Resumé

The article is a summary of a research programme utilising aggregate data at the
communal level to explore regional variations in Social Democratic and Communist
electoral strength. These variations have persisted since the party split of
1920. The most important factors explaining electoral strength seem to be
geopolitical, geoeconomic, or organisational in nature. Political developments
preceding the Civil War do not seem to be very significant.

The decision to form a Finnish Labour Party was taken during a conference of workers' associations in Turku in 1899. Four years later, at a conference held in Forssa, it adopted a socialist programme and a new name: the Finnish Social Democratic Party (SDP). In the 1907 election, the first after parliamentary and electoral reform, the SDP received 80 of the 200 available seats in the Finnish legislature. In 1916 it secured an absolute majority with 103 representatives. Relatively speaking, therefore, Finland at that time possessed the strongest socialist party in the world.

The two Russian revolutions - bourgeois and socialist - of 1917 had a profound influence upon developments in the autonomous Grand Duchy of Finland. As early as the spring of 1917 strikes began to be prevalent, particularly among agricultural labourers. Parliament was dismissed in 1917, and the outcome of the new election was that the Social Democrats lost their majority. The Civil War broke out at the end of January, 1918, and by the beginning of May the revolutionary troops ('Reds') were defeated by a government army ('Whites') supported by German troops. The war, as civil wars usually are, was very cruel. About 23,000 Reds died. But of these, only 3,700 fell in battle: 8,400 were executed, while 13,000 starved to death or died from disease in prison camps. Over 3,000 Whites fell in battle, and a further 1,700 were executed. The Civil War was very

clearly a class war: according to the calculations made by Viljo Rasila
(1968, 40-43), 78 per cent of those Reds who fell in battle were non-owners',
while 62 per cent of the Whites who died in action were 'owners'.

With defeat in the Civil War, most of the socialist leaders fled to Soviet Russia. In Moscow they formed in August of the same year the Finnish Communist Party (FCP), which immediately began to seek to build contacts with the workers' movements in Finland. However, despite the Red defeat and the flight of many former leaders, the SDP performed surprisingly well in the 1919 general election. With 80 representatives (cind 38 per cent of the vote) it was in the same position as twelve years earlier. Its strength, however, was soon dissipated further, for when the extreme left-wing radicals and communists failed to win a majority in the party conference of December, 1919, they departed to found early in the summer of 1920 the Socialist Workers' Party (SWP), which became the Finnish People's Democratic League of the 19205. From the outset the latter proved to be a serious rival to the SDP for the allegiance of the workers' associations and the support of the working class. Thus in the general election of 1922 supporters of the working class movement could choose between two socialist parties. The supporters of the SWP will be described as communists in this paper, although most had not or had only partly adopted communist ideology.

1. Design and Data

The immediate purpose of this note is to review some of the factors which may have influenced the division of the electoral support of the Finnish working class movement between the two socialist parties in the general election of 1922. More generally, it is a summary of an ongoing research project upon the contemporary strength and distribution of support of the Finnish socialist parties.¹ In this respect the first election after the split of 1920 within the working class movement is of more than passing interest, for the regional distribution of support for communism and social democracy retains the same landscape as in the early 19205.

The research design is based upon the 484 communes of mainland Finland as the units of analysis: the Åland (Ahvenanmaa) islands are excluded from the analysis. While the analysis is mainly based upon statistical data, use has also been made of the more qualitative findings of historical research. Two further brief points about the research design may be mentioned. First, it is built upon a theoretical frame of reference drawn mainly from the works of some classical theorists of socialism

(Engels, Kautsky, Marx, and Lenin), geography (Auer 1936), and the diffusion of innovations (Hägerstrand 1953; Rogers 1962). Second, the methods employed have been correlation analysis, cross tabulation, regression analysis, and discriminant analysis. Discriminant analysis, however, seems only to suggest and verify the same factors as regression analysis. The latter appears to be more suitable for a deductive and causally oriented research design.

