University of Oslo

1. Introduction

The role played by regional contrasts in Norwegian politics has been pointed out by several authors. In particular Stein Rokkan has underlined the importance of certain persistent differences in cultural orientations, referring to, among other things, referendum statistics showing very clear variations in the regional distribution of the vote in favor of alcohol prohibition. But he adds a reservation:

We have been able to assemble a variety of information on the overall contrasts between the regions and among the provinces within the regions, but to determine the importance of these differences in cultural conditions locality by locality has proved highly problematic. Our findings, therefore, can hardly be more than suggestive of promising lines of further inquiry.₁

More recently William M. Lafferty, in his book Economic Development and the Response of Labor, has compared the referendum statistics pertaining to the temperance question with other Norwegian referendum data as well as electoral statistics and roll-call data from Labor Party conventions.² He has followed Rokkan's suggestion and, in addition to studying referendum and electoral data at the level of the twenty provinces of the realm, has taken into account certain variations which can be observed between smaller localities. On the other hand he does not take into account the regional point of view.

In our opinion Lafferty's book merits a detailed discussion. He studies a question of general interest - the cause of labor radicalism - showing great imagination in the use of a number of techniques and in the search for possibly relevant data. However, it seems to us that the addition of a regional viewpoint will make the phenomena he has analyzed appear in a somewhat different light. In accordance with a proposition stated originally by Edvard Bull, Lafferty sees the strong radicalism of the Norwegian labor movement as an effect of industrialization. He explains it by referring to what he perceives as a connection existing generally between rapid economic development and cultural disruption. In our view, however, the contrast in cultural traditions between different parts of Norway makes it impossible to apply one single explanatory scheme.³

In what follows we shall first discuss in some detail the analysis performed by William Lafferty and then present our own view of regional contrasts in Norway during the period in question, i.e., the first quarter of the present century. We are particularly interested in the use made by Lafferty of referendum data, because regional differences easily acquire more importance in referendums than in elections. At an election of representatives to a legislative body the parties will have to fight within the separate constituencies for the seats to which each is entitled; this means that the antagonism existing between different groups within every constituency will inevitably be accentuated. In a nationwide referendum, on the other hand, one particular question is to be answered by voters in the whole country - the struggle is decided not by the majorities obtained in the separate constituencies but by the majority in the country as a whole. In consequence, it is rather the similarity of views that may exist within a constituency, in contrast to other constituencies or other regions, which will be accentuated by the advocates of one side or the other whenever they sense that a feeling of regional solidarity is likely to favor their cause.

2. Economic and Norm-related Factors

William Lafferty starts by studying two roll-calls - recorded at the Labor Party Conventions of 1918 and 1919, respectively - which determined the general principles that were to guide party work. He notes the distribution of votes for and against a revolutionary program within each of the delegations from different parts of the country. Those delegations which influenced the result most strongly in a revolutionary direction are termed the most radical, while those tending to influence it most strongly in the other direction are designated as the most moderate. Lafferty finds that there were, in all, 19 outstandingly radical and 12 outstandingly moderate delegations. On the basis of a study of the referendum statistics of 1905 and 1919 he then reaches the following conclusions: in the most radical localities - i.e., those sending the most radical delegations to the Labor conventions - voters proved to be on the average more strongly against alcohol prohibition in 1919 than were voters in the localities sending the most moderate delegations. Similarly, in 1905 there had been a nationwide referendum on another question, which revealed a similar pattern. What was then at stake was the issue of monarchy versus republic. It turns out that the average percentage in favor of a republican form of government had been a good deal higher in the most radical than in the most moderate localities (as defined above).

Lafferty sees this as indicating the existence of a 'generalized, norm-related factor'.⁴ The strong radicalization which occurred within the Norwegian labor movement during and after World War I has been interpreted as an instance of the disruptive consequences of rapid economic development. Lafferty's hypothesis is to the effect that 'isolated economic processes were not the decisive element in explaining voter-radicalism but rather... cultural and norm-related factors'.⁵ In this connection he refers to various phenomena, such as the extent

of migration from one Norwegian province (fylke) to another and the percentage rate of illegitimate births; but his reference to the nationwide referendums is of particular interest in our opinion. The analysis of the strength of monarchical or anti-monarchical attitudes in 1905 throws light on the question whether radical and moderate tendencies were in existence in the same localities even before the great wave of industrialization which occurred in Norway from about 1906 until the end of World War I. And the analysis of the prohibition referendum is seen by Lafferty as particularly revealing: 'In the radical areas ... it seems that the drinking issue revolved around... the staving off of anxiety associated with discontinuous cultural norms'. The high proportion of votes cast against prohibition in radical areas represents 'a measure of social disruption'; on the other hand, the opposite situation prevailed in the strongly moderate areas: 'Social disruption seemed nonexistent and voter preferences on the republic/monarchy and drinking question were both in the conservative direction'.⁶

