four months later. Consequently, this period is of particular interest for our analysis.

Of course, changes in aggregate party distributions do not give much information about the actual shifts which have been going on in the electorate. The available Gallup data are not very precise on shifts between voting and non-voting. It may be useful, however, to look into shifts among parties by active voters, i.e. respondents who report both party choice in the 1969 election and party preference at the time of the interview. Table IV presents the transition matrix for pooled Gallup data May-August 1971.

By comparing the marginals of Table IV we obtain differences in party distributions
between the Storting election of 1969 and party preferences May-August 1971. In size,

these differences deviate from the actual changes in distributions between the 1969 Storting election and the 1971 local election (cf. Table I). This is not surprising since Table IV does not refer to the actual election situation in 1971, but merely presents party preferences over a four-month period in which gradual changes were taking place. Moreover, Table IV includes only active voters, while changes in total distributions between two elections are due to several forces; shifts between parties, shifts between voting and non-voting, changes in the composition of the electorate due to mortality (death and emigration), and young people coming of voting age.¹³

None the less, the direction of changes for individual parties in Table IV fits with the
general tendencies between the two elections (Table I). This consistency supports the
possibility that the tendencies observed in the data reflect actual shifts in the electorate.

Table IV indicates a substantial number of shifts between Labour and the Socialist People's Party, largely in favour of the latter party. Further, Labour has a net gain in its exchanges with Liberals and Conservatives, but has lost to the Christian People's and the Centre Parties. In the Bourgeois camp all other parties have lost to the Centre Party.

Why did these new tendencies set in a few months before the local elections? Of course, a number of forces could conceivably affect the party preferences, e.g. local or regional issues generated by the impending elections, and delayed reactions toward the national parties due to the cabinet crisis and the general parliamentary situation. We do not have data at hand to test hypotheses of this character, but we can draw attention to one single issue, the question of EEC membership, which is most likely to have had a decisive impact.

Among Norwegian voters EEC membership has always been unpopular. As an example, we may look at the results from a nationwide Gallup poll in February 1971. The respondents were asked: 'This is a card with different alternatives concerning Norway's relation to the European Common Market. Which one do you prefer: (1) Full membership, (2) Association, trade agreement, mutual reduction of customs, etc., (3) No special arrangement, or (4) No opinion?'

The poll resulted in the following percentage distribution:

The proportion of people with 'no opinion' has been consistently high in all polls on the EEC question. Furthermore, the distribution on various alternatives has been remarkably stable for long periods. From 1961, when the EEC issue was first debated, until the end of 1970, repeated polls showed that 22-29 per cent favoured full membership in the EEC. Then this proportion suddenly dropped to about half. The new pattern, of which the poll of February 1971 is an example, remained stable throughout 1971." A slight variation in favour of the EEC occurred in January 1972.

Attitudes towards the EEC vary substantially within the population: favourable attitudes occur more frequently among people of high than of low social status, more frequently among people in urban than in rural areas, and more frequently in the Eastern region than in other parts of the country. Furthermore, political and social elites are generally more inclined than rank and file voters to be in favour of full membership. However, polarization on the EEC issue is also most articulated at the elite level. In the present Storting all representatives of the Centre Party are opposed to EEC membership, all the Conservatives are in favour of it, while the Liberal, the Labour, and Christian People's Parties are split on the issue. In the Communist and Socialist People's Parties, which at present are not represented in the Storting, the party organization leaders are unanimously opposed to membership.

Similar patterns of party differences occur on the voter level. As an example we present
data from a Gallup poll. In Table V attitudes toward EEC have been related to
party preference in September 1971.

Although distributions of attitudes within the individual parties differ slightly from one poll to another the basic pattern of party differences has remained stable over time.¹⁸ Voters of the Conservative Party are most in favour of EEC membership, followed by the Liberal Party, the Labour Party, the Christian People's Party, the Centre Party, and finally the Socialist People's Party and the Communist Party. The difference in attitudes between the three anti-EEC parties on the one hand and the Conservative Party on the other, is substantial. However, it is worth noting that even among Conservative voters less than 50 per cent favour full membership. If we exclude respondents who indicate 'no opinion,' more than half the Conservatives prefer full membership, while the corresponding figure is far less than 50 per cent in all other parties.