What follows is an overview of the most important findings. As a starting point one must emphasise the long-lasting and serious internal division of the working class movement after 1920. The correlation, for example, between the electoral support of the socialist parties in the 1919 and 1922 elections is as high as +0.96. In electoral terms the split slightly favoured the SDP: in the 1922 election it polled 25.1 per cent of the vote compared to the communists' 14.8 per cent. In 1975 the corresponding figures were 24.9 per cent and 18.9 per cent. The key factors which explain the divison of the electoral support of the working class movement seem to be organisational strength, industry, local leaders, the geopolitical location of Finland, the radiation effect of provincial centres, left-wing social democracy, and the civil war itself. In addition, it is relevant to mention events that preceded the Civil War and the role of the press.

2. The Findings

We may begin by referring to those factors which do not appear to be of
much significance: the press and pre-Civil War events.

Those factors which are related to the tradition of the working class movement before the Civil War (for example, strength of organisations, degree of co-operation with the bourgeoisie in the undemocratic municipal administrations, strikes, and voting activity) do not seem to explain the division more than minimally. The most important exception seems to be the impact of the agricultural strikes in the districts around Pori (see Figure 1). Thus the reasons for the division of the electoral support of the working class movement may more appropriately be found in the circumstances of and changes which took place during the Civil War and its immediate aftermath. One must, however, note that the main electoral reservoir for which both the communists and social democrats strived in the 1920s had already defined itself before the Civil War. The correlation between the support of the socialist parties in 1907 and 1919 is +0.86, and in 1917 and 1922 it is +0.93. In addition, one must keep in mind that most of the factors used as independent variables, for example industrial workers, existed

before 1917, but at that time they were not yet relevant variables to explain
the phenomenon under consideration.

The presence of a communist press does not seem to explain the division at all. Its distribution appears to spread mainly according to the spatial distribution of SWP members: first, one became a member and then subscribed to a newspaper. Before the general election of 1922 there were eleven localities where Social Democratic newspapers were printed. At the same time Communist newspapers were published only in Helsinki, Oulu, Vaasa, Kuopio, and Kajaani (see Figure 1): their total circulation was about 35,000 copies. At the same time there were about 25,000 SWP members. If the spatial distribution of the communist press had been completed by newspapers printed in Turku and Lahti, it would have been tempting to emphasise the role of the press in explaining the division of the electoral support of the working class movement. However, the press may have had some kind of indirect explanatory value: opinion leaders who received information by reading newspapers then diffused that information in discussions with other people (Tarkianinen 1971, 440).

Three factors which seem to have been significant are the nature of local
leadership, industrialism and related socioeconomic factors, and the
geopolitical position of Finland.

Leadership. One sequel of the Civil War was that the embryonic communist movement lost almost all of its leaders of importance, who, to stay alive, escaped to Soviet Russia. On the whole, individual candidates did not have a great influence, probably because of the practice at that time of presenting lengthy election lists. Regression analysis emphasises the importance of candidates in some parts of the country, but their influence is very much illusive. But some individual candidates obviously contributed to the electoral support of the communists. In the Turku district a former representative of the Christian Workers' Union ran as a communist candidate and attracted votes to the SWP. In Vehmaa (also in the Turku region) Emmanuel Ramstedt campaigned so vigorously that almost every leftist voted for the communists: in the early 1930s supporters of the Lapua movement forced him to emigrate (though he moved to Sweden, not the Soviet Union). To some extent, then, there were individuals whose authority and prestige was such that they persuaded their local association to separate from the SDP and join the communist organisation. Correspondingly, there were authoritative Social Democrats who prevented their own association from seceding from the SDP, or at least managed to ensure that the association remained 'independent', that is, it refrained

from affiliating with either of the workers' parties. Overall, however, there were not many authoritative individuals of this stature, and the Finnish working class movement seems to have divided mainly as a mass movement.