3. Some Basic Questions

As Lafferty himself takes care to emphasize, his book has an exploratory character
7 The subject studied is a complex one, and in order to evaluate the analysis
we shall have to proceed step by step. A number of questions arise:

a. The Norwegian Labor Party had 95 local branches, covering the
whole country, which sent delegations to the National Labor Convention.
Lafferty has selected 31 of these. Are his criteria of selection tenable?
b. In the corresponding constituencies, the reactions of the electorate at
two referendums are compared to degrees of radicalism or moderation
within the ranks of the Labor Party. To what extent is it legitimate to
presume that the reactions of socialist voters were identical with the reactions
of the electorate as a whole?

c. It is supposed that every affirmative vote cast at a referendum is an
expression of the same attitude on the part of every voter, and correspondingly
with regard to negative votes. Is the assumption tenable?

In addition to the above questions concerning presuppositions, it is also necessary
to ask:

d. Are the calculations performed on this basis meaningful?

We shall consider these questions one by one.

a. Lafferty restricts his analysis to 31 localities, those represented by the 19 most radical and the 12 most moderate delegations; or perhaps it is better to say 'the most effective radical' and 'the most effective moderate'. The result of a vote at the Labor Party Convention could be less strongly influenced by a small delegation (say, of 3 or 4 members, even if they all voted for either a radical or

a moderate proposal) than by a large delegation representing a populous locality

- even if its members were partly for, partly against the proposal. Lafferty is aware of the problem and tries to diminish it by omitting the capital from his analysis (the Oslo delegation disposing of no less than 70 and 90 votes at the 1918 and 1919 conventions, respectively).⁸ But there were also other large delegations. Even if their size was not to be compared with that of the Oslo group, they could influence the final result more strongly than the degree of radicalism or moderation existing within these delegations would seem to indicate.

It is also a question whether the degree of radicalism/moderation within a delegation corresponded to the attitude of socialist voters in the locality which it represented. Lafferty controls for this by looking at the record of votes cast in his 31 localities at the parliamentary election of 1921. By then the moderate wing had split off from the Norwegian Labor Party under the name of Social Democratic Party. Socialist voters could now choose between a moderate party and a radical, revolutionary one. Lafferty does find a certain correspondence between the degree of radicalism or moderation expressed by voters in 1921, and the degree of radical or moderate influence exercised by delegations from the same localities at the Conventions of 1918 and 1919.⁹ In niral as well as urban districts his most radical localities are above the national mean as regards the relation between NLP and SDP votes (whereas the moderate ones find themselves equally below the mean). The relationship was not quite as striking as the table makes it appear to be, however. In fact the radical superiority in the 'exceptionally radical' towns (2.3) was just barely in excess of the mean for all of the country's urban communites. Through an oversight, Oslo and Bergen have not been included in the figure of 1.5 given by Lafferty as representing the national urban mean. 2.0 (NLP twice as strong as SDP) is the mean for all towns. In other words, the voters in his radical localities do not seem to have been so very radical after all. We could say, using Lafferty's terminology, that there was not a very good correspondence between degrees of party-sector and voter-sector ideological preference. But perhaps it would be better to use other words. The lack of close correspondence may be due to the fact that the radicalism/moderation of delegations is expressed in absolute numbers (the surplus of radicals or of moderates in each delegation), whereby a big delegation may well come to weigh more heavily than a small one in a way which is excluded with regard to the radicalism/moderation of voters, the latter being expressed through the relation between NLP votes and SDP votes in each locality. Or, to look at the matter from a different angle, the surplus of radicals or moderates in a delegation does not represent a reliable measure of its 'degree of ideological preference', only of its degree of ideological influence at the conventions.

