Resumé

The Storting election of 1977 produced great changes in the partisan distribution of the vote compared with the preceding election of 1973. The latter, the most deviating election in Norway after World War 11, took place one year after the referendum concerning Norway's entry into the European Communities, and the EC issue still dominated the political scene.¹ While on the surface the election of 1977 restored the situation that existed before the EC dispute started, some notable changes are nevertheless evident.

The article provides the details of these trends and discusses the causes and
some possible consequences. It presents an initial analysis of the 1977 elections
in Norway and the political context in which it took place.*

1. The Political Background

The election of 1973 resulted in a narrow socialist majority in the Storting:
62 seats for the Labor party and 16 for the newly formed (leftwing)
Socialist Electoral Alliance (SV), out of a total of 155.

Although Labor's support had severely declined, the party managed to sustain a minority government throughout the subsequent four year period. During the first two years this minority government was faced with a highly divided opposition. The cooperative spirit of the bourgeois parties, which had formed the 1965-71 coalition government, had evaporated during the EC debate. Only in 1975 did the Conservative, Christian People's, and (Agrarian) Center parties open discussions in an attempt to reestablish a viable government alternative to the Labor party. The Liberals, which after the referendum had been split and reduced to two miniparties, did not join their former coalition partners in these discussions. (Valen & Rokkan 1974.)

---

* I would like to thank Helge Holbæk-Hansen and Bernt Aardal who carried out the necessary computer work.

The local elections of 1975 signalled a reversal back to the distribution that existed before the EC dispute began. (Valen 1976b.) On the parliamentary level political frontlines gradually became more clear during the next two years. The three bourgeois parties publicly confirmed their intention of forming a coalition government, provided they should obtain a majority in the Storting. However, they were unable to present a joint platform, and they refused to designate a slate of potential government ministers. Apparently, the idea of a bourgeois coalition obtained widespread support within the Conservative and Christian People's parties, while the Agrarians were severely split. While the latter's national congress of 1977 declared its willingness to cooperate in a coalition, a sizeable wing of the party, including the party's youth organization, opposed cooperation, claiming that a possible bourgeois coalition would be dominated by the Conservatives and would not benefit the interests and ideas of the parties of the center. The debate in the Agrarian party went on with great intensity practically until the day of the election.

Factional struggles in political parties are certainly not a new phenomenon, but in Norway after 1945 internal controversies have never been as visible as at the present time: the political youth organizations in particular have articulated opposition to the established leadership. The lack of political and organizational cohesion has been most conspicuous in the Socialist Electoral Alliance, which from its birth in 1973 consisted of highly centrifugal groups. (Valen & Rokkan 1974.) In the spring of 1975 the latter moved towards a unified party, the Socialist Left party (SV). Soon, however, an orthodox wing split away to reestablish the old Communist party.

2. The Election Campaign

Although two government alternatives were presented at the election of 1977, it was less clear what these implied in terms of policy differences. Clearcut policy positions could be identified with regard to only two single issues; the question of abortion upon demand, and that of a bank reform. The first issue had been on the agenda since 1969. It was clearly indicated that a socialist majority would introduce abortion upon demand, while the bourgeois parties were committed to making access to abortion more restrictive. On the second issue, the Storting earlier in the year had passed a bill which provided for the inclusion of government appointed public representatives on the boards of private banks. With the exception of Lange's antitax party, all parties supported the principle of public repre -

sentatives in private banks, but they were in disagreement over their number. The socialist majority had decided upon a majority position for public representatives. The opposition parties promised that a bourgeois government would reduce this number to a minority. It is not easy to assess the importance of these two issues in the campaign. On a great variety of other questions, the parties failed to clarify their priorities.

The lack of clearcut policy alternatives may partly be explained by the unusual character of the campaign. Politicians and parties had limited opportunities to discuss policy priorities, because the campaign was sidetracked by two important sets of events. A few weeks before the elections left socialist newspapers leaked stories concerning military intelligence and political surveillance by the police. The leakages led to police investigations and triggered a discussion on the activities and character of the socalled secret services. These events did not relate - at least not in a direct way - to the election. The other major event was the worsening of the economic situation. The Norwegian economy had been rather prosperous, despite the international recession, and unemployment was negligible. However, economic deterioration was dramatically highlighted by the devaluation of the Krone two weeks before polling day. Subsequently, politicians paid great attention to economic problems, particularly to the possibilities of maintaining full employment. Everybody emphasized the seriousness of the situation, but distinct policy options that could be identified with the two alternative sets of government parties did not emerge.

Nevertheless, with the notable exception of the abortion issue, themes reflecting left-right differences dominated the 1977 campaign. The election results once more indicated that the two main antagonists along the left-right axis, Labor and Conservative, tend to gain from this kind of polarization.