Industry and other socio-economic factors. In the 1920s the industrial development of Finland was about fifty years behind that of Sweden. Although the support of the communist (in percentage) and the proportion of industrial workers in Finland at the beginning of the 1920s were about the same (around 15 per cent), this did not mean that communists and industrial workers had a similar spatial distribution. The correlations of the percentage of industrial workers with communist support (+0.05), and with communist support as a proportion of the total socialist vote (-f 0.06), are almost zero. Other social and economic factors do not seem to have played any role at all in the division of the working class movement. However, unemployment and migration evidently increased slightly the chances of greater communist support. Many migrants were still relative newcomers in industrial localities, and so perhaps it was rather easier to recruit them as supporters of communism.

On first sight the linguistic division within Finland might seem to be of importance. But although in the Swedish-speaking districts almost every workers' association stayed with the SDP, it seems to be the case that those Swedish Finns who before the split of 1920 had supported the Social Democrats were more likely to support the communists than continue to vote for the SDP. This phenomenon is probably due to the fact that Swedish speakers lived within the spheres of influence of those centres (Helsinki, Turku, Vaasa) where the role of the communists became very important.

The role that the industrial workers played in the post-1920 alignments of the working class movement was particularly important in those provincial centres with relatively large concentrations of workers and/or larger industrial plants, where there also tended to be a high level of organisation among the workers. These developments prepared the ground for some provincial centres (Helsinki, Turku, Lahti, Vaasa, Kuopio, Kajaani and Oulu) to become points from which communist ideology 'radiated' into the surrounding countryside. There were similar industrial landscapes in Viipuri, Jyväskylä, Kotka, Pori, and above all in Tampere. Yet the political result was rather different. In Viipuri, Jyväskylä, and perhaps also in Tampere, the minor role of communism is to be explained by the strength of left-wing social democracy which emphasised

the necessity of the unity of the working class (see Hodgson 1967, 100-104). In Viipuri and Jyvåskylå the workers were highly organised, but left-wing social democracy tied them very strongly to the Social Democratic Party. The experiences of Tampere and Viipuri during the Civil War were very similar: so it is perhaps probable that the experiences of the Civil War led socialists in the former city to support left-wing social democracy. Nevertheless, the main obstacle to the establishment of mass communist support in Tampere was that less than one year after the surrender of 'Red' Tampere (the battle for the city was the greatest hitherto in the history of Scandinavia) the SDP had gained a convincing victory in the municipal elections, and this without doubt decreased any motivation to organise. Moreover, despite electoral strength, the organisational level of the workers in Tampere was low.

The indirect effect of the geopolitical location of Finland emerges in the fact that after the Civil War the illegal communist channels could immediately transmit incitements to action, which increased class consciousness, either across the eastern border or from the west through Sweden. It is quite possible that the first communist efforts were diffused in Finland from the west via Sweden, and not from the east as one might suppose. In any case, much of the communist stimulus came from outside. As one of the theorists of socialism, Karl Kautsky, stressed, the everyday struggle for bread can create a trade unionist consciousness, but not a socialist consciousness: that must come to the workers from outside the everyday struggle (see Lenin 1975, 65-66). Where illegal channels filtered such incitements to action through to localities with large industrial plants and highly organised workers, the localities became centres which radiated communist ideology into the surrounding countryside - but only if left-wing social democracy had not retained a strong foothold in the centre.

We may finally turn to those explanatory factors which seem to have the
most significance: the civil war itself, the political and economic organisations
of the working class, and the radiation effect of centres.

The Civil War is a very important factor in the cause-equation of the divison of the working class movement, when the whole country is considered as a single observation unit. However, if the same phenomenon is considered areally, one can find some interesting variations. There are many communes where one would have expected strong communist support because of their experience of the White terror during the war. On the other hand, there are to be found strong communist communes which had

been totally unaffected by the war. Ecological quantitative data are unfortunately not yet available for the numbers of Red guards executed or held in prison camps: these should improve the analysis considerably. In addition, the location of the front line during the Civil War does not seem to explain very much. Proportionately, there were obviously many communist votes on the White side of the front, but in absolute numbers the distribution of the votes is almost equal between the two sides of the front. The correlation between the number of Reds who died in the war and communist support as a proportion of the whole socialist vote is high (-0.39). The correlation of the number of Reds who died with communist support is -0.10, and with Social Democratic support +0.49. These correlation coefficients support the views, put forward mainly by the communists, that White terror restricted the possibilities of communism by destroying its probable future local elites. Overall, the Civil War seems to be a kind of background variable which within the limits of the whole country explains to a considerable extent why nearly one-third of the socialist voters (29.9 per cent) supported the communist movement in the 1922 election. But other variables - some of which are only weakly connected to the Civil War- more accurately explain where the support of the two branches of the working class movement was located on the surface of this wide observation unit (that is, Finland).