To sum up, it is not certain that an exceptionally radical laboring population was to be found in Lafferty's 19 localities. But at least they seem to have been somewhat above average in radicalism, as measured by the electoral results of 1921. We can therefore feel relatively secure at least in following Lafferty further when he shifts his focus to a study of referendum data from 1905 and 1919.

b. There is another pitfall, however. Even if a majority of the votes cast in a

certain locality was clearly in favor of prohibition, a majority of Labor Party voters may have been against it, and vice versa. Lafferty points out that this was unlikely to be the case when a large part of the electorate consisted of Labor Party adherents, which was the case in very many of the 31 localities selected by him for study. Still some uncertainty remains. There is a possibility of control, however, insofar as the largest towns are concerned. These were divided, during the period 1906-1918, into two or more single-member constituencies. The electoral results in each constituency can be compared with the referendum results of 1919. The latter are not to be found in the publication issued by the Central Bureau of Statistics and used by Lafferty, but they are available partly in municipal statistical records, partly in contemporary newspaper reports. From these data it is apparent that attitudes with regard to prohibition in Labor Party strongholds were not significantly different from attitudes recorded in other constituencies of the five towns concerned. The average anti-prohibition vote was highest in the capital and the nearby town of Drammen; then came the country's second largest town, Bergen, and the third largest, Trondheim, while Stavanger on the south-western coast ranked lowest among the five as regards the antiprohibiton vote:

It appears that the prohibition of alcohol was somewhat more in conformity with working-class than with middle-class norms, but the differences between Labor constituencies and bourgeois constituencies were much smaller than the differences between the various towns. Each one of these represents one of Lafferty's nineteen 'most radical localities' (except Oslo, which he excluded from his sample). Table I seems to indicate that it is possible to accept his presupposition with regard to the referendum data; reactions among the electorate as a whole seem to reflect pretty well reactions among the socialist part of the electorate, although the latter showed a certain prohibitionist bias compared to the rest of the urban population.

c. Answering 'yes' or 'no' to the question posed at a referendum does not always mean the same thing to one voter as to another. Lafferty is aware of the fact but regards it as unimportant. The 1905 referendum, which concerned the question of whether Norway should still remain a monarchy after the dissolution

of the Union with Sweden or adopt a republican form of government, he considers a clear 'radical/moderate test'.¹⁰ However, it was also a test of the voters' confidence in their government and, furthermore, nationalist feelings were certainly involved. When Norway's political leaders, in the spring of 1905, decided to secede from the Union, Sweden refused to accept the separation unless certain conditions were fulfilled. These included a demand for the demolition of Norway's fortresses along the common border. After prolonged negotiations the Norwegian government agreed in October to comply with the main Swedish demands, whereupon Sweden recognized the independent statehood of Norway. Her future constitutional status was decided in November, this being the only question which was formally put before the electorate. It appears, however, that certain other things may have been in the minds of many voters. Most of the politicians who advocated a republican constitution had been against the demolition of the border fortresses. They had preferred to risk war. The issue confronting the electorate seemed to be not simply that of a monarchical versus a republican form of government; rather, they were asked to choose between a pacific monarchy and a bellicose republic. It was also a question of personalities. The opposition bitterly assailed the Prime Minister, Christian Michelsen.¹¹ At this moment, however, he was a national hero in the eyes of the majority because of the leading part he had taken in dissolving the unpopular Union. And he staked his political life as well as that of his government on acceptance of the monarchy, a fact which must have meant a good deal to many voters.

Still it is also a fact that the results of the 1905 referendum reflect the contrast between radical and moderate attitudes. In most of William Lafferty's 31 localities this contrast seems rather clearly reflected, but in some cases a reservation is appropriate. One of the localities designated as most radical in 1918-1919, the town of Gjøvik, was the headquarters of the republican leader Johan Castberg in 1905. A wish to support the town's favorite son may explain in part the high percentage of anti-monarchist votes recorded just here. Among the 'most moderate' localities, the constituency of Aker near the capital had a relatively high percentage of republican votes in 1905. It is a question, however, to what extent this was an expression of radicalism. Statistically Aker represents a special case in that it formed two separate administrative entities in 1905 - the workingclass East Aker and the middle-class West Aker constituency. After 1905 they were merged into one, but at the November referendum, with the votes separately recorded, there appeared a much higher republican percentage in West Aker than in East Aker. It may well have been a manifestation of strong nationalism rather than radicalism. Yet on the whole Lafferty's assumption of a 'clear radical/ moderate test' does not seem too wide of the mark.

