1. Introduction

Throughout the year 1972 Norwegian politics was dominated by the issue of Norway's entry into the European Common Market (EEC). In fact, this issue had dominated the political scene since the summer of 1970 when the Storting decided to apply for membership in the Community.¹ In January 1972 an agreement on membership terms was reached between the Norwegian government and the EEC authorities. This proposal was submitted to a facultative referendum that took place September 24-25. As a result the membership alternative was rejected by 53.5 percent of the votes cast.

Shortly after the referendum the Labor government resigned, and a coalition government was formed by the parliamentary groups who had opposed Norwegian membership in the EEC, i.e. the Center Party (Agrarians), the Christian People's Party, and parts of the Liberal Party. The leader of the Christian People's Party, Mr. Lars Korvald, became Prime Minister in the new coalition, which enjoyed the parliamentary support of only 38 out of 150 members in the Storting.2 The primary mandate of the new government was to negotiate for a free trade agreement between Norway and the Common Market.

After the referendum several of the political parties suffered from internal controversies
stemming from the EEC issue. Only one of the parties, the Liberal, was split
over the issue. In 1973, however, the Labor Party suffered a minor split.

In this review article we shall first describe the character of the campaign. We
then present an initial analysis of voters' reactions to the referendum.

2. The Character of the Campaign

Referendum is not provided for by the Norwegian constitution. Occasionally, however, the voters have been invited to take a direct stand on important issues through facultative referenda.³ Since 1962, when the question of Norway's entry into the Common Market was first debated, it has been generally recognized that this highly controversial issue had to be settled through a referendum.

The campaign preceding the referendum was a most unusual experience in Norwegian politics. In general it might be expected that a campaign concerning a single issue will differ in character from an ordinary election campaign. This was certainly the case in 1972, because of the intensity and nature of the EEC issue. Although it was a matter of an international commitment, the decision obviously involved significant consequences for practically all areas of national politics as well.

The referendum called for a 'yes' or 'no' to Norwegian membership in the Common Market. Relatively few voters held the opinion that the country could stay totally outside the EEC. Consequently, the debate focused upon the question of how the nation should be attached to the Community: through a full membership or through a free trade agreement. Thus the seven parties were faced with only two alternatives. This situation would not have involved serious problems for the parties if opinion had been divided along traditional party lines. But the controversy cut across the parties and created bitter internal struggles in some of them, notably in the Labor Party, the Liberal Party, and the Christian People's Party. The Conservatives went heavily in favor of full membership, while the Center Party, the Socialist People's Party, and the Communist Party were close to unanimous in opposing membership.

The momentary transformation in the sheer number of voting alternatives - from a choice among seven to a choice between two - had a number of significant consequences for the campaign. For one thing, the political parties were less active and less visible than in ordinary elections, while huge ad hoc organizations, established by opponents and supporters of full membership, played a predominent role in the campaign. These organizations mobilized the support of political parties as well as the support of a variety of more or less well-organized political and social groups.

Second, opinions expressed by leaders were emphasized far more than in ordinary campaigns. Almost daily the newspapers printed statements by prominent members of political, social, economic, and cultural elites supporting or opposing EEC membership. Third, the conflict constellations deviated completely from the normal patterns and made for strange bedfellows. Thus the core of the opposition to EEC was composed of an alliance between the radicals in the cities and the farmers and fishermen in the periphery. Equally strange was an alliance in favor of EEC membership by two traditional antagonists, the Conservative Party and the Labor Party (the moderate wing) and their supporting organizations in the labor market, respectively the employers' organizations and the trade union movement. With this brief description as a background we may infer that not only were the political alternatives completely different from an ordinary election, the campaign organization and the channels of communication also differed widely from the normal pattern.

Now we may turn to the individual voter: How did he react in this rather unique situation? In the subsequent analysis we shall be concerned with the vote preferences of various subgroups of the population. More specifically we shall try to explore how the cleavages due to the EEC issue were reflected in the behavior of the voters and tc explain how these conflict lines relate to the 'normal' cleavage patterns in Norwegian politics.