The political and occupational organisations of the working class movement were of great importance. This importance was twofold. First, the more highly organised the workers, the easier it seems to have been for the communists to mobilise them, particularly in the more industrialised communes. On the other hand, when the communists had gained possession of an organisation, it radiated communist ideology to its immediate hinterland. Regression and discriminant analysis, and correlation coefficients clearly indicate the importance of organisation. In those communes where the communists managed to win a monopoly position in the organisations of the political working class movement, they polled on average three-fourths of the total socialist vote. Before the 1922 election there were local communist organisations in about 40 per cent of the communes.

The radiation effect of the provincial centres. After the Civil War Helsinki, Turku, Kuopio, Vaasa, Lahti and Oulu became centres which radiated communist ideology to their respective spheres of influence. In all of these towns industry was important and the workers highly organised. The radiation effect of these centres 'explains' nearly every communist representative who was elected in 1922. Perhaps only three out of 27 representatives are not explained by the radiation effect hypothesis,

two in the Pori region and one in the western constituency of Viipuri. In Pori the explanation can be found in the events of 1917-1918 and a prominent communist candidate: the bourgeois civil guards around Pori (especially in Satakunta) were, as early as 1917, very clearly hostile to the workers, in strong contrast to other parts of Finland. In the western constituency of Viipuri the communist representative is explained more by the presence of industrial workers. How did communist ideology spread to the surrounding spheres of influence? Most important was the diffusion effect of individual contacts; but the further away a commune was from the centre, the more important the industrial workers and their organisations became. Communism mainly spread in the patterns indicated by the curves around the diffusion of invation (see, e.g. Rogers 1962). Deviations from these curves are usually easy to explain if one considers factors associated with the organisations of the working class movement.

3. Conclusion

As a result of the division of the working class movement at the beginning of the 19205, communism and social democracy in Finland espoused those spatial features of electoral strength which are still easy to observe today. Why has the regional distribution of support stayed so firm for almost sixty years, even for example across the divide from 1930 to 1944 when communism was illegal? When communists polled over fifty per cent of the vote of the Left in 37.5 per cent of the communes, it is obvious that in different parts of the country there were areas where organisations and political tradition nourished and protected the continuity of communist support. During the period of illegality, the secret organisations were most powerful in those districts which both before and afterwards possessed a large communist vote. In Sweden the areas of relative communist strength are identical with those of overall leftist strength. In Finland the firmest support of communism is usually found in regions where the total support of the Left is weak.

The most important factors influencing the division of the electoral support of the Finnish working class movement seem to be organisations, industry, and the radiation effect of provincial centres. These are followed by a further cluster of three: the Civil War, the connections of the Finnish Communist Party with Finland (through these connections came both instructions and incitements which aided the development of consciousness), and the geopolitical location of Finland. There are not many pre-1917

factors which affected the later division of the working class. The easiest way for the exiled radical leaders and the newly-formed Finnish Communist Party to influence events in Finland was to seek connections with industrial regions where the occupational and political organisations of the left had been almost total. Through the organisational structures the communists could win over the workers and so build new organisations to radiate communist ideology to the surrounding districts. One decisive factor in the struggle between the communists and Social Democrats was the control of the provincial centres. When a provincial centre was occupied' by one side, the almost automatic consequence was the support of most of the workers in the surrounding countryside. The overall effect of these factors was that the communists fared more badly than the Social Democrats. In the traditionally firm Social Democrat regions the communists managed to occupy only one centre, Kuopio (see figure 1). On the other hand, the SDP succeeded in holding its preeminent position and at the same time the support of the surrounding countryside in Pori, Jyväskylä, Tampere, Hämeenlinna, Mikkeli, Joensuu, Sortavala, and Viipuri. This success guaranteed the SDP over one-half of the left representatives in the 1922 election.