3. Electoral Results

Table 1 compares the distribution of votes and seats for the Storting
elections of 1973 and 1977.

The election confirmed the status quo in the balance between socialist and non-socialist parties in the Storting. The Labor party won 14 seats, the same number as the SV lost. Consequently, the minority Labor government could remain in office, with a strengthened position after the election. It was indeed a narrow victory. In several constituencies the last

seat was won by a margin of only a few votes, but the final counts gave 78
socialist and 77 bourgeois representatives.2

The tendencies which appeared at the local elections of 1975 were to a
large extent confirmed and strengthened (Table 1). The main results of the
1977 election may be summarized as follows:

(1) The Conservative and Labor parties, which had lost heavily in 1973, enjoyed a remarkable recovery. Conservative gains were particularly notable. In fact the party obtained its greatest share of the vote since the election of 1930. The Labor party, however, was still some four percent below its level of 1969. (2) The extreme parties of the left and right, SV and FRP (ALP) respectively, declined sharply. The antitax party (FRP) obtained less than two percent of the votes and lost all its mandates in the Storting. Parties to the left of Labor obtained jointly about five percent of the votes, i.e. approximately the same share as in 1969. It is worth noticing that the decline of the extreme parties coincided with the reestablishment of two fixed government alternatives. (3) The Christian People's party was the only one of the center parties which maintained its level of 1973. The Agrarians and the New People's party suffered severe losses, and jointly the four middle parties declined by 4.6 percentage points. (4) Due to the electoral system the composition of the new Storting does not exactly reflect the distribution of the votes. Labor and SV jointly obtained 46.5 percent of the votes but 50.3 percent of the seats, while the three coalition parties obtained 45.8 percent of the votes, but only 48.4 percent of the seats. Thus both groups are overrepresented, but the socialist side more so. In general the changes in the distribution of seats were greater than those in the distribution of votes. The difference is greatest for the Agrarian party, which declined by 2.4 percent of the total votes, but by 5.8 percent of the seats. The electoral system is most sensitive for parties which are close to the threshold of representation, and the threshold varies from one constituency to another depending upon the number of seats and the relative strength of different parties. Thus in 1977 the Agrarian party fell slightly below the threshold in a number of constituencies, and consequently, the party's share of the seats became smaller than its share of the votes. Four years earlier the same party managed to climb above the threshold in almost all constituencies to make it relatively overrepresented in the Storting.3

(5) The turnout level, which had been relatively low in 1973 due to
crosspressures generated by the EC dispute, rose in 1977 to about the
'normal' level established in the 1960s.4

A considerable part of the changes between 1973 and 1977 must be

accounted for by the vanishing effects of the EC issue. The gains of Labor and the Conservatives are obviously related to this event. And it is a reasonable hypothesis that the situation created by the EC issue largely explains the rise and fall of the two extreme parties, SV and FRP (ALP). On the other hand, it would be an exaggeration to argue that this issue is the only cause of the recent changes in Norwegian politics. For example, the Christian People's party maintained its share of the votes, despite the fact that it had been a 'NO'-party and enjoyed a great victory in 1973. Furthermore, substantial political change and great electoral volatility are not unique Norwegian phenomena. During the seventies similar tendencies have been evident in many western countries. For the Norwegian case, it will be an important task in the analysis of recent elections to assess the relative significance of short-term forces, like the EC issue, as opposed to long-term tendencies arising from social and economic change. This basic problem cannot be dealt with in the present paper, which is limited to a description of the character of recent changes: the baseline for this description will be 1969, i.e. the last Storting election before the EC dispute began. Two trends seem to be of particular interest:

(a) The great changes among the bourgeois parties, i.e. the gains of the
Conservatives and the Christian People's party and the losses of the
Liberals and the Agrarians.

(b) The decline of the total socialist vote. At all Storting elections from 1945 to 1969 the socialist parties jointly won a majority of the votes, on average 51.6 percent. In 1973 the proportion declined to 46.9 percent and in 1977 it was 47.5 percent (including RV and NKP).

4. Geographical Variations

Norwegian politics have always been characterized by relatively large geographical variations between regions and between urban and rural areas within given regions. (Rokkan & Valen 1970.) These differences were recently articulated again, when the interests of rural areas and of peripheral parts of the country were mobilized in the opposition to Norwegian membership in the EC. (Valen 1973: 1976). The election results of 1973 clearly reflected this tendency: the losses of the 'YES'-parties were particularly large in areas where the EC-opposition had been strong. (Valen & Rokkan 1974.) One may ask to what extent the 1977 election restored the partisan vote distribution within territorial subgroups, or how and to what extent major electoral variations between 1969 and 1977 are reflected in terms of territorial variations. In an attempt to answer these

questions, some aggregate electoral data for the 454 communes will be
presented.s

First, variations with regard to urban-rural differences will be considered. In Table 2 the communes have been classified in four groups according to a simple typology, ranging from the most rural (fishery communes) to the most urban (cities, towns and suburbs).