A sizable proportion of the electorate must have experienced a conflict between traditional party identifications and loyalties with groups taking different stands on the EEC issue. An attempt will be made to explore how crosspressures generated by conflicting loyalties affected the level of voter participation.

The data in this analysis are drawn from three sources: a) public election statistics, b) data from the monthly Gallup polls, and c) a nationwide voter survey, where the fieldwork was done by the Gallup Institute from September 16 to 23, i.e. immediately before the referendum.⁴

3. EEC: Attitudes and Patterns of Conflict in the Electorate

During the entire debate a majority of the voters indicated a negative attitude toward membership in the Community.⁵ In the beginning of 1972 after the negotiations for membership terms had been concluded, there was an upward trend for the 'yes' alternative. But after April 1972 only minor changes occurred. The stability in the opinion was indeed remarkable. In light of the fact that most of the newspapers and most of the established elites were in favor of the membership alternative, a stronger swing in the 'yes' direction might have been expected.

Throughout the campaign, opinion polls also indicated great stability in the relationship of EEC attitudes to a number of background factors. Thus negative reactions occurred far more frequently in the peripheral than in the central parts of the country, more in rural than in urban areas, and more among farmers and fishermen than among people with urban occupations. We shall probe these relationships more carefully with data from the referendum.

Table I, based upon electoral statistics, shows a clear relation between region and
voting.

In Oslo and the provinces surrounding the Oslofjord close to 60 percent of the voters preferred full membership. In all other regions the majority went for the other alternative but substantial variations are evident: the 'yes' percentage declined from 45 in the East Central area to 28 in the North. Thus the heaviest opposition to EEC came from the peripheral provinces along the western and northern coast.

Apart from the regional differences, great variations occurred with regard to the structural character of communities. The 444 communes of the country have been classified according to density of population and type of economic activities. In sparsely populated fishing communes only 20.8 percent voted 'yes'; in other sparsely populated communes (largely agricultural), the corresponding figure was 34.5 percent. In densely populated rural communes (predominantly industrial), 41.5 percent voted 'yes', while the comparable figure in urban and suburban communes was 56.6 percent. These average figures for groups of communes clearly indicate that support for EEC increased with increasing urbanization. The variations in vote distributions are far greater for individual communes. Thus the percent in favor of EEC membership ranged from 6.5 to 70.7: the lowest figure was obtained in two typical fishing communes on the northern coast; the highest was registered in the commune of Bærum, which is a suburb of Oslo. Nonetheless, the relationship between EEC support and urbanization is evident in all regions.

A comparison of the 'yes' percentage between fishing communes and urban communes for each region shows a discrepancy of 20 for the East, 19 for the South and West, and 25 for the central and the northern regions. However, when the communes of a given region are classified into four structural subgroups, some of the categories are bound to contain very few units; for example, there is only one fishing commune in the eastern region. Because of the low number of units we might suspect that some aspects of the variations in vote distributions are due to chance. In Table II we have avoided this problem by dichotomizing the communes according to density of population ⁶ The dichotomy largely divides the communes according to predominant type of economic activity: in sparsely populated communes most people are occupied in farming and fishing, whereas densely populated communes are dominated by the secondary and tertiary branches of the economy.

In densely populated communes in the eastern and the southwestern regions more than half of the votes were cast in favor of EEC membership, and the difference between the two regions is only one percent. Densely populated communes in the central and the northern regions were substantially lower in EEC support. In sparsely populated communes sizable differences are evident between the eastern region on the one hand and the southwest and the central and northern regions on the other one. This difference is partly due to the strong opposition against EEC in the many fishing communes along the western and northern coast, but even in agricultural communes the support for EEC is consistently higher in the East than in other parts of the country. For all regions the proportion of 'yes' votes is substantially higher in densely than in sparsely populated communes.

