The three 'core countries' of the North all experienced dramatic elections in the autumn of 1973. In both Norway and Denmark there was a further deepening of splits on the left and a new and disturbing division on the right, and in Sweden the electorate produced an exciting 'cliff-hanger' in giving exactly one half of the seats to the left-oriented coalition of Communists and Social Democrats and the other to the three-party phalanx of opponents to the right.

It is tempting to view these three elections as further confirmation of the underlying
structure of differentiation among the three core countries:

The Danes had the shortest period of Social Democratic dominance (coalition with Radicals 1929-40) and already saw their system suffer some severe fragmentation in the fifties. By contrast, in Sweden the Social Democrats have managed to stay in power practically continuously for more than forty years: they even survived the deadlocked election in 1973, and today the Swedish party system still retains its basic 2 + 3 structure.

Norway moved gradually from the Swedish to the Danish structure during the 19605. The long period of Labor (Arbeiderparti) dominance was broken in 1961 and the two rounds of negotiations and controversy over entry into Europe (EEC) deepened the divisions on the left. This process was brought to a climax in the Referendum in September 1972: the 'No' front won a resounding victory, and so accelerated the process of fragmentation. The 'No' victory in fact split the party system on the left as well as in the center and on the right: the result was a fragmented Parliament and another period of minority government, this time under the Labor party.

1. The Electoral Consequences of the European Issue

The 'No' to Europe left the country in an extraordinary situation. The three leading 'Yes' parties - Labor, the European Liberals, and the Conservatives - refused to take part in any government, so the country was left in the hands of a 'mini-coalition' of two and a half parties - the Agrarian Center, the Christians, and the 'No' Liberals

- which managed to negotiate a treaty with the 'Nine' and push it through to final ratification before the new election in 1973. This alliance, however, could not hope to continue once the European issue had been settled: they controlled only 38 of the 150 seats in Parliament. All they could hope for was some improvement in their total strength but nothing was very certain: the total structure of alignments and coalition options had undergone profound change during the year since the 'No'.

The most spectacular change was the split within the oldest of the political parties in Norway: Venstre (the Liberals). They had been split at the parliamentary level during the European Campaign (9 'Yes', 4 'No') but the national party leadership had tried desperately to keep the actual mass organization united. With victory to the 'No' splinter the structure broke down. At a national meeting of party representatives at Røros in November 1972, the pro-Europeans decided to walk out and form a party of their own: this wing eventually decided to call itself the New People's party (DNF) but otherwise claimed to be the true inheritors of the Venstre tradition. As a result, the movement which led the opposition to victory in 1884 and produced a long series of governments until the 1930s came close to committing suicide at the election in 1973: there were DNF lists opposing Venstre lists or joint Venstre-Agrarian in all 19 constituencies, and this caused severe losses to both wings.

To the left of the party system, there was a merger as well as a split. The 'No' front within the Labor party split off and formed a movement on its own - the Workers' Information Committee (AIK) - and then moved into a loose electoral merger with both the older splinter movement of 1961 - the SF - and the smaller Communist party. The result was a remarkably successful Socialist Electoral Alliance (SV): the first example of a united left front in Norwegian politics. This party gained a total of 16 seats and acquired substantive leverage on a Labor party dependent on these votes in a majority coalition of 78 over 77.

The greatest surprise came on the right, however. The defeat of the European cause had triggered widespread resentment within the Conservative party and the loss of legitimacy made it even more tempting than before to try out alternative alignments. A high proportion of the voters on the right had been frustrated by the failure of the non-socialist governments since 1965 to bring about any reduction in the levels of taxation, whether direct or indirect. This made it possible for an old nationalist conservative, once an active leader in the anti-party 'Patriot League' of the thirties, to rally support for a party of his own - Anders Lange's party - against taxes and public expenditure. Professional politicians at first tended to treat this new development as a joke, but things changed on election night: the new party, a purely personcentered movement without any organizational basis at the time, got as many as 106,000 votes and gained 4 seats in Parliament.

These four anti-tax members could not be counted on in a broad coalition against Labor: they were 'outsiders' and could not be trusted to vote consistently with the others. Even if the Left-Labor flank had not reached the majority point, it would have been most difficult to establish a government of non-socialist parties.

2. Overall Results

With all these changes in the system of party alternatives, there was great excitement both during the campaign and on election night itself. The results hovered close to the majority point for hours and the final count gave the most fragmented Parliament in Norwegian history (Table I).