Table 2 confirms some well-known trends with regard to urban-rural differences. The Conservative party enjoys far greater support in urban than in rural communes, while the reverse tendency is true for the Christian People's party and, more particularly, the Agrarian Center party. Support for other parties in the system varies less consistently with type of commune. However, electoral changes between 1973 and 1977, considered

in relation to urban-rural differences, are less consistent than one should expect. Since the EC opposition and the subsequent losses for the 'YES'-parties were relatively strongest in rural areas, and particularly in fishery communes, one might hypothesize that the gains in 1977 for Labor and Conservatives should be relatively greater in rural communes. Table 2 shows no marked tendency in this direction. There is more support for the reverse hypothesis that the 'NO'-parties should suffer greater losses in rural than in urban areas. The trend in this direction is rather clear for the Liberal and Agrarian parties. However, the Socialist Left party has declined relatively most in densely populated and urban communes. But this deviation from the main pattern is not surprising considering the fact that this party in 1973 was particularly attractive to frustrated urban radicals who abandoned the Labor party over its position on EC entry. (Valen & Rokkan 1974.) The similarity in proportions of losses for SV and gains for Labor, particularly in densely populated and urban communes, suggests that a great number of voters must have moved directly from the one party to the other. The Christian People's party constitutes another interesting deviation from the main pattern. Despite the fact that the party opposed EC membership, it improved its position slightly in rural communes, and suffered only tiny losses in densely populated communes.

Thus the picture of electoral changes between 1973-77 that emerges from Table 2 is a rather complex one. The data suggest that these changes can be explained only to a limited extent in terms of short-term forces arising out of the EC issue. This observation is supported by a comparison of partisan distributions in 1969 and 1977. Thus the gains of the Christian People's party seem to be largely unaffected by the EC dispute. From 1969-71 the party increased its share of the votes by 7.8 percent in fishery communes, by 3.4 percent in other sparsely populated communes, and by less than three percent in more urbanized areas. This tendency towards greater gains in rural than in urban communes is most likely related to the party's stand on the abortion issue. Opposition to free abortion has always been greater in rural than in urban areas, and previous analyses clearly indicate that the Christian People's party has gained from this issue. (Martinussen & Valen 1976; Valen & Martinussen 1977). In a similar way the long-term gains of the Conservative party do not seem to have been much affected by the EC dispute. Between 1969 and 1977 this party increased its votes substantially in all types of communes, but most (6.9 percent) in fishery communes.

Finally, attention should be paid to change in overall socialist strength.

After 1973 the socialist parties jointly improved their position in rural areas, while remaining static in urban communes. By contrast after 1969 the socialist proportion of the votes declined strongly in fishery communes. In other types of communes the decline was less severe, and it did not vary much from one type to another.

As a general conclusion it may be stated that electoral changes at recent elections have been far greater in fishery communes than elsewhere. In a long-term perspective (i.e. 1969-77) the changes in fishery communes have largely benefited the Conservatives and the Christian People's party, while the socialist parties have lost. On the basis of aggregate data it is difficult to draw inferences v/ith regard to the impact of the EC issue upon electoral changes, but the fact that opposition to EC was strongest precisely in fishery communes,, suggests that directly or indirectly this issue triggered significant processes of change.

In a similar way communes may be classified according to region. There were substantial changes in partisan distributions, but again it is difficult to identify clear relationships between electoral change and the strength of EC opposition in specified regions. At this point it is sufficient to draw attention to the two major trends of recent years: the decline of the socialist vote, and the increase in the Conservative vote.

For the entire nation the socialist vote declined by four percent from 1969 to 1973 (from 51.0-46.9 percent), and then increased by .6 percent from 1973 to 1977 (Table 1). Variations between regions were not great, except for the North where the decline from 1969-73 was more than nine percent, a decline which was offset in 1977 by a reverse trend of two percent back towards the socialist camp. The great changes in the Northern provinces may be due to the very strong EC opposition there. It should be noted, however, that even in the 1960s volatility was particularly great in this part of the country, where the Labor party won a remarkable victory in the 1969 election. (Valen & Martinussen 1972, 287-8.) The Conservatives also experienced great changes in the North. But for this party gains were even greater in the South and West. In fact, in 1977 the party climbed above 24 percent of the votes in the latter regions, which have always been dominated by the parties of the center. The territorial variations in the growth of the Conservative vote may conveniently be illustrated by a map describing the changes between 1969 and 1977 in each individual commune (Fig. 1).