Table II reflects variations along two distinct conflict dimensions: a territorial centerperiphery
axis and an urban-rural axis. Both conflicts have long traditions in Norwegian
politics.7 But they have not been particularly dominant in recent elections.8

In this article we shall not try to explain why the center-periphery and the urbanrural
axes became strongly salient due to the controversy over EEC membership.9 Our
main purpose is to describe the phenomenon.

We now turn to the question of how other cleavages in the system affected the EEC issue. For this purpose we shall apply data from the nationwide voter survey carried out immediately before the referendum. As a first approach we may examine how reactions toward the EEC are related to occupational status. The differences reflected in Table 111 suggest variations along two well-established economic cleavages: the urbanrural conflict (the conflict in the commodity market) and the left-right conflict (the conflict in the labor market).¹⁰

Fanners and fishermen are less likely to be membership supporters than people in urban occupations. This tendency corresponds to the urban-rural difference we observed in Table 11. Furthermore, a substantial difference is evident between manual workers on the one hand and people in middle-class positions, i.e. salaried employees and independents in business, on the other. This difference reflects variations along the left-right axes. As we have noted, the radicals in the cities are strongly inclined to oppose EEC membership. Since radicalism is positively related to social position, it is not surprising that workers are more inclined than middle-class groups to oppose membership.

However, the location of a person in relation to various cleavages is not only a matter
of objective social position; it may also be defined subjectively in terms of the individual's
perception of his own interests.

In the voter survey an attempt was made to operationalize our two economic cleavages.
The urban-rural cleavage has been defined as 'essentially a conflict between

producers and consumers of products from the primary sector of the economy.'¹¹ In an attempt to operationalize this conflict the respondents were presented with the following statement: 'To secure equality between different branches of the economy, farmers and fishermen ought to get far higher prices for their products.' The respondents were asked to react according to a seven-point scale ranging from 'absolutely disagree' to 'agree 100 percent.' Of course, reactions along this scale are correlated with occupational position: nearly 90 percent of farmers and fishermen indicated agreement while the corresponding figure for other occupational groups varied between 30 and 40 percent.

On the basis of our previous tables we should expect that the stronger the identification
expressed with farm and fishery interests the more likely the opposition to EEC
membership. This hypothesis is clearly supported by Table IV.

Opposition to EEC is strongest among those who agree to the statement of price increase: in the most extreme group only one out of four intended to vote for membership. Of those who absolutely disagree to price increase, i.e. the most urbanistic attitude, three out of four voted 'yes.'

The concepts of conservatism, radicalism, and liberalism are subjects of considerable discussion. For our purpose the left-right controversy (radicalism-conservatism) w«il be defined in terms of acceptance of state interference in various activities in society, with particular emphasis upon the economic sphere.

The radicals are most in favor of public ownership and public control, while the conservatives, who strongly support private enterprise, prefer as little as possible of state control and interference. In the voter survey three items were designed to measure this attitude.¹² The responses have been combined into an additive index ranging from one (farthest left) to six (farthest right or conservative). EEC attitude has been related to position along this index. The results, which have been presented in Table V,

indicate a substantial correlation between these two variables, with people on the right
most in favor of EEC membership. The opposition to EEC on the left, however,
comes out less clearly than expected. On the leadership level the radicals constituted

the core of the opposition in urban areas. In fact, from the mass media one is left with the impression that the radicals were almost unanimous in opposing the membership. Table V does not confirm this impression. The reason may be that attitudes were less polarized among rank-and-file voters than among political leaders, but an alternative explanation may simply be that our measure of radicalism is not sensitive enough.

Introduction of control variables does not change the picture substantially. The relationship between EEC vote and left-right position has been studied separately within each occupational group. For farmers and fishermen there is no relationship between the two variables. For workers, salaried employees, and independents the relationship is essentially the same. If we look separately at urban communes around 40 percent of the left-wing people voted 'yes,' on the right wing the percentage varied from 88 in the East to 71 in the Southwest.