His assumptions regarding the alcohol referendum are more doubtful. For one thing, economic interests influenced the outcome quite considerably in certain districts, e.g., in the fishing districts of the North. But their inhabitants did not play a very important role in the Labor movement during the period in question; we are therefore not unwilling to follow Lafferty when he treats the referendum as if it were solely a test of the strength of conflicting normative attitudes. A

correct evaluation of the meaning and significance of these attitudes, however, is of real importance. It was not simply a question of teetotalist leaders trying in vain to influence the rank-and-filers. Lafferty is mistaken in his assessment of the temperance movement, it seems to us, because he has failed to note the pronounced regional variations in attitudes toward cultural norms. We shall revert to this question below.

d. So far we have considered Lafferty's presuppositions. Although they were not always quite convincing, we had no decisive objections to make except on the last point regarding the assumption that attitudes to alcohol prohibition can be understood without reference to regional traditions. This question is related to the question of Lafferty's statistical analysis, which he does not seem to carry far enough (but then, admittedly, he presents his result as only preliminary).

On page 309 he presents his product-moment correlations between certain factors he regards as crucial. He finds a positive correlation between radical (NLP) voter proportions and the percentage of votes cast against the monarchy in 1905 as well as the percentage of votes cast against prohibition in 1919. Similarly, there is a negative correlation between moderate (SDP) voter proportions and percent against the monarchy as well as against prohibition. The coefficients must be termed rather small: product-moment correlations of 0.22 and 0.36 in the first case and - 0.28 and - 0.11 in the second case. However, a calculation of averages gives another measure of the same phenomenon. Table II is a shortened version of Lafferty's presentation on pp. 301-302:

In other words: on the average, voters were more strongly opposed both to the monarchy and to prohibition in the 19 most radical localities than were voters in the 12 most moderate ones. However, this does not necessarily mean that the more republican a locality's electorate in 1905, the more it was against prohibiting alcohol in 1919, and the more monarchist, the more friendly to prohibition. It is even conceivable that the opposite was true, in the sense that there existed within the group of 31 a tendency for localities with high republican percentages to have high percentages in favor of prohibition, and for localities with low republican percentages also to have low prohibition percentages. This can be analyzed by calculating the product-moment correlation between the proportion of votes cast against the monarchy and the proportion of votes cast against prohibition in the 31 localities in question. Such a calculation gives a negative coefficient of some magnitude (-.45). In other words, there seems to be a tendency for localities

whose voters had been strongly in favor of the republican movement in 1905 to
support temperance strongly in 1919 (and vice versa) while Lafferty's hypothesis
points in the opposite direction.

4. Regional Distinctions

For our part we should like to emphasize another aspect of the matter. The 1919 referendum revealed a strong difference of opinion between the center and the rest of the country. As the center we designate the capital, Oslo, the surrounding area, and the next largest and third largest towns of Bergen and Trondheim. Only here were substantial majorities recorded against alcohol prohibition, in contrast to almost everywhere else in the country. If we divide the 31 localities selected by William Lafferty into central and non-central localities according to the above criteria, a similar contrast is visible:

The localities in which republicanism is designated as strong (weak) are those in
which the vote for the republic was above (below) the average within the group
of 19 most radical and 12 most moderate localities, respectively.

The attitude toward prohibition seems to have been related above all to the central/non-central difference, whereas no strong connection is apparent with the stand taken on either the issue of the monarchy in 1905 or the issue of labor radicalism in 1918-1919. The following somewhat different arrangement of the same data only serves to confirm the impression:

The geographical distinction made above is certainly much too crude. It is a well-known fact, for instance, that rather striking political differences exist between the provinces of the south-western and the provinces of the northern periphery. In Tables 111 and IV we have lumped together localities belonging to both under the common designation 'non-central. The regional differences have been visible not least in connection with matters of cultural policy; but as indicated by Stein Rokkan in the passage quoted in our introductory paragraph, it is also a fact that the latter kind of contrasts, revealed, i.a., through attitudes to alcohol prohibition, do not seem to follow exactly the provincial boundaries. We shall therefore indicate another scheme of regionalization. Tentatively, we have carried out a classification of the smallest administrative units - the communes - on the basis of voting behavior in the prohibition referendums of 1919 and 1926, as well as the constitutional referendum of 1905.