Apart from a few scattered communes the party has either maintained its position (i.e. remaining within a range of ±5 percent of the 1969 level) or it has made greater gains. The latter are largely located in the South and West, plus some pockets in the North and in the central area around the Oslofjord.

Obviously some of the changes described must be related to fluctuations in the support for other parties. In particular, the decline of the Liberal party is of great interest in this regard. In Table 5 electoral changes between 1969-77 have been considered, with the local units classified according to both region and type of commune. The background variables have been simplified into fewer categories: three regions, with two types of communes within each region.

Some main tendencies are evident:

(1) In the southern regions support for Labor as well as the total socialist vote has declined more sharply in urban than in rural communes. In Trøndelag and the North the trend is reversed, with the greatest socialist losses in rural areas. A closer inspection of the data reveals

that the pattern in Trøndelag is similar to that of the southern regions,
so it is actually only the North which is a deviating case.

(2) The decline of the joint Liberal and New People's party vote is consistently greater in urban than in rural areas. Traditionally the Liberal party had its stronghold in the South and West, and the decline is greatest in these regions. The New People's party still enjoys somewhat stronger support in the South-West than in other regions, while regional variations are less evident for the new Liberal party.

(3) The Christian People's party has consistently made greater gains in rural than in urban communes, although in the South-West the difference between urban and rural is negligible. This pattern is consistent with the fact that religious attitudes as well as opposition to free abortion are stronger in rural than in urban areas. Furthermore, such attitudes are strongest in the South-West region. (Valen & Rokkan 1974 b.)

(4) Throughout the country the Agrarians have declined more in rural than in urban communes. This tendency is reasonable enough, since the party traditionally has been very weak in urban areas, with consequently little to lose.

(5) For the Conservative gains, there is no clear tendency with regard to urban-rural differences. As a result the party has strengthened its position relatively more in rural communes, where previously it was rather weak. Furthermore, the election made the Conservative party one of the major alternatives in the South and West.

The picture emerging from Table 5 is rather complex. Obviously, the tendencies described must be inter-related in some way. For example, one may question to what extent the losses of the socialists are related to the gains of the Christians and the Conservatives in rural and urban areas respectively. Similarly, one may conjecture that the gains of the Conservative party are more or less related to the decline of both Liberals and Agrarians. Aggregate data can provide a structural context for studying partisan changes, but in order to understand the character of change among the parties, information about the behavior of the individual voter is needed.

5. Individual changes

Data on individual changes have been obtained through a panel study.6

Table 6 presents changes in electoral behavior between 1973 and 1977.
The marginal distributions in the bottom row and in the column at the

extreme right of the table indicate thai: the strength of individual parties deviates somewhat from the actual distribution of their electoral support (cf. Table 1). Thus the Conservatives and SV are underrepresented at both elections, while the Agrarian Center party is most overrepresented. However, by comparing for each party the marginal figures at the two elections, the differences come close indeed to the changes that actually occurred (Table 1). This consistency suggests that the trends in vote exchanges among parties are reasonably well reflected by the data.

The italicized figures along the diagonal of Table 6 indicate the proportions of the respondents who preferred the same party at both elections. These proportions add up to 69.6 percent. Presumably this figure slightly overestimates the stability at the polls since the youngest age brackets, which tend to be most unstable, are not included in the panel. One is probably safe in assuming that approximately one out of three respondents shifted his/her party preference between 1973 and 1977 i.e. a rate of change equivalent to that of the 1969-73 period. (Valen & Rokkan 1974.) These figures confirm that volatility among Norwegian voters has been very high in the 19705. By comparison the 1965-69 panel data indicate that only one out of four respondents changed preference.

The main trends in Table 6 may be summarized as follows:

(1) Most of the losses of the extreme parties, SV and FRP (ALP), went to
their closest neighbors in the party system, the Labor and Conservative
parties respectively.

(2) The Labor party gained some votes in its exchanges with the parties at the center, while it lost substantially to the Conservatives. The latter trend indicates a swing in a conservative direction, which is not unusual. But shifts between Labor and the Conservatives have not been a normal pattern at previous elections. Since these parties are perceived to be rather distant from one another along the left-right axis, people who left one of them tended to move to some party at the center, in particular the Liberal party.⁶ The faet that the Liberals more or less disappeared as a moderate alternative lying between Labor and the Conservatives apparently had an impact upon the pattern of change between left and right.

(3) The Christian People's party, whose share of the total vote remained unchanged between 1973-77, lost a number of votes to the Conservatives and Labor, but these were compensated by gains from the Agrarians and the Liberals.

(4) The Agrarian Center party gained very few votes from other parties,
but lost heavily to the Conservatives, the Christians and Labor.

(5) The Conservative party enjoyed net gains in its vote exchanges with all
other parties.