Three cultural cleavages have played an important role in shaping Norwegian politics: the language conflict, the moral controversies articulated by the temperance movement, and the fundamentalist lay religious movement.¹³ The voter survey included questions measuring position along these cultural cleavages.¹⁴

Table VI clearly indicates that all the three cultural movements tend to mobilize opposition to EEC membership. The relationship is weakest for religion: the proportion supporting EEC is only 10 percent lower for members of religious organizations than for non-members. For teetotalism there is a pronounced difference between active teetotalers and active non-teetotalers, i.e. people holding a liberal view on alcohol policies. However inactive teetotalers do not differ significantly in EEC vote from inactive non-teetotalers. For language position there is a marked and regular relationship from active nynorsk supporters over to active riksmål people.

A liberal moral outlook as well as an active riksmål position are rather characteristic
attitudes among higher middle-class groups in the cities. Traditionally these groups
have been most European in their cultural style and outlook, and it is precisely in these

groups we find the strongest support for EEC membership. The three cultural movements may not be directly affected in their activities by the EEC, but apparently they reflect values which are easily mobilized in defense against external threats. They all have their main bases in the less urbanized sections of the population, particularly in the southern and western regions. They all may be considered countercultures in opposition to the centralizing and urbanizing forces in modern social and political development ¹³ By opposing Norwegian membership in the Common Market the supporters of the countercultures are indeed consistent with old traditions.

The data presented so far indicate that position on the EEC relates to all the traditional cleavages in the system. But the relationships vary greatly both' in strength and form. The question is: Which of the cleavages have been really salient in' the EEC decision? Several of the background variables are interrelated. In order to asse'ss their relative contributions to the outcome of the referendum a multivariate analysis is required. Space does not permit such an analysis in the present paper, but our work suggests that three cleavages were very salient in relation to the EEC issue: the territorial center-periphery axis, the urban-rural axis, and the left-right conflict. For the cultural variables most of the zero-order relationships disappear when we control for economic and territorial background. Thus in the central and northern regions where the 'no' percentage was highest, the three countercultures are not very strong. Apparently the cultural cleavages added to the protest of the periphery, but independently they did not contribute much to the variance in EEC attitudes.

The observed conflict constellations differ substantially from the pattern of recent national elections. The territorial conflict has been declining in saliency since the beginning of this century.¹⁶ As to the urban-rural contrast, it probably provides a great potential for political controversy in society now, but a recent analysis suggests that in the Storting elections of 1965 and 1969 the urban-rural conflict did not play an important r01e.17 In these elections the left-right conflict and a moral-religious conflict were the most important cleavages. The controversy along the left-right axis was largely a struggle between bourgeois parties on the one hand and socialist parties on the other. In the EEC debate the constellations along this axis have been quite different; at least on the level of political leadership the extreme left was strongly inclined to oppose the EEC, while moderate socialists joined with the conservatives to support the EEC. Our data suggest that this relationship is less polarized among the voters than on the leadership level.

4. Electoral Participation

The level of voter participation at the referendum declined compared with the latest Storting elections. In 1972, 79.2 percent of the total electorate went to the polls; the corresponding figures for the elections of 1969 and 1965 were 83.8 and 85.4 percent respectively. On the basis of previous research one should have expected a record high turnout level at the referendum.

In a well-known article Angus Campbell has demonstrated that the turnout level at a given election depends upon the importance the electorate attaches to the choice between different alternatives.¹⁸ He distinguishes between 'low stimulus' and 'high stimulus' elections, and he finds support for the hypothesis that 'if the alternatives are seen as implying significantly different consequences, the stimulation to vote will be relatively high.'¹⁹ A similar relationship has also been demonstrated with regard to Norwegian elections. Thus at the Storting elections of 1965 and 1969, when the alternatives were clearly perceived by the voters, the turnout was several percentage points higher than at previous elections.-⁰