As a result five different regions appear, the lines which separate them cutting
across practically all of the provincial boundaries. We have given them the following
designations (see map):

1. Inner central area.
2. Outer central area.
3. Coast and outer fjords.

4. Valleys, mountains, and inner fjords.
5. The northernmost part.

Region 1, the 'inner central area', comprises the capital, Oslo, and its nearest surroundings, on the periphery of which there is a concentric area forming Region 2. We have also determined the next largest towns of Bergen and Trondheim to belong to this 'outer central area', which is characterized as a whole by a great deal of political similarity.¹² Apart from these two, the coastal towns are included in Region 3 together with the rural strip of land stretching along the coast from the Swedish border in the south-east all the way up toward the Arctic Circle. The interior of the country forms Region 4, while Region 5 consists

roughly of all the territory (including towns) north of the Arctic Circle, as well
as the coastal strip a little further to the south.

The five regions are distinguished from one another on the basis of the following criteria: At the referendum of 1905 there appears one largely contiguous area comprising those communes in which the percentage of republican votes is above the national average. This, in the main, is our Region 4, 'valleys, mountains, and inner fjords'. Throughout practically all of the other areas, with the exception of a few localities in the extreme North, the vote was overwhelmingly monarchist in 1905, the inner central area taking the lead in this respect. The latter was also

the leading anti-prohibitionist region in 1919 and 1926. In both those years Region 2 was clearly differentiated from 1, in which the temperance vote was lower, as well as from 3 and 4, in which it was higher. Between Regions 3 and 4, clearly differentiated in 1905, there was not much of a difference in the prohibition referendums. In the northernmost area, less homogeneous than the other four, there appears a very strong opposition to alcohol prohibition in 1926 along a great part of the coastline.

The latter phenomenon was a symptom of the importance of economic factors. Because of prohibition, the wine and liquor exporting countries of Southern Europe retaliated against Norwegian exports of fish. Norwegian fishermen, especially in certain parts of the North, suffered great losses and turned against the temperance policy. Losses were also inflicted to some extent on the population of Region 3, the western and southern coast; but here religious sentiment, expressed in one way through the temperance movement, was very strong. This seems to be a main reason why the prohibitionist stand was maintained so firmly. In Region 4, an area of small-scale agriculture, forestry, and some mining, religion was a much less important factor. While the humble, pacific Christians along the coast had upheld the monarchy in 1905, more pugnacious elements among the farming and mining population of the valleys and mountains had opted for the republic and a nationalist anti-Swedish stand. When they voted together with the coastal people against prohibition in 1919, it was less for religious reasons than as a protest against the towns and urban cultural norms as well as the big farmers and their conspicuous consumption.

The percentages of the vote cast within the five regions against the monarchy
in 1905, and against prohibition in 1919 and 1926, respectively, are shown in
Table V. It comprises all the communes of the realm.

In Table VI we have distributed among the same five regions the 31 localities
which were designated by William Lafferty as those sending the most radical and
most moderate Labor delegations to the Conventions of 1918 and 1919.

There is one aspect of the matter on which Table VI gives no direct information: the relative weight of delegations from different localities at the crucial Labor National Conventions of 1918 and 1919. The data concerning the faraway northernmost part (Region 5), which maintained only tenuous connections with the rest of the country, are so fragmentary that they can be disregarded. It can be inferred that the importance of this region at the Conventions was not great. But otherwise Table VI tells us nothing about the relative weight of delegations. Sufficient information is provided by Lafferty's book, however. Within the group of most effective radical localities, delegations from Region 4, with a great surplus of radical votes, played a very important part. The situation was the reverse within the most moderate group. Here localities belonging to the same region were unimportant; they provided only a relatively small surplus of moderate votes. In other words Region 4 represented a radical stronghold within the Labor Party. From Table VI it is apparent that localities in that region had the lowest anti-prohibitionist proportion of referendum voters in 1919. However, the radical faction at the Party Conventions of 1918-1919 also got a substantial part of its strength from delegations representing the central regions, in which a decidedly anti-prohibitionist attitude prevailed, as shown in the table.