As might be expected, a substantial part of the changes reflected in Table 6 can be explained as a result of the EC dispute. Thus the two 'YE S'-parties, Labor and the Conservatives, are the only parties which enjoyed net gains in the balance between voting and non-voting at the two elections. One may conjecture that they managed to mobilize previous supporters, who stayed at home in 1973 because they were disappointed with their party's stand on the EC issue. Furthermore, some of the losses of the 'NO' parties should have been anticipated. A number of opponents of the EC who had previously supported Labor or the Conservatives preferred some other party in 1973: Laborites went to SV, the Christian People's party and the Center party, while Conservatives favored the Center and Christian parties. (Valen & Rokkan 1974; Valen & Martinussen 1977.) It is not surprising that as the EC issue declined in saliency, deviators of 1973 were inclined to state a preference consistent with their basic political loyalties.

6. The Main Tendencies

As already observed, substantial changes did occur in partisan distribu
tions between 1969 and 1977. And again, the panel data presented in Table
7 seem to reflect these changes rather well.

Table 7 presents the vote distributions in 1977 for respondents grouped according to their voting in 1969. In evaluating vote exchanges among parties, it is essential to take into account the relative size of each party. Thus in 1977 three percent of the 1969 Laborites indicated that they voted for the Christian People's party, while ihe same percentage of Christians shifted to Labor. But because Labor is much larger than KRF, the actual number of voters who shifted from Labor to KRF is about six times as large as the number changing in the other direction.

The general picture emerging from Table 7 suggests two main tendencies
in electoral changes between 1969 and 1977:

(a) A change along the left-right axis in favor of the Conservative party,
and

(b) A change along the religious axis in favor of the Christian People's
party.

These changes will be considered separately for the socialist parties and
for the bourgeois parties.

(1) The decline of the socialist vote. Taken together the socialist parties lost considerably to the bourgeois parties. They gained somewhat at the cost of the Liberals, while gains and losses in relation to the Agrarians evened each other out. Consequently, the losses went to the Christian People's party, the Conservatives and the new anti-tax party, FRP. The net loss of socialist votes went to these three parties in the ratio: FRP 1, KRF 2, and H 4.

A comparison of the Labor party and the leftwing SF indicates that both parties lost most heavily to the Conservatives. From a spatial point of view it is interesting to observe the differences in the patterns of change for the two parties. Labor has exchanged votes with all parties, to the left, at the center, and to the right. But people who left SF either went to the closest neighbor, Labor, or to some party on the right wing. Although the figures for the latter party should be interpreted with caution because of the low number of cases, the data suggest that SF/SV owes its existence purely to differences along the left-right axis, while Labor is related to other dimensions in the party system as well.

(2) The changes among bourgeois parties. A great number of respondents have shifted back and forth among these parties, two of which, the Conservatives and the Christians, have gained from the exchanges. The former party has enjoyed net gains in its exchanges with all other parties, both socialist and bourgeois, the greatest gains coming from the Liberals. The second greatest loser has been the Agrarian Center party: indicated shifts between the two parties are in the ratio of 2:1 in favor of the former. The shifts between the Conservatives and the Christian People's party are more even, with only a slight balance in favor of the Conservatives.

The Christian People's party has gained most from the Agrarians, followed by Labor and then the Liberals. The complicated patterns of change among the bourgeois parties are consistent with previous observations indicating that these parties differ along several cleavage dimensions (see Valen & Rokkan 1974b; Converse & Valen 1971).

To some extent the individual change reflected in Table 7 contributes towards understanding tendencies revealed by aggregate data (Table 5). Thus the fact that one out of four 1969 Liberals in the panel shifted to the Conservative party may explain why the latter has been particularly successful in the South-West region, the stronghold of the old Liberal party. Furthermore, the strengthening of the Conservatives in rural areas is likely to be related to the decline of the Agrarians. Finally, one may speculate about a possible relationship in urban areas between the decline of the socialists and the gains of the Conservatives. But considering the

fact that voters tend to move back and forth among several parties, one should hardly expect that aggregate data for specified geographical units will reflect all tendencies in electoral changes. Nonetheless, by combining the two types of data one is presumably in a better position to understand the conditions for political change. In the further analysis individual changes will be studied in a geographical or ecological context. For example, the character of change will be explored separately for urban and rural communities as well as for different regions.

7. First-Time Voters

As noted above, the reported panel study does not include data for people who became of voting age after 1969. There is no way of assessing the relative impact of changes among voters who became members of the electorate between 1969 and 1973. First-time voters in 1977, could not possibly have been involved in individual electoral changes at Storting elections. Nonetheless, this group may indeed have contributed to changes in overall party distributions from one election to another. Information about the behavior of first-time voters has been obtained through a nationwide voter survey after the election of 1977.⁷ Table 8 presents party choice of first-time voters in 1969 and in 1977.