Following Campbell's classification we may safely assert that the 1972 referendum
was a 'high stimulus' election. The campaign was conducted with unusual intensity, and

the alternatives were clearly defined. Furthermore, the alternatives were seen as implying different consequences, although opinion as to the character of the consequences was divided between the two camps. The involvement expressed by the electorate was unusually high. When asked about their interest in the EEC issue, 29 percent of the respondents indicated that they were 'very interested.'²¹ However, in order to evaluate the voter's involvement in 1972 one must make comparisons with previous elections. A set of identical questions was asked in election surveys of 1969 and 1972.²² In 1972, 28 percent of the respondents indicated that they were 'very interested' in international politics, while the corresponding figure in 1969 was 17 percent. In a similar way the proportion indicating great interest in national politics increased from 17 percent in 1969 to 25 percent in 1972. As might be expected, however, interest in local politics remained stable: the proportion saying 'very interested' increased only by two percentage points - from 14 to 16.

Despite the fact that the electorate was highly stimulated by the EEC issue, the turnout level declined from 1969 to 1972 by almost five percent. Why did this happen? One explanation might be that many voters abstained from voting because of the complexity of the EEC issue; ordinarily the electorate is not faced with problems of an international character. However, if lack of knowledge was the main reason, one should have expected a decline in turnout all over the country. This was not the case. Substantial territorial variations are evident in electoral participation, and, as we shall return to shortly, these variations seem to be related to the distribution of votes for and against EEC membership. This relationship leads us to hypothesize that the relatively low participation level at the 1972 referendum was due to crosspressures generated by the EEC controversies.

Generally speaking, the level of participation at Norwegian elections is high.²³ Under normal circumstances, crosspressures do not seem to affect participation significantly, because the multiparty system is so well tuned to the basic cleavage structure.²⁴ At the referendum a two-party situation occurred, which created a lot of tensions within practically all parties. In Table VII, which is based on data from the EEC study, vote intention at the referendum has been related to party preference. First we notice that the distribution for the total sample closely reflects the actual outcome of the referendum, in which 47.5 percent of the votes cast went in favor of full membership. Table VII indicates great variations among supporters of different parties. The two

extreme parties on the left wing together with the Center Party (Agrarians) constitute the strongest opposition to EEC, while the Conservatives are most likely to be supporters. The Labor, Liberal and Christian People's Parties are severely split over the issue. The pattern in Table VII is consistent with attitudes reflected on the political leadership level.²⁵ All parties, except the Liberals, took a stand in their 1972 congresses either for or against EEC membership. From Table VII we may infer that a substantial

proportion of the voters disagreed with their party's stand on the EEC. One way of measuring the amount of dissonance is simply to add the number of respondents in the various parties taking a stand different from that of the official party line. Excluding the Liberals and those who did not indicate any particular party, the deviators constituted 18 percent of the sample. Considering the intensity of the debate and the involvement of the electorate many voters must have experienced serious conflict over split loyalties, and above all, over loyalty to party and loyalty to groups sharing one's own position on the EEC.

In the survey an attempt was made to study the amount of psychological conflict felt by individual voters. The respondents were asked: 'Which party or parties do you think come closest to your own stand regarding Norway's relation to EEC?' Sixtyeight percent of the respondents indicated agreement with their own party; 18 percent were in disagreement; only 14 percent did not perceive any differences between the parties with regard to the EEC issue. As might be expected, it was largely supporters of the Labor, Liberal, and Christian People's Parties who indicated disagreement with their own parties. The data suggest a remarkable fit between the voters' perceptions and the actual stand taken by the various parties. Apparently this fit is due to the intensity of the debate and the clarity of the alternatives.

Used in conjunction with aggregate statistics for individual communes, the survey data are likely to provide an excellent basis for testing our crosspressure hypothesis. Unfortunately we cannot go into this analysis here. An attempt will be made, however, to elucidate the problem by describing territorial variations in turnout level.

Voter participation in Norway is clearly related to social structure: the turnout level has always been lower in peripheral than in central parts of the country, and within given regions turnout tends to increase with increasing urbanization.²⁶ Consequently, in order to study local variations in the referendum it is not sufficient to look at turnout figures in this particular election. The results also have to be considered against a baseline established at 'normal' elections. For this purpose data from electoral statistics will be compared commune by commune for the referendum and for the Storting election of 1969.