There were three main victors in this contest: the new Socialist Alliance (SV) on the left, the Christian People's party, and the anti-tax party on the right (ALP). All three gained a total of 26 seats - a record number of gains in any one election. The main losers were the Labor party and the two wings of the old Venstre: Labor lost 12 seats, the Liberal wings 10. Two parties remained at roughly the same level of strength: the Agrarian Center gained only one seat, while the Conservatives lost some votes but otherwise kept their 29 seats.

3. Results by Constituency

Table II shows the distribution of gains and losses across the 19 constituencies.1

It can be seen that the Left Socialist gains correspond roughly with the Labor losses, which were heaviest in Telemark - a traditionally radical district dominated by large industry - and in the far north. Practically all the SV gains in seats were at th& expense of the Labor party. There were two sets of exceptions only: in the central constituencies of Akershus and in Oslo where Labor did not actually lose seats, but failed to gain any of the new ones allocated to those districts in 1973, and then in the two constituencies where the new alliance gained two seats right away, in Sør-Trøndelag (where the Agrarians also lost a seat) and in Nordland (where both Venstre and the Conservatives lost).

The two wings of the old Liberal party suffered spectacular losses in the south and west - the traditional strongholds of counter-cultural politics. Most of these seats went to the Christians and the Agrarians, but the actual shifts at the voter level were no doubt much more complex. Calculations indicate that the two wings of the old Left

would have gained eight seats if they had stayed together: they would have remained above the threshold in Akershus and Oslo, as well as in the south and west and even in the north. The dramatic losses came about because both parties fell below the threshold in so many constituencies: the SAINTE-LAGUE divisor of 1.4 for the first seat does offer a barrier against excessive splits.

4. Results by Type of Commune

A full analysis of shifts at the micro-level is not possible at this stage: the Central Bureau of Statistics has not yet (February 1974) made available the tapes for all results by commune, and the survey data collected before and after the election have so far only been subjected to superficial analysis.

In order to give a few first impressions of the likely findings of detailed analyses, we shall present, first, a table showing variations by type of commune published by the Central Bureau of Statistics² and then a set of findings from surveys carried out by the Bureau3 and within the Norwegian program of electoral research.⁴

Table 111 gives the distribution of votes by type of commune: unfortunately this
analysis covers the entire territory and does not allow regional differentiation. The
trends in the data are fairly clear, however:

The Left Socialist wing clearly gained most in the fisheries districts and in the industrial areas in the periphery: in our terminology the European issue tended to increase the chasm between the 'peripheral' proletariat of fishermen, smallholders and isolated industrial workers, and the 'central' industrial working class.⁵

The differences between the two wings of the old Venstre party also come out quite clearly in the table: the 'No' yang turns out to be very similar in its strengths and weaknesses to the Christians; the 'Yes' wing is much more heavily concentrated in urban areas.

The split-off on the right is much more difficult to analyze in ecological terms: the
Anders Lange party presents a profile very similar to the Conservatives at this level of
aggregation.

The Labor and Venstre splits both reflected increased tension between the central and the peripheral areas: this comes out clearly in a comparison of SV vs. Labor strengths and in a comparison of the 'Yes' and 'No' wings of Venstre. This reaccentuation of the center-periphery polarity in Norwegian politics was a direct result of the protracted conflict over entry into the EEC. Tables published by the Central Bureau of Statistics show that the two parties which suffered greatest losses in 1973 were at the same time the ones most evenly split on the European issue: Labor 65-35% in 1972, Venstre 42-58%. What is particularly interesting in the figures from the Bureau survey is that the split-off to the right seems to reflect tension over the European issue: the Conservatives voted overwhelmingly 'Yes' while voters for the Anders Lange party were fairly evenly split 45 to 55% on this issue. This contrast clearly reflects less a geographical center-periphery split but does point to an underlying difference in social and economic position: the Anders Lange voters less integrated, the Conservatives more integrated into the networks of large-scale enterprises most likely to benefit from entry into Europe.