Because of the low number of cases, the differences in Table 8 should be
considered with caution. Nonetheless, some main tendencies deserve to
be considered.

(1) The Conservative party, which was strongly favored by the electoral
winds of 1977, was particularly attractive to first-time voters.

(2) Two parties, the Agrarians and the Socialist Left party, which both

lost substantially in vote exchanges with other parties, managed to obtain greater shares of first-time voters than of older age groups. At the election both parties appealed to so-called 'green values', e.g. pollution control, decentralization and non-growth (economically). The differences reflected in Table 8 suggest that the two parties have been successful in mobilizing the support of young voters for such values. However, a third party that appealed to green values, the Liberals, was less successful in attracting first-time voters.

(3) The Christian People's party, which maintained a status quo between 1973 and 1977, has been less attractive to first-time voters than older voters. This tendency is consistent with a well known trend: support for the Christian People's party tends to increase with increasing age (Valen & Martinussen 1972, 252-7).

(4) Most surprising is the strong underrepresentation of the Labor party among the youngest voters. Previous research suggests that Labor used to be an attractive party for young people, although there have been deviations from this pattern.⁹ It will be recalled that despite the fact that Labor in 1977 recovered from its defeat in 1973, the party did not manage to regain its 1969 position. The party's relatively weak showing among first-time voters in 1977 must have contributed to this result.

(5) Table 8 indicates that at the election of 1977 first-time voters deviated relatively more than in 1969 from the rest of the sample. This finding is consistent with the tendency of recent years towards greater volatility in the electorate. Young people are more inclined than older ones to follow shifting election winds. (Valen & Martinussen 1972, 274-6); Butler & Stokes 1971, 65-89).

(6) The socialist parties taken together were in 1969 slightly overrepresented
among first-time voters, but in 1977 they were substantially underrepresented.
This tendency is consistent with the declining socialist vote.

8. Concluding Remarks

To a large extent the Storting election of 1977 marks a swing back to the situation that existed before the EC issue was placed on the agenda. Most important, a great number of EC opponents, who in 1973 had left the Labor and Conservative parties, returned to the fold. At the same time, the two extreme parties to the left and to the right, respectively SV and FRP (ALP), whose advance in 1973 had been largely due to the EC issue, suffered a sharp decline. But although much of the electoral change over these four years may be explained by the vanishing saliency of this issue,

the data presented in the preceding pages suggest that other forces must
have been at work as well: for example, the Christian People's party,
which had opposed EC entry, held on to its gains of 1973.

If Norwegian politics are considered in the longer time perspective of
1969-1977, it is evident that significant changes have occurred:

(1) The old Liberal party has been reduced to two miniparties.

(2) The socialist proportion of the total vote, which used to be a little over
50 percent, has declined by some four percentage points.

(3) The Conservative party has improved its position strongly.

(4) Likewise, the Christian People's party has increased its share of the
votes substantially.

Again, however, the tendencies in recent years of a declining socialist support and a corresponding strengthening of parties to the right are not exclusive to Norway. In fact, similar tendencies are evident, although with varying strength, in a number of nations: including all the Scandinavian countries. (Berglund & Lindström 1978, 174.) From a Norwegian perspective this similarity is indeed surprising considering the domestic turmoil generated by the EC dispute. Could it possibly be that the latter mainly served the function of triggering a process of change that was in any case destined to occur? In a comparative perspective one may wonder if recent electoral changes are caused by general political forces. Most likely, the tendencies reflect the impact of some common issues or shifts in ideological values. In this respect, the reactions of first-time voters in the Norwegian system may be symptomatic of a generational shift. At the end of the 19605, when student unrest and the youth protest movement reached their peak, 55 percent of first time voters supported some socialist party: in 1977 the corresponding figure was 42 percent (Table 8). Since young people tend to be particularly receptive to new ideas, these figures suggest a general ideological shift in a conservative direction. It would indeed be a challenging task for comparative analysis to study how the ideological values and ideas of our time manifest themselves in different societies and their political consequences. Even with a lack of adequate comparative data, valuable knowledge can be acquired through studying long-term as well as short-term tendencies in individual nations. The Norwegian case, with a data base that extends over a series of consecutive elections, ought to contribute towards a general understanding of these problems. The further analysis will fccus upon two main questions:

(a) Who have been changing party preferences, i.e. what characterizes
changers as compared to stable voters?

(b) Why did they change?

Obviously, the EC dispute still provides challenging problems for analysis: in which ways and to what extent are recent electoral changes related to this issue? Another issue which has caught a lot of attention during recent years is the question of abortion upon demand. Consistent with previous research, one may expect that this issue accounts for a major part of the increase in the Christian vote. But possible other effects ought to be examined as well; for example, the extent to which active pro-abortionists have mobilized support for specific parties.