When communes are classified according to type of community, an interesting tendency is evident: the decline in turnout from 1969 to 1972 is around 5 percent in urban communes, in densely populated rural communes and in sparsely populated communes with agricultural economy; in fishing communes turnout increased by 1.6 percent. Thus, the gap in participation between fishing communes and other types of communes tended to disappear at the referendum. If we look at the two extreme types, the turnout percentages in 1969 were 75.3 and 84.8 in fishing and urban communes including suburbs respectively. The corresponding figures in 1972 were 76.9 and 80.1, i.e. the difference in participation was reduced from 9.5 to 3.2 percent. This tendency is consistent within all regions. If we classify communes only by region no systematic tendencies are evident. The decline in turnout is lower in the North and in the Southwest than in other regions, but apparently this difference is due to the fact that the two former regions have the highest number of fishing communes.

We have seen that in coastal communes where the referendum had some mobilizing effect upon the electorate, the opposition to the EEC was particularly strong. This tendency suggests that turnout is related to vote distribution in the local community. In order to test this hypothesis the 444 communes have been classified in decils according to 'yes' percentage at the referendum. Variations in turnout are presented in Figure 1.

The curve indicates that there is indeed a relationship between attitude toward the
EEC and voter participation expressed as a difference in turnout between 1969 and
1972. In communes where opposition to EEC membership is exceptionally strong

a substantial vote mobilization occurred; with around 20 percent of 'yes' votes participation is about as high in 1972 as in 1969, but from then on the 1972 level declines. With increasing support for EEC the gap in turnout between 1969 and 1972 tends to increase, but when the 'yes' percentage reaches 45 the trend shows a reversal. With minor deviations the trend in Figure 1 is consistent within all types of communes, and all regions.

How does this tendency fit the hypothesis that the low turnout at the referendum was mainly due to crosspressures? Obviously, inconsistency between own EEC attitude and identification with some national party cannot have been a direct source of crosspressure, because such inconsistencies are equally likely to oc,<;uf regardless of the

amount of support for EEC in a given commune. For example, the Labor Party enjoys strong support along the North Norway coast, where opposition to the EEC was overwhelming and where vote participation at the referendum tended to be relatively high. A great number of voters in this area must have felt a conflict between their own EEC position and party loyalty.

Figure 1 suggests that pressures generated in the local community have greatly affected the participation level: the more unanimous the opinion, the higher the participation. This regularity is very clear in communes where opposition to the EEC was particularly strong. The most significant decline in turnout occurred in communes where the two camps were about equal in strength. The picture is less clear in communes with a majority favoring the EEC, although the main tendency is toward improving the relative participation level. But the curve in Figure 1 is not symmetrical.

One obvious reason is that the degree of consensus is much higher in communes opposing the EEC than in those favoring membership. A great number of the communes in the last category are urban or at least highly industrialized. The largest ones are bound to contain a variety of local communities with different political inclinations. For one of these communes, Oslo, information is available about vote distribution as well as turnout for each separate ward or polling place, 82 altogether.

When the wards are ranked according to strength of EEC support the turnout curve fits nicely to the curve in Figure 1: in wards where the 'yes' percentage was very high (80 percent or more) the relative voter participation was also high, and with declining 'yes' percentage the turnout level tended to decline.

On the basis of aggregate statistics we may conclude that voter participation was closely related to the distribution of attitudes toward the EEC in the local communes. This finding suggests that the individual voter has been exposed to great pressures from his social environment in the community. But how did the voter relate local pressures to identifications with national parties and leaders? How did various pressures produced by a most unusual situation affect his attitudes and behavior at the polls? These and similar problems, which will be the focus of subsequent reports, call for contextual analysis based on a combination of aggregate statistics and survey data.