5. Changes at Voter Level: 1969,1972,1973

The Norwegian electorate has clearly been on the move for some time. Opinion polls indicate increased volatility from the summer of 1970 onwards:⁶ it was at this time that the EEC came back on the agenda after the lull since the General said 'No' in January 1963.⁷ There was clear evidence of a widespread mobilization of opposition to Europe: this came out even clearer at the local elections in 1971.⁸

Analytically, we can distinguish three types of changes at the voter level:
- changes in party preferences among active voters voting at both elections;
- changes in party strength due to shifts between voting and non-voting: gains due

to extra spurts of mobilization or losses due to lower levels of turnout;

- changes in the balance between parties due to the continuous renewal of the elec

torate: shifts in the distributions for young people entering the elections as against

shifts due to differential losses through death and emigration.9

On the basis of a nationwide probability sample interviewed at the elections of 1969 and 1973, an attempt will be made to study individual changes between the two elections ¹⁰ Some 1200 respondents were interviewed at both. No additional sample was drawn for citizens who came of voting age after 1969. Furthermore, the data are not yet ready for an analysis of shifts between voting and non-voting: we have not yet been able to check the sample against the electoral register. Consequently, we shall focus this first analysis on the changes in party preferences of 'active' voters: among respondents who stated that they voted at both elections.

As might be expected, the frequency of individual changes was very high: 28% of

our 'active' respondents indicated that they changed party between 1969 and 1973. If we include reported non-voting, the proportion of those who changed was 33.5%.ⁿ Corresponding figures for changes between the two preceding elections (1965 and 1969) were 18 and 24%.ⁱ 2i2

First we shall consider the changes in overall party distributions within our panel. Since we do not include first-time voters and shifts between voting and non-voting, we cannot expect to obtain distributions which would approximate those of the total electorate. Nonetheless the tendencies indicated in Table IV are quite similar to the

changes which actually took place (cf. Table I). The proportions of votes for the Labor party, the Liberal party and the Conservative party declined, while the Christian People's party, the Center party and the Socialist Election Alliance (or rather the parties constituting this alliance) increased their share of the vote. Even the degree of change for each party comes close to the differences from 1969 to 1973. A notable exception is the Labor party, which lost more in the aggregate than in our sample of 'active' voters. Furthermore, the ALP obtained support from only 2.5% of the active voters in our panel as compared with 5% for the total electorate. Our data suggest that the discrepancy between the two parties is largely due to the impact of shifts between voting and non-voting. However, since the changes described in Table IV come so close to the overall election results, we can assume that the data by and large reflect the tendencies which actually occurred in the electorate.

The aggregate figures presented in Table IV reflect a variety of shifts at the individual
level.

Table V offers a summary of some of the major tendencies in the transfer of votes
from party to party. The great number of shifts between Labor and left-wing parties

fc

demonstrated in the table is not a new phenomenon. Similar changes back and forth between the socialist parties have been observed over a number of elections.¹³ In 1973, however, the changes went almost exclusively in one direction: from Labor to the left. Only a few voters changed between the non-socialist parties and the parties to the left of Labor; this tendency is also consistent with earlier findings. The changes between these groups were in fact slightly higher than usual in 1973. Similarly, shifts between Labor and non-socialist parties were greater in 1973 than normal. On balance Labor lost substantially in all directions. Table V does not present shifts among the non-socialist parties: these were remarkably large. 44.7% of our active respondents voted for one of these parties at both elections, but more than one quarter of them changed from one election to the other. The Liberals had a net loss to all non-socialist parties as well as to Labor. The Conservatives lost to all non-socialist parties except the two Venstre wings. Finally, the Agrarians gained votes from the Venstre wings and the Conservatives as well as from La.bor, and the Christians enjoyed a net gain from the same three groups as well as from the Agrarians.

The aggregate figures suggest that most of the changes in 1973 reflected reactions to the EEC stand taken by the parties. To test this, we have analyzed the shifts separately for supporters and opponents of EEC membership. The respondents were asked about their position on EEC at the time of the election.¹⁴ Around 40% said 'yes': this is lower than the result at the referendum. One result stands out with great clarity: the opponents of the EEC were far more inclined than the supporters to change party from 1969 to 1973. This confirms a tendency observed in earlier data. Of EEC supporters who voted at both elections, 76.3% voted for the same party. If we include shifts between voting and non-voting, only 72.5% remained stable. The corresponding figures for EEC opponents were 67.3 and 60.4% respectively. To allow detailed analysis of the character of these changes, transition matrices for party choice between 1969 and 1973 will be presented for supporters and opponents separately.