Since all parties are deeply split over the EEC issue, we may ask: How did this issue
affect preferences?

Gallup samples of June and August 1971 have been pooled in which identical questions on the EEC issue were asked. The data have been classified according to position on the EEC, and for each group recalled voting in 1969 has been related to party preference at the time of the interview (cf. Table IV).

The total shifts in party preferences differ substantially from one group to another: of people who indicate 'no opinion' on EEC, 94.1 per cent are consistent in their party preference between 1969-1971, the corresponding figures are 95.1 per cent for those who prefer 'full membership,' 89.7 per cent for those who say 'association, etc.', and 89.4 per cent for those who say 'no special arrangement.' In other words, respondents who oppose membership have been most inclined to shift party.

We shall not go into details about the patterns of shifts within the various subgroups,
but it may be useful to look at the resulting party distributions, i.e., total distributions
at the election of 1969 compared to the preference in June and August 1971.

Table VI confirms the impression of great stability in party loyalties, despite the EEC issue. None the less some significant changes did occur. Changes were small indeed among people who indicated 'no opinion' on the EEC. Supporters of 'full membership' tended to move in the direction of the Labour Party. Opponents of EEC membership were attracted to the anti-EEC parties, and defected from Labour and the Conservatives. Although the changes are rather small, they follow a pattern which could be expected on the basis of the public EEC debate, and they also are consistent with the main trends reflected in the local elections of 1971.

4. The 1971 Local Elections and the EEC

Since the parties were more or less split on the EEC issue, and since local and not national elections were being held, the issue never became the main focus of the elections. None the less it was frequently mentioned during the campaign, and our analysis indeed suggests that attitudes toward the EEC were a major force in bringing about changes in party distributions. The election resulted in progress for the anti-EEC parties and losses for Labour and the Conservatives who were most in favour of EEC membership. Secondary analysis of Gallup poll data from the period preceding the election indicates a similar trend in the shifts of party preferences. Furthermore, the analysis indicates that respondents tend to move between the parties in accordance with their position on the EEC issue. Although this tendency is weak, it is consistent.

This does not mean that the EEC issue completely explains the election results. Some of the regional variations can hardly be explained in terms of this particular issue, for example the fact that Labour improved its position in the interior East, but lost in all other regions.

In general, an inconsistency seems to exist between the election results for the entire nation and for subgroups of the electorate, and this item needs further elaboration. The most conspicuous inconsistency occurs when we group communes according to degree of urbanization (cf. Table II). The pro-EEC parties maintained their position better in rural than in urban areas. This tendency is surprising when we take into account that the urban population in general holds more favourable attitudes toward EEC than do people in rural areas. As an explanation of this unexpected result we shall propose the following hypothesis: whether or not attitudes toward EEC shall induce voters to break with their traditional party loyalties depends upon the character of their involvement in the EEC issue. In general, involvement in politics is greater among people in urban than in rural areas.¹⁸ We may assume that the same difference occurs with regard to the EEC issue. And we may expect that the more strongly people are involved in this issue, the more likely they are to break with their old party, provided they perceive the stand of the party as different from their own attitude. Furthermore, resistance against EEC membership is drawn from different sources: in rural areas anti-EEC attitudes are largely based upon pragmatic expectations concerning the future of primary economies, agriculture and fishery, inside or outside the EEC; in urban areas EEC attitudes are apparently based more on ideological considerations, in particular left-wing socialist ideology. It is reasonable that people who disagree with their old party on the EEC, are more likely to defect from the party if the basis for disagreement is ideology rather than pragmatism.

The decrease in turnout may be explained by the same reasoning. Presumably the large number of non-voting was due to cross-pressures: people who disagreed with their old party on the EEC issue preferred not to vote at all instead of voting for some other party. One should expect that the more strongly people are involved in the EEC issue the more likely they are to find themselves subject to cross-pressures, provided they disagree with their old party. Consequently, it is reasonable that the decline in turnout

should be strongest in urban areas where degree of involvement as well as ideological
considerations are most predominant.