5. Concluding Remarks

Our analysis clearly supports the notion that the political conflict structure in Norway changed in character because of the issue of entry into the European Common Market. Differences in the EEC vote have been observed for all traditional cleavages in the system, but there are three eminent cleavages: the center-periphery axis, the urbanrural conflict, and the left-right conflict. The new conflict constellation produced strange alliances, which could be observed both on the level of political leadership and among the voters. The most important groups in favor of the EEC membership were the Conservative and Labor Parties and their supporting organizations in the labor market - i.e. the employers' organizations and the trade unions. The core groups in the opposition to the EEC were the radicals in the cities and farmers and fishermen in the periphery.

After the referendum, negotiations were opened for a trade agreement between Norway and the Common Market. The conflict structure produced by the EEC debate remained essentially unchanged. Only the Conservative Party kept advocating full membership in the Common Market. All other parties accepted the results of the referendum. Nonetheless several of the parties suffered from bitter internal struggles. In November 1972 the Liberal Party was split. In the Labor and Christian People's Parties strong efforts were made to bridge the gap between the two camps. The conflict was brought to the fore in the nomination process before the 1973 Storting election. In the selection

of candidates both parties sought a balance between supporters and opponents of EEC membership. Despite these efforts, a minor anti-EEC faction ousted the Labor Party and started negotiations with the Communist Party and the Socialist People's Party for the purpose of running joint left-socialist tickets.

The monthly Gallup polls have indicated instability in party preferences in recent years. Compared with the 1969 Storting election, changes in party preferences have occurred far more frequently among opponents of the EEC than among supporters. The Storting election of September 1973 will give a first indication of the impact of the EEC issue upon the electorate, but it will not give the final answer to a more basic question: Which long-term changes will now occur in the party system? We are not likely to get the answer until Norway's relation to the EEC has been finally settled and until several elections have passed.

University of Oslo

Henry Valen

---

NOTES

* The author wishes to express his appreciation to Mr. Ragnar Waldahl, Mr. Eyolf Steen and Mr. Jan Hovden who have done the computer work and Dr. Andrew Cowart and Mr. Willy Martinussen who have read the manuscript and made valuable suggestions. The author, however, takes full responsibility for the contents.

1. For a more detailed account of events in 1971 and the development of the EEC issue, see H. Valen, "Norway: Local Elections in the Shadow of the common Market,' Scandinavian Political Studies, Oslo, Universitetsforlaget, 1972, 272-282.

2. Throughout the debate attitudes have been more favorable toward EEC on the elite level than on the level of the ordinary voter. Thus EEC membership was rejected by more than half of the voters but by only a fourth of the members of parliament. For a discussion of the contrast in EEC attitudes between elites and rank-and-file, see O. Hellevik and N. P. Gleditsch, The Common Market Decision in Norway: a clash between direct and indirect Democracy,' in the present volume, pp. 227-235.

3. This has happened four times: in 1905 a referendum was held concerning the dissolution of the Union with Sweden and another one concerning the question of the future form of government, republic or monarchy. In 1919 a referendum was held on the introduction of prohibition. In 1926 prohibition was abolished after a new referendum.

4. A probability sample, 1641 persons, was interviewed. The study was directed by Henry Valen and Daniel Katz. A report from this study is forthcoming.

5. See note 2.

6. No single cell in Table II is based upon less than 28 communes.

7. The character of cleavages in Norwegian politics has been analyzed in a number of publications. See particularly: S. Rokkan and H. Valen 'Regional Contrasts in Norwegian Politics,' in E. Allardt and S. Rokkan (eds.), Mass Politics, New York: Basic Books, 1970; S. Rokkan, 'Geography, Religion and Social Class,' in S. M. Lipset and S. Rokkan (eds.), Party Systems and Voter Alignments, New York: The Free Press, 1967; P. Converse and H. Valen, 'Dimensions of Cleavages and Perceived Party Distances in Norwegian Voting,' Scandinavian Political Studies, Oslo: Universitetsforlaget, 1971; and H. Valen and S. Rokkan, 'Conflict Structure and Mass Politics in a European Periphery: Norway,' in R. Rose (ed.), Comparative Electoral Behaviour, New York: The Free Press, 1972.