The importance of the EEC issue is clearly reflected in the split-up of Venstre: supporters of EEC switched almost without exception to the DNF (the New People's party), while approximately three out of four EEC opponents remained in the majority Venstre. The trends are less clear in the shifts between other parties. If the EEC issue had had a major impact on the 1973 vote, it would have been expected that parties in favor of EEC would have strengthened their position among EEC supporters but lose among EEC opponents. For parties opposing EEC the opposite tendency would have been expected, i.e. decline among EEC supporters and gains among opponents. As far as opponents are concerned the data indeed support this hypothesis. Table VI shows that the proportions of votes for the Labor party and the Conservatives were substantially reduced in this group, and that the anti-EEC parties, the Agrarians, the Christians and the Socialist Election Alliance gained correspondingly. Among supporters of EEC, the changes in party distributions are far less spectacular (Table VII), even contradictory in some cases. The Conservatives did enjoy some gains while the Agrarians suffered a moderate decline among supporters of the EEC, but the trend seems to have been the reverse for the Socialist parties: Labor in fact declined by 1.6%, while the left-wing Socialist Alliance increased its share of the votes by approximately 1%. These findings indicate that the EEC issue had a much stronger impact upon the electoral choices of its opponents than of its supporters. The fact that the Anders Lange anti-tax protest party drew a large proportion of its vote from EEC supporters does not contradict this general conclusion: the figure for the ALP share is lower in the sample of active voters than in the electorate and so suggests that the party drew disproportionate strength from groups who had not voted in 1969. What is interesting is that the ALP seems to have drawn a roughly equal share of its vote from both Labor and the Conservatives.

A full analysis of transition matrices is not possible at this stage. A few trends in the
shifts among the established parties deserve comment.

The Labor Losses

Here we observe an interesting contrast. Among EEC supporters only a tiny proportion of left-wing voters have moved over to the Socialist Election Alliance. The bulk of defections in this group went to the parties to the right, to the Conservatives and to the ALP. Among anti-EEC people Labor lost largely to three parties: the Socialist

Election Alliance, the Agrarian Center and the Christian People's party. These tendencies suggest that Labor lost support along several of the dimensions which were salient during the EEC debate: presumably most Laborites who went to the left-wing Socialist Alliance were either urban radicals or voters in coastal peripheries, while the bulk of those who shifted to the two non-socialist parties must have been smallholders or farmers. As we saw from the table for results by type of commune (Table IV), the SV was strongest in peripheral industrial and in fisheries communities: our survey data suggest that this was where the greater part of Labor losses occurred. But the Christians also made inroads on Labor in the peripheries. In this case the EEC issues may not have counted as much as the position of the Labor party on the liberalization of the abortion law: this was a salient issue in the campaign.

The Liberal Losses

The data for the Venstre wings are interesting not only because of the split-up over Europe but also because of the directions of losses at voter level. A number of EEC supporters among the Liberals switched to Labor and the Conservatives, while most of the EEC opponents who did not stick to the reduced Venstre voted for the Agrarians and the Christians. This supports our earlier interpretation of the tensions within the Venstre party: the EEC polarized it on the center-counterculture cleavage.¹⁵

The Christian Gains

The fact that the Christian People's party enjoyed more marked gains than the Agrarians was indeed one of the major surprises at the 1973 election. The voters may to some extent have been influenced by the fact that Mr. Korvald, the leader of the Christian party, was Prime Minister of the 'mini-coalition' which negotiated a trade agreement between Norway and the Common Market, and had been successful in keeping his party united despite a strong pro-EEC wing. The abortion issue may have had a general impact in addition to the concerns over EEC. Among EEC supporters the Christians maintained their position, while the proportion of Agrarian voters declined. Among EEC opponents the Christians enjoyed a substantial net gain from Labor as well as from all the non-socialist parties including the Agrarians. It is worth noting, however, that Conservative opponents of the EEC were more likely to defect to the Agrarian party. This finding suggests that Conservative elements opposed to the EEC were attracted to the Agrarians, but that voters in the traditional countercultures were more inclined to prefer the Christians.

Our conclusion must be that the ELEC issue had a definite impact on electoral choice in 1973. Further analysis is required, however, to determine to what extent reactions toward the EEC interacted with issues of a purely domestic character. Furthermore, the pattern of shifts between parties suggests that voters moved along different conflict dimensions. Some of the shifts may be accounted for by issues in the 1973 campaigns. Presumably this was not frequently the case for pro-EEC voters. More often the shifts seem to follow cleavages which were salient during the EEC controversy. In the detailed analyses to be undertaken later it will be essential to study the shifts in party choice as well as shifts between voting and non-voting within subsets of the panel, particularly the subsets which can be contrasted along the major conflict dimensions of the system.¹⁶

6. A Note on Possible Comparisons Within Scandinavia

The three elections held in such close succession in the autumn of 1973 invite thorough
comparisons on a number of points. We cannot go into complete detail at this juncture
but would like to point out some interesting possibilities.