The election results of 1971 riay herald a realignment in Norwegian politics due to the EEC issue. Undoubtedly this issue has led to a restructuring of traditional conflict lines. Conceivably the new alliances may lead a number of voters to defect from their traditional party loyalties. But will this development continue after the question of Norway's entry into the EEC has been settled?

University of Oslo

Henry Valen

---

NOTES

1. I am indebted to the Norwegian Gallup Institute which has generously given me access to their data.

2. See S. Rokkan and H. Valen, 'The Elections to the Norwegian Storting in September 1969', Scandinavian Political Studies, Vol. 5, Oslo: Universitetsforlaget, pp. 287-300.

3. See A. Campbell, P. E. Converse, W. E. Miller and D. E. Stokes, Elections and the Political Order, New York: John Wiley, 1966, pp. 40-62. For Norwegian elections a similar regularity has been demonstrated by H. Valen and W. Martinussen, Velgere og politiske frontlinjer (Voters and Political Frontlines), Oslo: Gyldendal Norsk Forlag, 1972, pp. 13-31.

4. See Rokkan and Valen, op.cit.

5. An estimate of how non-partisan voters were likely to distribute among the parties at the 1967 local elections have been presented by the present author: Henry Valen, 'Norway: The Local Elections of 1967', Scandinavian Political Studies, Vol 3, Oslo: Universitetsforlaget, 1968.

6. See Valen, op.cit.

7. See T. Hjellum, 'The Politicization of Local Government', Scandinavian Political Studies, Vol. 2, Oslo: Universitetsforlaget, 1967, and S. Rokkan and H. Valen, 'The Mobilization of the Periphery', in S. Rokkan (ed.), Approaches to the Study of Political Participation, Bergen: Chr. Michelsen Institute, 1962, reprinted in S. Rokkan, Citizens, Elections, Parties, Oslo: Universitetsforlaget and New York: D. McKay, 1970.

8. See Hjellum, op.cit.; Rokkan and Valen, op.cit.; and Valen, op.cit.

9. For a measurement of party distances, based upon rank-ordering data, see P. E. Converse and H. Valen, 'Dimensions of Cleavage and Perceived Party Distances in Norwegian Voting', Scandinavian Political Studies, Vol. 6, Oslo: Universitetsforlaget, 1971. According to this study, the standardized distance score between Communists and the Socialist People's Party in 1969 was 2.90. Similar data obtained through a Gallup poll in June 1971 gave a score of 2.38.

10. Respondents who report no voiing in 1969 and/or no preference at the time of the interview have been excluded. Thus our figures have been computed on the basis of 'active' voters.

11. For a discussion of different measurements of stability, see Valen and Martinussen, op.cit., pp. 280-283.

12. For a detailed analysis of these changes, see Valen and Martinussen, op.cit., Chapter 10.

13. The processes of change between two elections have been described with admirable clarity by D. Butler and D. E. Stokes, Political Change in Britain, New York: St Martin's Press, 1969.

14. For an interesting analysis of attitudes toward EEC in the Norwegian electorate see T. Høivik, O. Hellevik and N. P. Gleditsch, 'Folkeopinionen og EEC', Samtiden 80 (1971), pp. 239-260; O. Hellevik, T. Hzrtvik and N. P. Gleditsch, 'Folkemeininga om EEC i 1971', Syn og Segn 78 (1972), pp. 103-117; and Willy Martinussen, Norsk EEC-opinion i forhandlingsåret, Oslo: Institutt for statsvitenskap, Universitetet i Oslo, 1972 (mimeo).

15. See Høivik, Hellevik and Gleditsch, op.cit., p. 245,

16. See W. Martinussen. 'Political Awareness and the Distribution of Other Social Resources; The Case of Norway', Scandinavian Political Studies, Vol. 6, Oslo: Universitetsforlaget, 1971, pp. 153-170; and Rokkan and Valen, op.cit., 1962.