We are forced to the conclusion that, at the end of World War I, there existed in Norway two types of labor radicalism, very different one from the other in regard to the cultural norms associated with the liquor question. In our opinion the difference had to do with regional contrasts. William Lafferty sees it from another angle, however. He is aware of the fact that there were two sharply

divergent views on the liquor issue within the labor movement, but he regards
this as a case of disagreement between leadership and rank-and-file. He writes as
follows:

Based on the general adherence of the Labor Party leadership to temperance values, it was assumed that this factor would render results in the direction of what that leadership thought (or rather hoped) the drinking issue implied; i.e. an attempt to establish working-class status through the abstention from 'debasing' habits. There is no doubt that they preached this line and it seems likely that later on in the 30's their views actually began to take hold farther down in the party membership ranks. But in 1919 the issue seems to have been affected (in the radical areas) by a force more powerful than status consideration; i.e. norm loss. A consciousness of status differences implies enough stability to, first, know what the general ranking of other statuses in the community is and, second, to care whether or not you have attained a certain status level... but hi the radical areas... it seems that the drinking issue revolved around more than status; i.e. the staving off of anxiety associated with discontinuous cultural norms.¹³

The conclusion which Lafferty draws is no doubt appropriate with regard to some radical areas. In places where newly established industries attracted a large number of young workers, rebellious political attitudes seemed to be combined with a rejection or loss of the cultural norms of the surrounding rural society. But there were other places of quite a different character in which radical attitudes had long been prevalent, for instance, certain old mining districts whose population remained in close contact with the tillers of the soil. Here traditional cultural norms were deeply imbedded - which does not mean, however, that the people were politically conservative. On this point Lafferty has not followed up his excellent idea of tracing earlier left-wing strongholds through an examination of the 1905 referendum.

Lafferty points to the Trondheim area (Trøndelag) as the single most important area of labor radicalism. He also underlines the leading part played by Martin Tranmæl, chairman of the South Trøndelag provincial party branch, when the radicals took over the direction of the Labor Party on the national stage in 1918. Tranmæl was a strict temperance man, but he was far from being alone in taking that stand. In a large part of the area which he came from, this was the normal thing. In the important Trøndelag mining district of Røros-Aalen, for instance, one of the earliest strongholds of labor radicalism, where a compact majority voted against the monarchy in 1905, there was also strong prohibitionist sentiment, 82 per cent of the valid votes being cast in favor of alcohol prohibition at the 1919 referendum in the commune of Røros and no less than 95 per cent in the commune of Aalen. The inhabitants of these areas held fast to the norms of their forefathers. Stern and frugal they had to be in order to eke out a living on the barren mountain plateau or in the snow-covered valley, whether as miners,

as cottars, or sometimes as a combination of both. Most of them were fully convinced that temperance was the only right thing.¹⁴ They rejected the lifestyle of modern city people - of merchants, shipowners, and factory owners, and of the prosperous farmers who were prone to imitate them - drinking imported wine and liquor and tempting poor people to ruin their economy by doing likewise. To the mind of a man like Tranmæl, the consumption of alcohol appeared as an essential expression of 'corrupt capitalistic society'. If the rejection of traditional rural norms was a driving force behind the radicalism of some groups of workers,¹⁵ there were others of whom it might be said that the affirmation of old lower-class norms, characteristic of their part of the country, provided a driving force. And perhaps it was no less potent.

5. Postscript

After this article had been written, William Lafferty's second book, a sequel to his Economic Development and the Response of Labor, was published under the title Industrialization) Community Structure, and Socialism.¹⁶ It contains a wealth of material on labor movements in different countries and on Norway in particular. An analysis of Norwegian regions, which was lacking in the first book, has now been carried out. Lafferty divides the country into four regions. They comprise (see his map on page 177):

Region A. Six southern and western provinces, plus Bergen.
Region B. Five provinces around the Oslofjord, plus Oslo.17
Region C. Four provinces to the north of Oslo.
Region D. The three northernmost provinces.

As a result of the regional analysis, Lafferty finds that he has to abandon his hypothesis about the connection between anti-prohibitionist sentiment and radicalism. He writes resignedly: '... the temperance question does not seem to be the powerful discriminator... that was originally anticipated'.¹⁸

There is one main reason, in our view, why Lafferty obtains this negative result: in several respects his analysis was restricted to the level of the province.₁₉ As indicated already, we believe that only a regional grouping of smaller localities transcending the provincial boundaries will make it possible to grasp certain peculiarities of politics and culture which were characteristic of Norway during the period in question.²⁰

---

NOTES

1. Stein Rokkan, 'Geography, Religion, and Social Class: Crosscutting Cleavages in Norwegian Politics; in Lipset and Rokkan, eds., Party Systems and Voter Alignments (New York: The Free Press, 1967), p. 415. See also Stein Rokkan and Henry Valen, 'Regional Contrasts in Norwegian Polities', in Allardt and Littunen, eds., Cleavages, Ideologies and Party Systems (Helsinki: Transactions of the Westermarck Society, 1964), pp. 162-238.