However, the observed shift along the left-right axis from socialist parties to' the Conservatives cannot be explained as a result of either the EC dispute or the abortion issue. It is more likely that this tendency is related to social and economic change. For years the political debate has been dominated by questions relating to public activities and state control of business as opposed to private initiative. Thus at the 1977 election topics like taxation, state bureaucracy and bank 'democratization' or 'socialization' (depending upon the partisan viewpoint) were on the agenda. One may ask to what extent questions of this kind have contributed to the shift in a conservative direction. Equally interesting is the question of how various groups of voters reacted to these topics. During recent years Norwegians have enjoyed a substantial increase in net income and in standard of living. At the same time, and partly due to inflation, a sharply increasing proportion of the electorate has been hit by the relatively steep progressive taxation. The question is: to what extent is there a relationship between the individual's level of income and his/her reactions towards taxation and public expenditure? And to what extent have negative reactions in these areas been strong enough to make people change political loyalties? In a more general perspective these questions relate to central problems of the modern welfare society. Public expenditures are bound to increase all the time due to a flow of demands concerning education, health, pension schemes and social welfare, transportation, and other public services. Partly as a result of welfare measures, partly as a consequence of industrial development, the average level of education has tended to increase steadily, and so has the proportion of people in whitecollar jobs and middle class positions. The question is: what impact do these social and economic changes have on political constellations? How do they reflect upon the moderate socialist parties, which pioneered the welfare society? And in particular, what are the reactions of the youth, which has been brought up during the last two or three decades, and has experienced nothing but a society of social welfare, of growing individual opportunities, and of growing prosperity?

A few words should be added about the interesting changes at recent elections among the bourgeois parties: the growth of the Conservatives and - to a lesser extent - the Christians, at the cost of the Agrarians and the Liberals. Of course the high frequency of change among these parties reflects controversies concerning major issues like the questions of EC membership and free abortion. Nonetheless, it may look as if the voters' attachment to specific bourgeois parties has declined, or rather that bourgeois voters see fewer party differences than before within this family of parties. The fact that in the 1960s the four parties for the first time managed to establish a coalition government, indicates that former differences and barriers had been broken down, or perhaps that the party system had become more unidimensional. Previous research suggests that the leftright axis at this time constituted the dominating conflict dimension, and the four bourgeois partiews did not differ much from one another along this axis. (Converse & Valen 1971; Valen 1976; Berglund & Lindström 1978, 105-68.) During the EC dispute other cleavages became salient as well, but by 1977 when three of the; bourgeois parties declared their willingness to create a new coalition, it may seem as if the former spatial constellations had, by and large, been restored. Once more, differences along the left-right axis constituted dominating themes in the political debate, and the Conservative party became the most articulate spokesman for bourgeois policies along this dimension. The dominating position of the Conservatives is not surprising, considering that this party traditionally has been predominantly concerned with issues relating to the left-right cleavage, while the parties at the center have been relatively more concerned with other cleavage dimensions, the Christian People's party with moral and religious issues, and the Agrarian Center party with rural interests. Furthermore, the Liberals, which used to articulate a moderate position between left and right, had been split and no longer constituted a challenge to the Conservatives. As the most prominent spokesman in 1977 in favor of reestablishing an alternative to the Labor government the Conservative party profited greatly from vote exchanges among the bourgeois parties. One may wonder if the establishment of a coalition in the 1960s contributed to reducing differences among bourgeois parties in the perceptions of the voters. Possibly this event started a process towards realignment among these parties, although the impact of this process may have been delayed by the emotions of the EC dispute.

---

NOTES

1. On the EC dispute and the referendum, see Valen, 1973; Hellevik & Gleditsch, 1973; Valen 1976. On the Storting election of 1973, see Valen & Rokkan 1974; Valer & Martinussen 1977.

2. Throughout the exciting election night the result remained uncertain, but in the end the final projection seemed to be 78 socialist and 77 bourgeois representatives. However, the next morning the situation changed dramatically when the final count in the province of Møre and Romsdal indicated that the Christian People's party and not Labor had won the last and marginal seat. For a few hours the likely distribution of seats seemed to be: socialists 77, the coalition parties 76, and the Liberals 2. Thus the latter party, which had not committed itself beforehand to either alternative, held the balance, and it signalled support for a bourgeois government. However, late in the afternoon another surprise occurred: it appeared that some 50 votes for the SV had been omitted from the first count in the province of Nordland. The result was that SV took the last seat in Nordland from the Conservatives. Again the socialists held the majority, but now the result was final.