Erik Allardt has written several articles which deal with the mass support of communism in Finland during the 19505. The essence of his views and theoretical generalisations are summarised in 'Types of Protest and Alienation' (Allardt 1970). It seems that the regional variation in Communist support upon which Allardt's typologies about diffuse and institutionalised deprivation and different types of alienation are partly based are perhaps explained primarily by political traditions maintained for the most part by local communist organisations. This question cannot be resolved here for we do not have any variables on organisations in the 19505. Although Allardt emphasises the role of organisations in the development of 'industrial communism', he does not include any data on the local communist organisations. Also Nousiainen (1956) and Rydenfelt (1954) seem to pay too little attention to the role of organisations: they both emphasise social and economic factors in explaining the regional support of communism, probably because they see organisations as part of or so closely connected with elements of political support that they cannot be used independently to explain the latter. This point of view arises perhaps from the fact that in political sociology the main interest has been in the social basis of political phenomena.

The thrust of our research programme, by contrast, is towards causal
relations (see Nurmi 1974). In this sense it seems that Allardt's distinction

between Southern and Western Finland on one hand, and Northern and Eastern Finland on the other, is not entirely satisfactory. On industrial communism Allardt speaks of institutionalised relative deprivation, and on backward communism, diffuse deprivation. Undoubtedly, this interpretation explains in some measure internal distinctions in Finnish Communism in the 19505, but for the 1920s it probably loses its significance. The argument by which Communism after the Second World War lacks tradition in Northern and Eastern Finland seems strange because it was explicitly in Northern Finland - in and around Kuopio - that the communists, as early as the beginning of the 19205, took under their wing almost all of the organisations of the old unified working class movement. Many factors other than the purely social and economic seem to have determined the division of the working class movement outside the provincial centres, and these determined the social environment in which communism was located spatially.

One must agree with Allardt in one important respect: in Finland the support of communism is very much a spatial phenomenon. Analyses which try to explain its mass support while ignoring the spatial context will very likely fail. At any rate it seems to be correct to take as the starting point of an ecological mapping of Finland the electoral support of every party in their electoral 'debut' and the factors which affected that support, before moving on to study the changes over time in the spatial support of each party and the factors producing those changes.

---

NOTE

1 The research programme has been financed mainly by the Finnish Academy in Mikkeli. The computer runs were carried out at the University of Turku. The analyses are expected to be concluded in 1979, with subsequent publication in Finnish (including an English summary).

REFERENCES

Allardt, Erik, 1970. 'Types of Protest and Alienation', in Erik Allardt and Stein Rokkan
(eds.), Mass Politics. New York.

Auer, Vainø, 1936. Suomen asutuskeskusten talousalueista (German referat: iiber die
Wirstschaftsräume der Siedlungszentren in Finland), Terra 48:3. Helsinki.

Hägerstrand, Torsten. 1953. Innovationsförloppet ur korologisk synpunkt. Lund.

Nousiainen, Jaakko, 1956. Kommunismi Kuopion läänissä (Communism in the Province of
Kuopio). Joensuu.

Nurmi, Hannu. 1974. Causality and Complexity: Some Problems of Causal Analysis in the
Social Sciences. Turku.

Rasila, Viljo. 1968. Kansalaissodan sosiaalinen tausta (The Social Background of the Civil
War). Helsinki.

Rogers, Everett M., 1962. The Diffusion of Innovations. New York.
Rydenfelt, Sven. 1954. Kommunismen i Sverige. Kristianstad.

Tarkiainen, Tuttu, 1971. Eduskunnan valitseminen 1907-1963. Suomen kansaneudustuslaitoksen
historia, IX (The General Elections 1907-1963. The History of the System of
Representation in Finland). Helsinki.