What makes Scandinavia such a fascinating field for comparative research is the commonalities in basic political culture and the contrasts in the sequences of development towards full-fledged mass democracy as well as in the current alignments within the international system.

The two simple two-dimensional diagrams depicted here help to summarize the differences.

The contrast between Swedish and Danish developments toward full-suffrage mass politics has received considerable attention in the literature: in Sweden near-continuous dominance of the Riksdag and a gradual domestication of the lower strata through step-by-step extensions of the suffrage; in Denmark a long period of absolute monarchy, a sudden introduction of near-manhood suffrage in 1849, and a long and difficult struggle for supremacy between Parliament and Cabinet.¹⁷ The Norwegian development fell somewhere in between: absolute rule under the Danish kings until 1814, a remarkable democratic constitution that year and then a succession of extensions of the suffrage from 1884 onwards. As Tingsten has shown in his classic treatise

on Swedish Social Democracy,¹⁸ Norway shared with Denmark an early development of an urban intelligentsia and a greater openness to the radical secularism spreading from Europe: Sweden was less integrated into the European urban network and was much more of an 'agrarian bureaucracy' in Barrington Moore's terminology.¹⁹

The consequences of these initial differences in the development of mass politics can
be traced through several steps:

First, a much more sudden mobilization of the peasantry in Denmark and in Norway, and a much smoother incorporation in Sweden; secondly, much stronger radicalagrarian fronts and longer periods of Liberal (Venstre) dominance in Denmark and Norway, with hardly any such front in Sweden and a much shorter 'Liberal transition'; thirdly, an earlier radicalization of the intelligentsia in Denmark and Norway and greater problems of integration within the growing labor movement; smoother domestication of the rising Social Democrat party in Sweden and much greater integration at the elite level.²⁰

In the years after World War II these differences in early developments combined
with the differences in international alliances to produce contrasts in the strains on the
established party systems:

- in Denmark and Norway, violent conflicts over the Atlantic Alliance to the left
of the Labor movement; in Sweden very little disagreement on the policy of neutrality;

- in Denmark and Norway, great dissensions over the question of entry into the
EEC; in Sweden again quite manageable differences.

Largely as a consequence of these difficulties over foreign policy, the Danish and Norwegian Labor parties lost their dominant positions and had to cede to coalitions of non-socialist parties for several periods. These experiences of 'bourgeois' rule brought small consolation to the ardent anti-socialists, however: the coalitions to the right could neither change the structure of the welfare state nor lower the taxes appreciably. The result was the 'poujadism' of Glistrup in Denmark and Lange in Norway. In Sweden there was no such splinter on the right: the Conservatives did not take part in any coalition and as a consequence could absorb within its ranks a lot of the anti-tax, anti-welfare sentiments which were channeled into separate parties in Denmark and Norway. In Sweden the Social Democrats had been in power for more than a generation and could be made the sole target for all the anti-tax protestations. In Denmark and Norway both the established blocks - the non-socialists as well as the Labor party - could be blamed: the protest movements had to find new channels.

Such waves of frustration and protestation can obviously find many expressions. The British election in February 1974 provided an example of an old party - the Liberals — re-emerging on a wave of resentment against the policies of the dominant parties. The three Scandinavian elections in the autumn of 1973 offer exciting data for comparative analysis, not only within Scandinavia but also across a wide range of competitive systems in the throes of conflict over policy priorities in the welfare state. There is a great deal of evidence of a widespread 'defreezing' of the alignment structures of the fifties and early sixties and. of increasing volatility within national electorates. This opens up challenging tasks for comparative political analysis. In the Norwegian case we cannot come to a final conclusion in our search for explanations of the great changes from 1969-725 without extensive comparisons with other countries. It is easy enough to assert that the EEC conflict can explain the change but to test this we have to review the evidence across several countries. What we want to know is how far the EEC was an issue in its own right and how far it primarily acted as a catalyst for deeper tensions which had been latent in the system since the beginning of mass politics.