3. The method of proportional representation is a modified version of the Sainte Lagiie method, which was introduced in 1953. The votes for all party lists in a given constituency are first divided by 1.4, and the party which obtains the largest fraction takes the first seat. When a party has obtained one seat, the total number of votes for the party is divided by 3. After it has obtained two seats the number of votes is divided by 5, etc. On the electoral system, see Rokkan 1970.

4. The turnout level was rather stable between 1930 and 1961: at Storting elections on average 79 percent of the electorate participated. In 1965 the turnout level climbed to 85.4 percent, and in 1969 it was 83.8 percent. The light decline from 1965 to 1969 was due to a lowering of voting age from 21 to 20 years before the latter election. On the decline in turnout during and immediately after the EC dispute, see Valen 1973; Valen 1976 b.

5. Due to recent changes in communal boundaries the number of communes has varied somewhat from one election to another. In 1977 the number of units was 454. In order to obtain comparability, vote distributions have been estimated for prececing elections in communes affected by local reforms.

6. In a representative nationwide survey, originally 2000 persons were visited at the elections of 1965, 1969, and 1973.⁵ The members of the panel were again asked about their vote at the 1977 election, this time by a mail questionnaire. Approximately 1000 respondents, i.e. 90 percent of the active sample of 1973, responded to this enquiry. It should be noted that in the 1969 study a supplementary sample had been drawn for people who had become of voting age between 1965 and 1969, but in 1973 a similar adjustment for first time voters was not obtained. By 1973 the sample did not include respondents for the four youngest age brackets, and by 1977 the eight youngest age brackets were lacking. Nonetheless, changes in the electorate seem to be reasonably well reflected by the panel. The panel study is part of a Norwegian program of electoral research at the Institute for Social Research in Oslo. The program was started at the Storting election of 1957 by Professors Stein Rokkan and Henry Valen. Field work of all nationwide voter surveys from 1957-73 was conducted by Norsk gallup Institutt. On this program see Rokkan 1966; Valen & Rokkan 1967.

7. On party distances, see Converse & Valen 1971. On the patterns of change, see Valen & Martinussen 1972.

8. In a nationwide probability sample, some 2200 voters were visited immediately after i;he election. The study is part of the Norwegian program of electoral research. Sampling and field work for the study was conducted by the Interview Organization of the Central Bureau of Statistics.

9. One such deviation occurred at the Storting election of 1965. See Valen & Martinussen 1972, 252-257.

REFERENCES

Berglund, S. & Lindström, U. 1978. The Scandinavian Party Systems. Lund, Studentlitteratur.

Butler, D. & Stokes, D. 1971. Political Change in Britain. London, Macmillan.

Converse, P. E. & Valen, H. 1971. Dimensions of Cleavage and Perceived Party Distances
in Norwegian Voting'. Scandinavian Political Studies 6, 107-152.

Hellevik, O. & Gleditsch, N. P. 1973. 'The Common Market Decision in Norway: A Clash
between Direct and Indirect Democracy. Scandinavian Political Studies 8, 227-235.

Martinussen, W. & Valen, H. 1976. 'Makt og moral'. Tidsskrift for Samfunnsforsking 17,
19-46.

Rokkan, S. 1970. Citizens, Elections, Parties. Oslo, Universitetsforlaget.

Rokkan, S. 1966. 'Political Research in Norway, 1960-65'. Scandinavian Political Studies I,
266-280.

Rokkan S. & Valen, H. 1970. 'Regional Contrasts in Norwegian Polities', in E. Allardt & S.
Rokkan, eds. Mass Politics. New York, Free Press, 190-247.

Valen, H. 1976, 'National Conflict Structure and Foreign Politics: The Impact of the EC Issue
on Perceived Cleavages in Norwegian Politics'. European Journal of Political Research 4,
47-82.

Valen, H. 1976b. 'Norway: The Local Elections of September 1975'. Scandinavian Political
Studies 11, 168-183.

Valen, H. 1973. 'Norway: 'No' to EEC. Scandinavian Political Studies 8, 214-26.

Valen, H. & Martinussen, W. 1977. 'Electoral Trends and Foreign Politics in Norway: The
1973 Storting Election and the EEC issue', in K. Cerny (ed.), Scandinavia at the Polls.
Washington, American Enterprise Institute, 39-77.

Valen, H. & Martinussen, W. 1972. Velgere og politiske frontlinjer. Oslo, Gyldendal.

Valen, H. & Rokkan, S. 1974: 'The Elections to the Storting in September 1973'. Scandinavian
Political Studies 9, 205-18.

Valen, H. & Rokkan, S. 1974b. 'Norway: Conflict Structure and Mass Politics in a European
Periphery', in R. Rose, ed., Electoral Behavior. New York, Free Press.

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