---

2. William Lafferty, Economic Development and the Response of Labor in Scandinavia: A Multi-Level Analysis (Oslo: Universitetsforlaget, 1971), (Cf. his article in this Yearbook, 'Industrialization and Labor Radicalism in Norway', Scandinavian Political Studies 7/1972, pp. 157-175.)

3. To be fair to Lafferty we must add that he has presented much more than a one-factor analysis. For instance, he has been careful to point out the political circumstances which contributed in great measure to the radicalization of Norwegian labor (i.a., the country's electoral laws). He has also drawn a well-balanced picture of the economic development. But his account of the significance of cultural factors is one-sided. He has over-emphasized the importance of what he terms the anomic results of industrialization (op.cit., p. 19).

4. Lafferty, op.cit., p. 303.

5. Ibid., p. 317.

6. Ibid., p. 318.

7. Ibid., p. 319.

8. His results are to a certain extent vitiated by this procedure, however. It is true that opinions were rather evenly divided within the Oslo delegation in 1918 (40 moderates, 30 radicals), but there was a large radical majority in 1919 (69 radicals, 21 moderates). Cf. Lafferty, p. 242, note 14, and p. 260, note 6. As for the question of terminology, Lafferty does sometimes use the expression 'most effective radical (moderate)'.

9. Lafferty, op.cit., pp. 307 ff. See in particular his Table 8.10.

10. Ibid., p. 300.

11. Prominent shipowner and politician. Later, founder of the Michelsen Research Institute in Bergen.

12. Unlike the others, Region 2 is not a spatially contiguous area; but in our opinion it should be regarded as a separate entity because, in important respects, its inhabitants had a common attitude as a result of their numerous commercial ties and easy communication with the 'inner center'.

13. Lafferty, op.cit., pp. 317 f. Italics and parentheses as in the original.

14. They did not all possess the strength of character of a Martin Tranmæl, who seems to have hardly touched a glass of beer, wine, or liquor in his whole life. But they all upheld the principle as such, even if there were, apparently, some who modified it to mean that only what had been imported from abroad or manufactured in the town was wholly bad: it might be permissible once in a while to make a little something for home consumption on one's own farmstead.

15. Among such groups were those employed in the construction of water-power plants and factories using hydro-electric power. Some of the plants were located in purely rural districts far from any urban conglomerations, constituting an entirely new element which does not fit into our regional pattern. These were very few in number, however, and their importance should not be exaggerated. (Cf. note 16 below.) But with reference to such localities, whose population structure had changed completely from 1905 to 1919, a comparison of referendum results in the former and the latter year has little meaning.

16. Lafferty, Industrialization, Community Structure, and Socialism: An Ecological Analysis of Norway, 1875-1924 (Oslo: Universitetsforlaget, 1974). As mentioned in our introductory section, Lafferty's starting-point was Edvard Bull's proposition to the effect that a sudden disruption of peasant society and the foundation of new industrial centers near hydro-electric power sources had brought forth a revolutionary working class in Norway. [See E. Bull, Die Entwicklung der Arbeiterbewegung in den drei skandinavischen Landern, in Archiv fiir die Geschichte des Sozialismus und der Arbeiterbewegung, 10 (1922) pp. 329 ff. Original Norwegian text in E. Bull, Den skandinaviske arbeiderbevegelse 1914-1920 (Kristiania: Det Norske Arbeiderparti, 1922), p. 4]. Lafferty comes to the conclusion that this was by no means the whole picture: '... the one category or setting which seemed to correspond most directly to Bull's "disrupted" picture was a grouping of communes which could account for no more than 15 per cent of the total radical vote hi 1921.' (Industrialization, Community Structure, and Socialism, p. 312.)

17. Bergen and Oslo were regarded, administratively, as separate provinces.

18. Lafferty, op.cit., p. 306.

19. Cf. p. 289. Only in some respects and with regard to some regions does he analyze smaller localities, making a number of interesting observations.

20. In particular, a detailed commune-by-commune study of the Trøndelag area seems of interest. It is our intention to present the result of such an analysis in another article.