University of Oslo

University of Bergen

Henry Valen

Stein Rokkan

---

NOTES

1. The city of Bergen constituted a separate fylke until 1971; it is now part of the surrounding fylke of Hordaland.

2. Statistisk ukehefte, 46/73, 14 Nov. 1973, Table 7.

3. Statistisk ukehefte, 51/52/73, 19 Dec. 1973, Tables 4-5.

4. Henry Valen, 'Norway: "No" to EEC, pp. 214-226 in Scandinavian Political Studies, Volume 8/73. Oslo, Universitetsforlaget, 1973.

5. H. Valen and S. Rokkan, 'Cleavage structures and mass politics in a European periphery: Norway', in R. Rose (ed.) Electoral Behavior: A Comparative Handbook. New York, Free Press, 1974.

6. H. Valen & W. Martinussen have carried out a secondary analysis of changes in party preferences during 1969-72 basedon data from the Gallup Institute. The data came from the nationwide probability samples interviewed every month. See H. Valen & W. Martinussen. Velgere og politiske frontlinjer. Oslo, Gyldendal, 1972, pp. 329-360.

7. See especially S. Rokkan and H. Valen, 'Regional contrasts in Norwegian polities', in E. AUardt & S. Rokkan (eds.) Mass Politics. New York, Free Press, 1970.

8. H. Valen, 'Norway: Local elections in the shadow of the Common Market', pp. 272-282 in Scandinavian Political Studies, Volume 7/72. Oslo, Universitetsforlaget, 1972.

9. For an elaborate discussion of the sources of electoral change, see D. Butler and D. E. Stokes, Political Change in Britain. New York, St. Martin's Press, 1969, pp. 275-292.

10. A nationwide sample of voters was interviewed before and after the election of 1965 and 1969 and after the election of 1973. The study is part of the Norwegian program of electoral research: on this program see S. Rokkan, Scandinavian Political Studies, Volume 1/66, pp. 266-280; H. Valen and S. Rokkan, 'The Norwegian program of electoral research', Scandinavian Political Studies, Volume 2/67, pp. 294-305. Major publications within the program: H. Valen and D. Katz, Political Parties in Norway. Oslo, Universitetsforlaget, 1964; S. Rokkan, Citizens, Elections, Parties. Oslo, Universitetsforlaget, and New York, D. McKay, 1970; H. Valen and W. Martinussen, Velgere og politiske frontlinjer, op. cit.; and W. Martinussen, Fjerndemokratiet. Oslo, Universitetsforlaget, 1973.

11. This figure may have to be slightly revised when we have checked reported non-voting against the public electoral register.

12. See Valen and Martinussen, Velgere og politiske frontlinjer, op. cit., pp. 300—328.

13. For changes from 1965-69, see Valen & Martinussen, Velgere og politiske frontlinjer, op. cit., pp. 298-328.

14. The question was: 'Let us assume there were to be held a referendum on EEC membership tomorrow. How would you vote, would you vote "yes" or "no" to Norwegian membership in the Common Market?'

15. Valen & Rokkan. 'Conflict structure and mass politics in a European periphery', op. cit.

16. An analysis along these lines will be presented in a monograph by Henry Valen and Daniel Katz on Norway and the EEC: this is currently in preparation.

17. Stein Rokkan. 'The growth and structuring of mass politics in W. Europe', Scandinavian Political Studies, Volume 5/70, pp. 65-85, especially pp. 67—68, also his attempt at constructing a 'conceptual map' of Europe in his chapter 'Dimensions of state formation and nation-building' in Charles Tilly (ed.) The Formation of National States in Europe. Princeton, Princeton Univ. Press, 1974. For further comparisons of the Scandinavian countries see W. Lafferty, Economic Development and the Response of Labor in Scandinavia. Oslo, Universitetsforlaget, 1971, and Stein Kuhnle, Social Mobilization and Political Participation in the Scandinavian Countries 1850—1970. Bergen, Institute of Sociology, 1973.

18. Herbert Tingsten. Den svenska socialdemokratins historia. Stockholm, Tiden, 1941, Vol. I, pp. 13-24.

19. Barrington Moore, Jr. The Social Origins of Democracy and Dictatorship. Boston, Beacon, 1967, esp. Ch. 7.

20. This hypothesis has been developed by Bernt Hagtvet in a recent paper: he is currently assembling evidence for Norway and Sweden.