8. P. Converse and H. Valen, note 7.

9. See Converse and Valen note 7. Similar conflict constellations were apparent in 1961-63 when the EEC issue was first debated.

10. See Valen and 'Conflict Structure and Mass Politics' note 7.

11. Ibid.

12. The following statements were used: 1) 'If society has no opportunity to control private business, the leaders of banks and industry will get far too great influence'; 2) 'It is easier to secure full employment if the state has control over banking and business activities'; and 3) 'In this country we ought to work for far greater equality regarding people's income.' The respondents were asked to give their reactions to these statements according to a scale ranging from 1 (absolutely disagree) to 7 (100 percent agree). A factor analysis indicated unidimensionality in the responses toward the three items. The index was constructed simply by adding the responses for each individual.

---

12. The following statements were used: 1) 'If society has no opportunity to control private business, the leaders of banks and industry will get far too great influence'; 2) 'It is easier to secure full employment if the state has control over banking and business activities'; and 3) 'In this country we ought to work for far greater equality regarding people's income.' The respondents were asked to give their reactions to these statements according to a scale ranging from 1 (absolutely disagree) to 7 (100 percent agree). A factor analysis indicated unidimensionality in the responses toward the three items. The index was constructed simply by adding the responses for each individual.

13. See particularly Rokkan, 'Geography, Religion and Social Class,' Valen and Rokkan, 'Conflict Structure and Mass Politics ...'

14. Religious involvement: the respondents were asked about membership in religious organizations. Teetotalism: Q. 'Do you happen to drink alcohol of any kind, or are you a teetotaler?' (If non-teetotaler): 'In this country the authorities have, as you know, introduced a number of rules, regulating the use of alcohol. Do you feel the rules are too strict or are they too lenient?' and (If teetotaler): 'How interested are you in teetotalism? Would you say that you are very interested, fairly interested, or not particularly interested?' Non-teetotalers who say they are 'very' or 'fairly' interested have been classified as 'active', those who say they are not particularly interested are called 'inactive' teetotalers. Non-teetotalers who accept the present rules or say they are 'lenient' are called 'inactive', and those who think the rules are too strict are called 'active' non-teetotalers.' Language Q.: 'Do you prefer bokmål (riksmål) or nynorsk when you write?' 'How interested are you in the language issue? Would you consider yourself very interested, fairly interested, or not particularly interested?' Respondents who indicate that they are very or fairly interested have been classified as 'active', while those who say not particularly interested are called 'inactive.'

15. Rokkan and Valen, 'Regional Contrasts in Norwegian Politics,' op.cit.

16. Ibid.

17. Converse and Valen, 'Dimensions of Cleavages and Perceived Party Distances in Norwegian Voting,' op.cit.

18. Angus Campbell, 'Surge and Decline', in A. Campbell, P. E. Converse, D. E. Stokes, and W. E. Miller. Elections and the Political Order. New York. Wilev. 1966. t>r>. 78-95.

19. Ibid.

20. H. Valen and W. Martinussen, V elger e og politiske frontlinjer, Oslo: Gyldendal, 1972, pp. 18-31.

21. Question: 'How interested are you in the EEC issue? Would you say that you are very interested, fairly interested, a little interested, or not interested at all?'

22. The respondents were asked: 'Of course, people are not equally interested in all areas of politics. Let us look at local politics in this commune. Would you say that you are very interested, fairly interested, or not much interested in local politics?' A similar wording was used for measuring interest in domestic national politics and in international politics.

23. For an analysis of the development of vote participation in Norway, see S. Rokkan, Citizens, Elections, Parties, Oslo: Universitetsforlaget, 1969, Chapters 6 and 7.

24. See Valen and Rokkan, Conflict Structure and Mass Politics in a European Periphery, op.cit.

25. See note 2.

26. See note 23.